So, my uncle needs a new computer and he needs Windows 10 on it. How can I make sure that he is not being spied on and psychologically manipulated by ads right from the beginning? Is there a professional version of Windows 10 with none of those shenanigans I keep hearing about?
Not really, at least not for consumers. There are 'decrapifier' scripts you can run, but all that shit just gets re-installed on the next major build update anyways in my experience.
Go to Settings -> Privacy, and slide many (most?) things to off. Including Diagnostic & Usage Data to Basic.
Settings > Personalization > Start and slide "Occasionally show suggestions in start" to off.
Settings -> Personalization -> Lock Screen and slide "Get fun facts, tips, and more from Windows and Cortana on your lock screen" to off.
Settings -> System -> Notifications & actions -> "Get tip, tricks, and suggestions as you use Windows" and "Show me the Windows welcome experience after updates and occasionally when I sign in to highlight what's new and suggested" to off.
Go to Cortana (Taskbar) -> Settings (Gear Icon) -> "Taskbar tidbits" to off
I wouldn't use the built in "Mail" app with or without ads, it is terrible. Just have him simply not use it.
Be careful with those scripts/software. Many of them essentially break Windows 10 by design.
If you stay within the Settings UI then you are working within Windows 10's supported design. When you run a script that disables a system service (e.g. Cortana) things start to get quirky/broken.
Several of those break Windows Update for example.
Not the biggest fan of 8 series, but the best thing about it was some of the energy optimisations - as they were going to target mobile. The kernel hibernation is good, for quicker start up. But sadly that's at the expense of hibernating proper.
8.1 brought back things like the closing button on metro apps, and some other fudging, to make it a bit more Win 7, and it just felt like even more of an abomination. The control panel/metro settings mixed UI - is the thing that killed Windows for me. In Windows 10, I can't find anything setting wise particularly easily. And on a recent new laptop start up, I cringed at all the options that I had to turn off, and at the apps I tried to remove.
Here's unpopular opinion: start menu really did need a remake and I like the designs 8.1 had compared to 10. Also, IIRC 8 has some optimizations (including some tweaks to better support ssds).
I wonder if it's possible to use it with modern hardware. May be someone will make a modern Windows 2000 distribution with backported drivers and fixed vulnerabilities. Source code was leaked, so should be possible.
As far as I know, the Windows 2000 source code leak was incomplete, actually; or if it isn't, then not the entire thing ended up getting leaked. For example, the code to the setup program was missing.
To the best of my knowledge, the most recent complete leak was NT 4.0. That one did have a short-lived burst of retro affectionados under the name "OpenNT 4.0", though it's not entirely clear if it died on its own or if a cease & desist helped that along.
I liked 8.1 because they compartmentalized the new junk—with a start menu replacement app (of which there were many), you could get away with never touching the Metro apps and just using a faster, cleaner, more secure version of Windows 7.
I'm going to keep running it until the SSD it's installed on melts, and burns a hole in my desk! Win10 is a monstrosity and I loathe having to support all my customers who use it.
This wouldn't be bad if I could uninstall that app but the button is gray on my machine. Same for the Xbox apps that I am guaranteed to never need. Weird feeling.
Having to do this has in the past been the main argument I've seen used against using Linux based OSes. If you have to do it anyway you might as well actually own your computer.
Microsoft continues to demonstrate that it's worst enemy is itself. There's no way that throwing ads in the default Mail app is going to make them more money than it costs them in the long run.
While the ads and bloatware Microsoft was specifically trying to get away from in the 2000-2010ish era coming back, but this time packaged by Microsoft themselves, is absolutely baffling, the Cortana voice playing when you install Windows 10 has me even more confused.
How did that get integrated? Is there some sort of Microsoft VP who decided to stake their career on Cortana and overrode everyone to merge that awful design decision into the installer, or do Microsoft employees like it as well?
Who ever wanted to run a Windows installer via voice? Has anyone ever actually used the voice installer?
For other cases, I'd consider filing feedback requesting an easier way to mute during the setup process (if there isn't one already; I don't remember looking for that option and instead just muting the device using a hardware switch).
To be fair, the only serious competition for home users has laptops that (now) start at $1200, plus tax, applecare, and dongles (oh, you wanted a USB-A connection like basically every device in the world uses? Jony is about to courage your wallet. You want to be able to connect it to the odd tv via HDMI, what everything in the world uses? More courage!)... so basically $1500 to walk out the store. I'm aware you can get an old air with 5 year old guts for $1k, but that will prolly not be sold for much longer.
My parents do not need $1.5k of laptop to browse the internet, email, and exchange photos of grandkids. Particularly in a world when a decent laptop doing the above should cost $750 or less out the store.
My parents live in rural America, and the internet connection is not good. It's manageable for email and browsing, but even the school has problems with chromebooks and their bad internet connection.
Plus... if you're fleeing Windows because of the advertising, you're gonna go to a google product? How on earth does that help?
At the same time, how many cheap Windows laptops are only so cheap because they come with a lot of pre-installed bullshit?
Also, 5 year old Air for $1k? That's almost how much they cost new. I'm seeing half that on Ebay/Craigslist. My sister has a new $450 Toshiba and my 5 year old Air runs much better. I got a new battery for it and gave it to her.
Unfortunately you went way too far with your rant and buried your point about Windows' real competition being more expensive. But the low end of Windows laptops is pretty atrocious. Someone stuck buying those aren't doing it because Airs lost their SD card port. They're doing it because they can't afford anything other than a computer that serves them ads, has 3hrs battery life, and struggles to buffer 1080p video on Youtube like my sister's new Toshiba.
In this case, I don’t think that’s fair. A great number of people can’t devote the time to learning a new operating system or cannot afford a high priced OS X machine.
> A great number of people can’t devote the time to learning a new operating system
Can't or won't? My experience has been that several people I know would rather stay in their comfort zone rather than learn something new. I'm not just talking about OSes here. I qualify this as a won't.
I suspect this reluctance to switch from the status quo might apply to a wider population than my acquaintances.
Can't or won't is a distinction without a difference. Desktop/laptop computer ownership is flat and has been for the past decade [0] with smart phones and tablets growing [1].
What's the use in learning an entirely new operating system on your desktop whenever most of your time (65%) is being spent on Mobile anyway? [2]
What is there to learn on a Gnome/Ubuntu distro? There's a button you click to open firefox/writer/calc and a button you click to open the (arguably inferior to wpa_gui) wifi network GUI. There is a finder/explorer like shell on pretty much any "popular" DE. External drives are auto mounted and there's better point and click support for pretty much any way to move files over the network.
There really isn't any learning unless you're a power user, and those are the people that can handle learning.
And even if there was, Microsoft moves their UI around often enough that people have had to "learn a new OS" at least three times by now.
Pretty much everything made by Adobe and Microsoft (but I can sort of live without the ones from Microsoft). Honestly, if Adobe made a Linux distribution as part of Creative Suite (which I would think wouldn't even be hard for them to port to Linux) I bet they could make a strong running taking over the desktop operating system market.
Ubuntu not only came (or still comes?) with an Amazon app right on the launcher, but also offered Amazon product suggestions when you were using the search function.
I am sure that providers with paid products like Netflix or HBO are watching closely. Until now the deal was that the customer pays to get rid of ads but maybe they can make the customer pay and still show ads?
I don't think that Microsoft serving ads within their native mail application has strong ramifications for Netflix or HBO. They are different market segments with different expectations.
I’m fairly sure Hulu has done this from the beginning, right? I never understood the allure of paying for Hulu but still seeing ads. I just went to check and apparently you can pay extra to get rid of ads.
Recently Amazon has been playing ads at the beginning of shows that are offered for free via Prime. I feel like I even saw them when I had bought a season of a show, but it is now offered for free via Prime.
So far it seems like buying shows and movies through Apple is ad-free.
In practical terms, any software engineer knows how to work in a POSIX environment, so no issue here.
>setting up
If you mean "installing Linux", then any mainstream distro (e.g. Linux Mint) is ready to go out of the box. If you mean "configuring things up to your liking" then it is no more time-consuming than any other setup.
>operating
Don't know what you mean by this, but a POSIX environment is probably the most efficient/least painful way to do software development.
What does HN recommend as a better Win10 email client? I'm familiar with Thunderbird and I'll probably switch to it if nothing else, but I'm not a huge fan of the UI/UX.
This might sound glib, but as someone who has used both Thunderbird and Outlook extensively, I don't recommend any desktop email client. Email clients were designed for a whole different era, and only work well in that context.
Gmail, Outlook.com/O365/etc, are designed for you to "keep everything." Problem is that all desktop email clients aren't designed that way, they want you to have a very small limited subset of email that they can index and constantly have in memory.
When these two worlds converge desktop clients choke and choke hard. Outlook is so terrible that it almost needs to be completely retired. They've barely updated it in ten years.
Best case scenario right now is full webmail (with anyone) and browser notifications for new email etc if you need that.
> Problem is that all desktop email clients aren't designed that way, they want you to have a very small limited subset of email that they can index and constantly have in memory.
with large amounts of mail (~90k in one of my inboxes - and if someone asks just yesterday I had to go and fetch some info from a 2007 mail), GMail chokes much more than thunderbird, e.g. it takes 2/3 minutes and sometimes does not work to mark a large set of email as read/archived/whatever while this is almost instant with TB.
clients still make sense for users with multiple accounts. and modern clients only sync a subset of the db but still permit online search and retrieval. also for certain tasks the web clients aren't much better (gmails interface is garbage for working with multiple messages for example)
the default answer to op's question in enterprise land (where windows is still treading for relevancy) is Outlook.
thunderbird suffers the issues the comment mentions above -- barfs with large dbs, no support for exchange, not sure if it even supports push notifications, etc
it should be noted that mail on mac works well and the win 10 mail app this article is all about works pretty well too, which is why its a travesty MS is going to ruin it.
> Problem is that all desktop email clients aren't designed that way, they want you to have a very small limited subset of email....
I have over 20 years of email stored locally in Postbox. I've kept every non-spam email since I first got online in 1997. I regularly use it to search for things people sent me (or I sent them) months or even years ago. It's really not a problem at all.
I've got Postbox open right now and it's using 625MB RAM. (That's a lot compared to the Eudora days, but I have 16GB RAM now.)
A desktop client on your laptop also means you can write emails & be able to search your email history while in flight mode, and queue the messages to be sent when you land & have WiFi again.
I remember that more than 15 years ago it was a pain to set up thunderbird to be closed in the systray and display a notification when new mail arrived. It wasn't a core feature and you had to mess around with extensions or binaries that would just check out your mail and launch thunderbird (but couldn't open the mail mentioned in the notificatin).
15 years later and 3 major windows iterations later and we still don't have it.
Possibly Postbox, it's a commercial app but derived from Thunderbird. I've been using it for years on Mac since Eudora was discontinued, and plan to use it on Windows now that I'm switching back to PC this month:
For me, email is absolutely mission critical. $40 is an absolute steal for the software I use the most in my business & spend lots of time in every day personally. It's cheaper than Sublime Text too.
If you do feel that's a lot, try the demo first to see if you like it, and maybe wait for Black Friday in case they have a sale.
Ten years ago I would not have said this, but Outlook is really pretty solid. Pretty fast, uses hardly any RAM, excellent Calendar integration (big deal for me), extremely flexible account setup.
I swear half of the clients on the market turn out to support only a small list of authentication methods, or don't let you specify port!
It is of course overkill though if you have one email account and only process a couple emails a day.
Sorry, but Outlook 2016 is broken compared to 2013. I am forced to use it at work with exchage, and it has weird bugs that simply did not exist in the 2013 version.
i still dont think it supports the iCalendar/calDAV standard either.
They support it, but in a very strange and broken way (lots of MS-specific fields with poor documentation). IIRC from my time trying to do appointment sending from line of business app, if you send someone an ICS attachment with the right MIME type, Outlook treats it as a regular appointment request.
Thunderbird works very well for me and with the new (to me?) CalDAV/CardDAV plugins, it is able to sync with my remote repo seamlessly for appointments and contacts in addition to email. I recommend it highly.
I use Mailspring which is available for Windows, Mac, and Linux. I use the free version but they have subscription model with many features (which I don't necessarily need).
Is it at all ironic that I get a page that is trying to force me to turn off my ad blocker when I am trying to view a link about Microsoft putting ads in their Mail app?
That's probably true, I haven't touched the AdBlocker extension in a while and I get anti-ad block modals all the time. Probably time to look into an updated list with a working anti-ad block blocker
The latest Gmail revamp is more touch-friendly than it used to be, especially in Edge (despite Gmail's increasingly obnoxious insistence that you should only use Chrome).
The article suggests that the Mail app will never have ads for Office 365 users. Office 365 Home is often a good deal for a Surface owner anyway, as it boosts your OneDrive space, and gives you up to date copies of all the Office programs, and stuff like that.
> despite Gmail's increasingly obnoxious insistence that you should only use Chrome
Fuck Edge touch support - they don't support the event API used by Safari/Chrome, instead you need to use an MS specific touch API.
I support touch on IE11/Edge (because our business made a bet on Windows Phone) but I have wasted an obscene amount of time to do so (we have very few windows phone/laptop touch users).
I hate supporting Edge more than IE11 - weird bugs and continual problems with the debugger and touch support.
Are you talking about Pointer Events versus Touch Events?
Edge supports both (and has for some time). IE supports only Pointer Events. (There is a polyfill for Touch Events on IE: https://github.com/WebReflection/ie-touch)
Both are W3C recommended standards. Pointer Events has more capabilities, and an easier abstraction for multi-modal input (ie, mouse and touch, or touch and pen). So I often see Pointer Events recommended over Touch Events if you can use Pointer Events. (React recently moved to recommending Pointer Events, as one example.)
The big reason not to use Pointer Events is that Safari/WebKit has lagged far behind on Pointer Event adoption (it is finally in experimental preview in Safari, years after Firefox and Chrome/Blink, and seems to be part of the increasing Blink/WebKit split). There is a Pointer Event Polyfill for Safari (and other older browsers that support Touch Events but not Pointer Events): https://github.com/jquery/PEP
Hmm... I'm an Office 365 customer, so I guess I'm not getting ads then. Good.
Also, I love Edge to death (it's the only browser I'll use when I'm on my Surface), but I forgot another reason why I don't want to deal with Gmail in the browser: Gmail doesn't let me have my personal and work email open at the same time.
OSX: Pay here. Not enough, we increased our prices! Oh, don't forget your dongles!
Windows: Oh you want to rename a file that is opened by an application? Fuck you. We're not even capable of telling you what application. Have some ads instead!
Linux: Ooh you thought you could use $APPLICATION_NAME? Sorry, that isn't available, but you can try $MEDIOCRE_ALTERNATIVE instead.
I agree. OSX user here as well. It's the "least bad" option IMO if budget isn't much of an issue.
If OSX wouldn't exist, I would likely try Linux, get sick of it, try Windows, get even more frustrated, and get back to Linux. It is very likely I'd be using ElementaryOS or Ubuntu Desktop in the end.
As a Linux user, I find macOS infuriating. As soon as you change a setting, your chances of something breaking increase dramatically. The window manager is sometimes okay, but other times infuriating (maximize vs. zoom vs. "fullscreen" vs. actual fullscreen), and has major bugs in every release ranging from funny (the underglow for windows in Exposé being somewhere other than under the window you're hovering over) to day-ruining (hard killing the application you're switching to, sometimes [saw this with 10.11 a few times]). Worst of all, they insist on changing everything every couple years, and given the kinds of bugs they tend to introduce while doing this, I suspect they are repeatedly rewriting the whole window manager, almost from scratch, for seemingly no reason.
Now, that's not to say that X11 desktop environments are all sunshine and rainbows but at least a) if you like it the way it is, you can probably keep it exactly as it is with no consequences, and b) you can always choose something simpler, where there's less to go wrong.
I went with option b here, I've been using mostly the same version of the same lightweight window manager since 2012, it is just fine, it isn't ever sluggish on any machine, and it isn't "too little" for my workstation. I use it on OpenBSD as well, and when I tried out MINIX 3 in 2013, it wasn't any effort to use exactly the same window manager. There are no commits in it (mine is a fork, but it didn't really need to be) since February 2016, because it didn't need to change.
So crazy how far behind Windows is to OS X. Even basic GUI stuff (animations, smoothness, etc) are not even a comparison. This is just another example. Ridiculous.
But seriously, I still refer to even SSDs as hard drives. Spinning rust or memory chips. They're interchangable right now, until spinning rust phases out completely.
Anyone remembers just 5 years ago, Microsoft had a huge campaign about how everyone got "scroogled" by Google? Especially this ad about Google reading your mail to devliver you ads? https://www.youtube.com/watch?v=iI1ominSL_c
Well, everyone thank Satya Nadella, it's a miracle we were able to draw the line somewhere! /s
I'm sorry, this snark is not really directed at you, but how exactly are they determining your interests if they're not reading your e-mails? Where does that data come from? I'd be happy to take back my /s if someone can tell me it was all voluntarily collected, but I suspect the answer is really "they just found a different way to spy on you other than reading your email."
There are plenty of ways to spy besides email... Bing searches, for instance. I believe anything you type into the start menu is sent to Bing by default, so that could be more data than you think.
I personally use Bing, but not Windows. Looking on my interests page for Bing, I see they have a decent chunk of info on me, nothing too crazy personal though.
There is still no evidence that they are reading your emails so it's not the same. I am fine with ads, just as long as they don't mine my data and follow me around.
Duckduckgo does not. Hacker News does not. There are many examples. Ads on blogs such as Daring Fireball and Coding Horror do not. I gladly whitelist these sites.
Sure there are fringe cases where it's not true. In substance ad platforms are now tracking & targeting mechanisms. That's the "value" they sell, not their ability to deliver ads.
Really? You can do the experiment yourself by sending certain emails and eventually getting ads target off what words were in them. I don't think this was controversial.
All I want is to be able to give companies my money in exchange for not being spied on, manipulated, or locked into some walled garden. I can't for my OS - I have to use free gnu/Linux because everyone else treats me like crap.
Is my money not good enough for them? This is getting absurd
It seems like they can get people to sell their privacy and autonomy cheaply if they can convince them it's the normal thing to do. Having an option that doesn't do any of those things would undermine the illusion.
The money you would likely pay would be less than what they can make by selling you ads (as a whole). If you were willing to pay > $1000 for software and services there was no reason for companies to maintain an Ad business.
Is that really true? I feel like they would need _everyone_ to be willing to pay that much, and on a recurring basis somehow to make support sustainable.
The value of marketing to Windows users is in the network effect. If one Windows user can pay $1000 to avoid advertising and change nothing else, the other users will probably hear about this and take notice. But if every Windows user is subjected to ads, I very much doubt that Microsoft is making $1000 in advertising revenue from just one person.
Microsoft would never convert 100% of users to "ad-free" paid subscriptions, at any price. But they can make money directly from advertising to 100% of their users, and there is no "loss" except from attrition (which I would guess is still probably quite a bit smaller than their numbers for overall user-base growth.)
> The money you would likely pay would be less than what they can make by selling you ads (as a whole). If you were willing to pay > $1000 for software and services there was no reason for companies to maintain an Ad business.
So people used the same argument with games and loot boxes. But even if videogames cost $500 a copy, there is so much money to be made with loot boxes that they would still exist in these games at that price.
If you were willing to pay > $1000 for software and services there was no reason for companies to maintain an Ad business.
I don't think this is true, but if it is, please provide a link to prove the assertion.
I doubt that companies are selling software at a loss with the intent to make it up in ad revenue. They're selling the software at a profit, and then getting ad revenue for additional profit.
All this just to "maximize revenue potential" for a bunch of faceless investment funds.
> . If you were willing to pay > $1000 for software and services there was no reason for companies to maintain an Ad business.
Let's be honest though, if a company can make money doing something they most likely will do it. It wouldn't matter if you paid $100,000 a month to use windows, if they could make an extra $1000 selling your personal information and pushing ads at you they will.
Ubuntu did infest their OS with ads (via amazon). Last I checked they still did but backlash caused them to disable it by default. I can't speak to apple's data collection and ad practices but I'm guessing they've been doing some form of both.
maybe they could compare my purchase history from various services along with the ads they've served, and at some point realize that I've never bought anything from an ad, nor ever bought anything, ever, that's been promoted through these ads, then finally stop showing them to me. It would be better experience for me, and also make their clickrate ratios go up for their advertisers.
> All I want is to be able to give companies my money in exchange for not being spied on, manipulated, or locked into some walled garden. I can't for my OS - I have to use free gnu/Linux because everyone else treats me like crap.
AFAIK a license is paid with every new computer coming WITH Windows, so it isn't like people get Windows for free. I'm not sure it is even possible to transfer Windows's license from computer to computer anymore, so why is Microsoft doing this? just because they can. Imagine tomorrow some f\*cking ads pop up each time you open the start menu, or on the background of the desktop, or on the login screen, or in Visual Studio Code because why not? Absolutely outrageous. "Pay for the product and you're no longer the product they say". and while we are at it, why not display ads on websites hosted on Azure even when you paid for hosting?
What do you mean tomorrow? When I open my start menu today I see ads for a bunch of games and apps that I don't have installed. When I see my lock screen I see a pretty nature photo with a couple of clickbait phrases on it like "If you think that this place is peacefull, then..."
While what you say is true, the gp's point is that Windows is already pre-installing crapware. The Start Menu icons were still pre-installed, even if you can remove/uninstall them and even if the apps they represent weren't pre-installed.
The first time I opend the start menu on Windows 10 after mostly using linux, it included news about Kim Kardashian. Since then, my thoughts have been sealed about Windows.
Same thing! After using Linux abroad I started a new job using a Windowns 10 workstation and it showed me a picture of the President of the United States. It took a while to get the menu cleaned up too.
I used to work for Microsoft on Edge Browser. Our tests ran on a clean version of Windows. Seeing the ads pop up in start menu, MSN homepage on Edge, little squares ads on bing page made me a bit miserable, day by day.
The default experience was garbage thrown in your face.
The leadership had to drive revenue growth and msn/ads were a money maker. That’s how they get promoted and we had to play their game.
Seeing that ladder chasing culture made me quit.
Very few companies actually legit give a shit about their users and the user experience. It’s so easy to blindly do things that annoy users but make a quick buck.
OOBE on Windows 10 has gone to the dogs. Live Tiles on the default Windows 10 Start Menu — especially if you buy an OEM laptop — look gaudy and distract you from the task at hand. I just wish Microsoft realised that this is making Apple’s systems more attractive.
I've been a member for a while now. And we even ran a domaition drive for a local food nonprofit, and our COO would match to a nonprofit org we said. More to the FSF.
I had to spend several days manually ripping out huge chunks of Win10 to get something resembling the usability of XP. That included removing all of the updating utilities, hosts-file blocking of microsoft servers, removal of most of the preinstalled apps, as well as disabling and deleting all of the scheduled tasks which try to re-install and re-enable those settings.
All this trouble so I can play modern PC games... Next system upgrade cycle I think I'll be moving to linux.
Oh man I get just a glimpse of what the Windows team has to deal with when applying patches it's absolutely insane. Never in my life have I had to worry that yum update was going to trash my system.
Steam support for linux is really going places. I mostly play indie games and older games (so nothing too resource intensive) and Steam's custom wine distribution has been handling them really well. I think I might wipe my windows partition in the next month or so.
I keep hoping I'll run into someone that's played around with that. If you're at the point where you can use their software and expect a random game to work more often than not, that's a massive leap forward for Linux on the desktop!
I've been playing doom 2016 using steam's wine offering, proton I think it's called. There currently aren't a lot of games the steam client beta supports, but it at least was painless to install and run. Hopefully they'll continue to support wine going forward, and start working with publishers to get more big budget games working.
Blizzard has officially stated they won't boot you (permanently, apparently there are occasional false positives) for using other wine offerings. I've been tempted to try to install Overwatch using Lutris as well (productivity be damned!).
I'm lucky in that I tend to prefer indie games, which are more likely to support steamos, but I'm certainly feeling optimistic about gaming on linux. I've found that I need to boot into my win7 partition less and less in the last year for games reasons.
My employer uses Windows Enterprise and it also had Candy Crush and other crapware preinstalled. IIRC one of my coworkers wrote a script to remove the bundled applications.
Vendors aren't considered threats under HIPPA which, tbf, is pretty rational because if you don't trust the entity that wrote the software or made your hardware you've got much bigger problems.
Employer was supposed to be using Windows deployment tools to create custom enterprise images... But I'm stuck in the same boat. The maintenance overhead is too much for a small business. I am very upset that the default enterprise image is full of crap-ware, and the only alternative was LTSB with no support for Edge or Windows Store.
Other than using classic shell, I pretty much like the W10 toolbar. The harder learning curve is I still have many years of searching for X, when the UX is now in Y for a given setting. It's arguably in a better place now, just hard to re-learn.
As to linux gaming, a lot of effort has been made with Steam, another walled garden, but a nicer one. I just wish some of the UI/UX in linux would solidify a bit. I've been running elementary with some tweaks on my spare, but still not enough for my main desktop.
Elementary is a great project. I am about to build a box for a single mother with two kids and I am thinking about using elementary for it because of its ease of use and similarity to mac.
That being said, you really want money to go to developers making the software. Not to discredit the work of packagers and integrators (mostly at Debian) but actually writing the code at the end of the day is the deal breaker.
For that you can donate to Gnome, KDE, specific projects like Krita, or Software in the Public Interest that funds a bunch of the non-umbrella projects like LibreOffice and Postgres.
I don't know about the parent; but, as an MBP owner for almost 15 years, for me "fiasco" means "best-in-class keyboard and trackpad have become second- to third-rate." (Most complaints are focussed on the keyboard, and the Touch Bar's insistence on interfering with touch typing muscle memory; but I've found the trackpad has its own weird set of problems, at least on my machine.)
I think my old 2010 mbp (stolen in 2013) was probably my single favorite touchpad and keyboard for any laptop before or since. My current (late 2014) is okay, just not quite as good, and I don't like the newer ones at all... I use a physical keyboard at my desk and reducing the travel more and more just irritates me and slows me down.
May take a look at the Huawei Matebook X Pro, or it's successor next year. (Worries about chinese spying aside).
It's odd to me that so many people seem to dislike this comment. My experience with MacOS has gotten more and more negative with each update. I upgraded to Mojave a few weeks ago (because the OS would not stop nagging me to) and I've had nothing but problems since.
I've had nothing but problems using Windows, Linux and iOS. But that's not very solid in a thread with people talking about ads and privacy violations.
I've not seen a single nag to upgrade to Mojave. I'm still on Sierra (and will likely move to High Sierra this weekend - I like to stay a major version behind for Mac software for stability's sake on my Mac Mini).
I resurrected my late 2012 Mac Mini recently and "upgraded" it to Mojave.
I can't even SSH into the box. The SSH daemon accepts connections, firewall is off, and login sharing is enabled. But the daemon refuses my key, my password. I can't even figure out how to get it to display the reasons for the denied access.
At least on Linux I can turn off the daemon, turn on logging, turn the daemon back on, and get somewhere with denied access.
use ssh -vvv when you connect. Several types of ciphers were deprecated for security reasons over the past few releases, you can reenable them in /etc/ssh_config or /etc/sshd_config
SSH is no different on macOS to Linux. Go and look in the usual places: /etc/ssh/sshd_config and /var/log. If you really are still stuck, disable Remote Login from the Sharing panel and start sshd yourself in a Terminal in verbose mode.
You might find that running sshd in a Terminal is probably the best way to get a sense of what's really going on if the logging level isn't high enough by default - things like cipher disagreements would probably be visible on the client, whereas authentication errors are more likely to be logged on the target side. Do come back with how you get on!
I don't think that's true. You have to screw around with launchctl. The original poster probably has to add the ability for a specific user to ssh to the box in "remote access"
The "Remote Login" checkbox in System Preferences actually does something equivalent to the `systemsetup` command. If you uncheck it, the daemon stops running, and you can then make changes or run it yourself.
I could be mistaken but I think you're crashing into Apple tightening up the security (in the good "require better encryption" way, not the bad "walled garden" way).
Aha! I had googled around a bunch but nothing helpful came about. You and @neilalexander are saying basically the same thing (I think). Thanks for the link! I'll check it out this evening :)
very picky if you’re on a single user system. If you’re on a shared system, where admin has set default home permissions open, it makes complete sense to require .ssh to be 700.
I think consumers don't give Apple enough credit for securing their OS because the other consumer-grade OS we compare it to was such a malware-magnet for decades that we've set our expectations low.
Several of the latest OSX releases have progressively locked down some executables/directories to Ring -1 (requires more than just root/sudo privs). I see it as Apple locking down the OS against potential malware, as I've always been able to work around their changes with some Stack-Overflow-fu.
With macos, you need to buy the hardware, and the hardware locks you out of using Linux lately. You also have to spend like crazy on pre-upgrades since everything is soldered in. Then they fight tooth and nail to keep 3rd party repair from being viable, so the apple store says you need a fully new machine and yet the guy down the street can fix it for $700 (would be less but everything is soldered). The machines and os are great, but the other stuff is a buzz kill.
With red hat, I'm not sure its up to date enough for desktop use - dedoimedo likes to talk up his centos install but he doesn't actually recommend it. I'm going to look at red hat one day for production.
For desktop I would say SUSE is the way to go. You can get different levels of support. Much more worth it than Windows license - which is the price you pay to get Microsoft stickers on your machine (and to have their spyware on). Calling MS support was the worst experience ever for me.
You shouldn't really use RHEL as a desktop OS (they even dropped KDE recently). On servers if you have the money RHEL is the best you can get.
I've never looked into SUSE. Could you elaborate on what they offer? All I know about them is that they are one of the really old enterprise distros and got purchased recently.
To be honest, if a tool could run a scan over all my github contributions, and then use that to find a really well matching job, that might be potentially very interesting. Finding a job which is a good match for your skills and interests is hard. If most of your contributions are public then making use of them would be potentially beneficial so long as it wasn't super creepy.
The main challenge would be the information asymmetry between both sides (as is already the case). It's one thing for the recruiter to be able to scan all your contributions. Would you as a potential recruit be able to do the same, and see what skills and techniques were /actually/ in use by the company. A tool which could generate an anonymised summary for both parties and compare them for mutual compatibility might be very nice.
The problem with that is the same as every other time anybody is trying to exchange something for something else: sturgeons law. The greatest developers are not on the market and the companies worth working for don't need recruiters.
Ads can be great in a (very) few areas: nobody is going to talk about a new great tampon or condom. Heck most people haven't even heard about the menstural cup, but other than things humans feel really embarrassed talking about, exceptional products don't need ads, and neither do exceptional programmers or companies.
> The greatest developers are not on the market and the companies worth working for don't need recruiters.
Not sure about that part...at this point most companies that are growing at a significant rate use recruiters. They may be internal recruiters (but even then they often use a minimal amount of external recruiters too). All of the companies we all know (Facebook, Google, Microsoft, whatever) use recruiters, most VC startups do, and a portion of small to medium size companies also do.
The other thing here is, of course, that these "companies we all know" are not the be all and end all for software development. There are hundreds of thousands of smaller companies the world over doing development in every field you could imagine. I know nothing about them, but some of them are very likely to be doing things I would find fascinating. There's very little to connect me with them except by chance. Even if I'm aware of the main companies in the specific field I work in, there are still far more that I'm unaware of. That's a lot of missed opportunities.
The current process is inefficient and suboptimal for all concerned. Recruiters have trouble identifying good candidates. And on the other side we have trouble identifying good companies/projects. On both sides we dumb down the skills provided/needed to bullet points of technologies and tools, often completely ignoring the actual problem domains. While I don't think the suggestion here about gleaning insights from git repositories is a solution by any means, it might make a small improvement by providing some deeper insight into real demonstrated skills and experience. As someone else mentioned, of course this could lead to gaming the system, but I suspect most fraudulent stuff could be filtered out--there would be clear differences between real projects contributed to over the course of years and "resumé fodder".
I think its more that if the owners of software respected you at all they would not be trying to sell it you in a proprietary form to begin with. It should be no surprise that those trying to exploit and abuse copyright in an era when it makes no sense don't respect you just because you participated in the scam.
Also, they (Apple, MS) don't want your money. They want your freedom, which is virtually priceless. I'm sure a fat monthly fee will suffice, but yeesh.
The last update changed the association of web email links back to microsoft mail, still cant figure out a way to change it back.. Why cant microsoft concentrate on security updates and leave the rest of my system the fuck alone.
I know you're not necessarily looking for specific advice to this one problem, and I agree and would rather pay for software than be the product, but since email is the topic, have you looked into FastMail? You pay for it, there's no ads, and they don't sell your info.
I used to think that being open-source was enough to avoid being spied on but it is not. Ubuntu showed this with the Amazon search issue [1] and is apparently collecting telemetry with 18.04 LTS [2]. My understanding is that Brew also gatherings telemetry by default.
Enabling telemetry collection as the default setting needs to end -
even if it is well intentioned.
Canonical doesn't seem as concerned with privacy as it could be, but at least the collection is opt-in at install time instead of being sneaky (let me know if I'm mistaken).
I wish the fsf distros could get more contributors. I understand why corporate contributions aren't aimed in that direction, but I wonder why Trisquel doesn't have more hobbyists pitching in.
18.04 prompts you very clearly and the first report they let you see looks innocent. Is there something concretely wrong with 18.94 telemetry beyond objecting to even optional telemetry on principle?
I've avoided Ubuntu since the Amazon-search issue, so I haven't tested 18.04 personally, but the second article I linked to says:
"This data collection is turned on by default"
I took this to mean that telemetry was being collected and, if someone desired, could be turned off after the fact. I have a problem with telemetry being opt-out (it should always be opt-in). If Ubuntu is prompting to allow or disallow the collection of data, then that is good news, as people are being asked for their informed consent.
No. Because if you have money to spend, then your ad spots are worth more than the average. And if you are so rich you can afford to pay not to see ads, those spots goes up even more. If you are so rich you can afford to pay this increased rate, those goes up even more.
There used to be a time where ads on the internet were, if not outright illegal, then at least considered like spam is today. I am not sure we wouldn't be better of if we didn't go back to to those classifications.
I suspect the appeal of an ad model is that it's perpetual recurring revenue.
It's the same direction they go when they desperately push subscription licenses (see: Office 365) They want a business model less dependent on people replacing their PCs every three years.
I don't doubt they see a rapidly approaching future of management saying "these computers are more than adequate for office work and web browsing, so we won't replace them until they physically fail in 5+ years." It's an easier and easier argument to make, especially when we're running out of easy, cheap across-the-board performance boosts. (Office and Chrome don't run that much better on a new 32-core Threadripper than on a second-generation i5)
Amazon has lately been putting ads on prime video for other prime video shows. I have a feeling we're rapidly going towards most paid services having ads.
Also, on prime video, occasionally, pressing the fast forward button won't skip the ad. I'm sure it's one of those "bugs" that just happens to force the user to do something the company wants but the user doesn't but I have a feeling it might just stop working all the time and we'll be forced to sit through the whole ad on a paid service.
This is infuriating. I'm paying for the service--that's supposedly what's sustaining it--why are there ads? If you (Amazon) decide you want more money, let me decide to pay more to eliminate ads.
Similar things abound with Hulu. Even the more expensive "ad free" version has some ads for certain shows/movies due to "licensing agreements" (read: Hulu wanting more money).
Once these companies decide their best path to new profits doesn't include new or improved products or growth it's just a question of who they are going to fuck, their employees or users.
Wow, glad to hear about that Hulu situation. I used Hulu for a while, then dropped it when they added ads to my paid account. When I heard they had an ad-free version again I thought I might come back - but if the ad-free version is not ad-free then I'll just keep my money.
Depends on the program source... some have an ad in front of the episode even for the ad-free tier. The "live tv" option sometimes keeps you from skipping some ads as well. Most frustrating to me is they have an updated app on the fire tv devices, but won't update the android tv app (NVidia Shield TV).
Its not new whatsoever. I'm sure double dipping predates newspapers who combine recurring billing and ads in between the articles. Cable TV today is a really obvious offensive case of it because unlike with the papers you aren't as constrained by recurring distribution costs.
Newspaper subscriptions mostly pay for the delivery costs... the printing costs are largely from the ad revenue.
Also, Cable is a bit of a mixed bag as the cable company you pay doesn't pay for ALL the content on 100+ channels. Even if you only want like 10 of them. That's where ads pick up the slack.
That said, the ever shrinking runtime is really annoying... late 70's hour long programs had around 48-50 minutes of content... today, it's inching under 42 minutes of content (and overlays on credits most of the time).
What are you talking about? cable TV has always had ads in the very beginning in the 1950s and 60s on through the 1970s when the first cable TV stations came out.
Nope. local channels continued to have ads but cable did not.
Cable TV was advertised as being ad free because you were paying for it. It was part of the sales pitch.
My household was a subscriber in those early days and I remember the concern that ads were coming. I remember that they started by putting the ads between shows and only advertising upcoming shows on that channel. Then it moved to general ads about shows on that channel, then to ads about shows on other channels. At that point I knew the cause was lost and sure enough soon it was normal TV ads interrupting shows.
Here is a newspaper article from around that time asking how much longer cable TV will be ad free
Netflix is going down the same route. They are not interrupting continuous playback yet but they are already promoting their Originals with autoplay ads (sometimes even with sound) in some clients. Luckily their clients' range of functions is horribly inconsistent, so I can still chose different clients to avoid this.
I feel like we're diluting the atrocity of shitty ads when we refer to anything and everything as an ad.
Netflix's above-the-fold, skippable "Now Showing" preview on the homepage can't be compared to the unskippable videos that Amazon and Hulu force you to watch. Hell, it even remembers that you muted it. And it only plays if you sit there staring at it.
Since almost all of Netflix's competitors force you to watch ads after you click play, I'm pretty happy with what Netflix does. Seems like a perfectly fair trade-off to me.
If you don't like it, just add ".billboard-row" to your content blocker. I bet you haven't. I bet you kinda like it. ;)
If it's not aiding me in navigating directly without delay or distraction to what I want then it's an ad. Whether it looks interesting or not is irrelevant.
I've switched almost entirely to watching content I own, whether its iTunes purchases or ripped DVDs. I control the entire experience, I don't have to worry about content suddenly vanishing, I don't have to deal with anyone trying to steal my attention.
I feel the same way about dropping Netflix now as I did about dropping cable TV 5 years ago. I'm never going back.
I have a feeling we're rapidly going towards most paid services having ads.
This is why I won't subscribe to Hulu. If I'm paying for it, I'm not going to watch ads unless I absolutely have to. And "absolutely have to" is why I pay for satellite TV. Can't wait until I move to a place with OTA TV again. I don't mind watching ads on broadcast TV because it's free.
It's free 2-day delivery for me, and I don't always get that. Funny how when you order it will say "guaranteed delivery by <date>" but if they don't make that all you get is a "sorry" email. Not much of a guarantee.
For me it was free two-day shipping on Prime-eligible items until Amazon at some point decided to make only some Prime-eligible items eligible for Prime shipping. In effect they’ve silently and somewhat arbitrarily changed the definition of “Prime”.
when someone says "let's monetize X", what they're really asking is, "how can we burn all that developer goodwill we've worked really hard to earn". So maybe go looking for another revenue channel?
They also do this on the web version of Outlook. [1].
And you could see that the ads are littered all over the place. And their ads don't really make sense anyway, as for example, the ads that I see are something that:
- I am not interested
- Irrelevant as the products have no use for me
- Unable to purchase as the products themselves are not available locally where I am located.
They should probably train their algorithms for ads to ensure that it is something interesting (or at least would compel the consumer to look). For example, I noticed that in Instagram, the ads are more targeted, and they usually remind you of stuff that you have recently searched.
Outlook.com is also a mail provider, so it's not really a proper comparison. Outlook Web App (OWA), the Web client for Exchange and Office 365, and Outlook on the desktop have no ads.
As a vocal critic of the Linux Desktop, your horrifically misguided attempts to "improve" Windows since Windows 10 have so degraded the experience of using it that I am now only 2 reasons[0] away from switching anyway, so at least I don't have to pay for the privilege of using crappy software.
I would ask you kindly to cease your endeavors to kill off the Windows Desktop and personal computing in general, but having dealt with you I know that it is your policy to ignore user feedback.
[0] 1. I hate package managers as an application installation paradigm, and 2. GPU drivers still perform worse for my card, if you must know.
Legitimately asking; does AMD still do that thing where they unceremoniously drop support for an older card on newer kernels - so you do a apt-get upgrade at some point and reboot to a black screen, and your options are either never upgrade your kernel or get a new GPU?
This, unfortunately is not as true as I wish it was. I bought an RX 560 and tried to install Ubuntu 18.04. It could not propery drive my 4k display that uses DP 1.2 MST. That display has been working fine with a few different cards on Mac and Windows since 2014 when I bought it.
I even tried the AMD pro drivers. Granted, the open source nvidia drivers couldn’t do it with a 1030 on 18.04 either, but the proprietary ones had no issue.
edit: I should note, the RX 560 runs fanless up to 60c on Windows, but the fans always spin on Linux.
I have no problems running my RX 580 with the amdgpu driver and recent versions of Mesa.. If I had to guess, Ubuntu 18.04 has some archaic versions of both the kernel and userspace drivers.
Do you have a 4K display that requires DP 1.2 MST?
I tried the latest proprietary amdgpu-pro drivers also.
I even bought the RX 560 because some other Linux users assured me the AMD Linux drivers were excellent. My post here is to help other people so they can make a more informed choice in the future.
The Mesa drivers Ubuntu (and Debian, and many other distros) provide CAN be rather old. As you mentioned Ubuntu, follow the directions here to have a stack that's as updated as possible.
> 1. I hate package managers as an application installation paradigm
wat
so you prefer go to vlc.com for vlc, 7-zip.com for 7-zip, update them by hand or add custom system services constantly running for every program to install themselves, etc etc instead of a `pacman -S 7z vlc krita libreoffice [every other software you need]`
In open source world, this is fine. In the closed source world, this is one step closer towards walled garden. Just the other day, I could not install something because my Windows 10 was set to "install only from store" and I had go in and turn that off. How long before that's no longer even an option?
Yes. I like having software that is up to date and delivered direct from the developer. I like having portable applications that don't even need installation. I like being able to put applications on different disks. I like being able to have multiple versions of an application.
There is not a single package manager in existence for Linux that gives me these things.
Yeah, they do actually. I really like AppImage, it's a shame they aren't really embraced at all by the community. Good luck getting them to show up with icons in your file manager (even Nitrux, which supposedly embraces AppImages, can't seem to do it).
AppFS is a package manager for Linux that exists and solves all these problems, except the "putting them on different disks", which it doesn't prevent you from doing but it doesn't really make any sense either.
If its one thing I have learned about software development, its that you must account for people doing completely non sensical things.
I knew a guy who had MS word 97, 2000, 03, and 07 all installed on the same machine, and he used all of them for various tasks. Insane? Most certainly, but welcome to humanity.
AppFS has no problem with that. The problem comes from keeping different packages on different disks -- since it caches them all into the same directory by default. It would be trivial to make it cache files from certain packages into different places as part of the configuration -- that's easy enough. The issues come into place for files that are shared (having the same SHA1) between packages, then you either have to duplicate the file or have some policy of searching for it.
A better approach, if you care about this, is simply to have a lifecycle policy where more recently used blocks live on the SSD and less recently used blocks move to slower storage (bcache can do this for you), or are deleted (this is a planned feature: http://appfs.rkeene.org/web/tktview?name=930821d0fb ).
Aside from the obvious missing feature, there isn't even enough information to very your claims. Can I really have two versions of, say Krita, installed? Can I make a copy of that to take to an offline PC?
> I like having software that is up to date and delivered direct from the developer.
I really think your argument against package managers on Linux is incorrect. Many linux distributions provide "some recent version" of common tools, but of course you can get the latest package directly from the developers.
Nothing stops you from getting the latest packages directly from the developer (assuming the developer provides packages in the first place). Some developers even offer their own repository you could add to the package manager so you can stay up to date regardless of whatever version you don't like in the OS repository.
> I like having portable applications that don't even need installation.
You can build from source and run without installing. That's probably not quite what you mean though.
> I like being able to put applications on different disks.
No problem here. You can mount disks to wherever you want and you can install using a prefix path to those locations.
> I like being able to have multiple versions of an application.
I've spent the past five years building a lot of things from source. I agree that there are definitely a lot of things that it's nontrivial for. However, I've found that the hardest part is satisfying dependencies. Once dependencies are satisfied, building is usually really easy. And the software that isn't easy to build after satisfying dependencies is usually not worth building in the first place.
Of course, that's just my experience as a C++ developer.
> but of course you can get the latest package directly from the developers.
If you're willing to compile it from source, which I'm not since it isn't the 70s anymore.
>...
The rest of these are all the same thing: you can do these things as long as you compile from source. As I've already mentioned, I consider that a backwards and tedious way to deal with application installation. I mean ffs, getting build environments right is such a pain that many devs ship a docker container of a working environment. Not to mention it is only viable with OSS software.
I like how homebrew cask does it, where each cask is a formula to download the app into /Applications . You get the reproducibility of package lists, the speed of just typing an install / search command vs searching a website and the connivence of normal app installation and management.
Occasionally, I still run into a piece of software which makes me yearn for the days when I built my own /usr/local from source. But as the repos have got better, this is getting to be a rare event. I still don't like the feeling of being boxed into the package management system. And the extra level of complication trying to manage issues from building software from source which reference package managed libraries and other dependencies.
There's no reason that they need to come from the developers. I'm maintaining over a dozen packages that build directly from HEAD that are all integrated with my package manager and allows anyone else to leverage my work and get my packages.
Nix is the package manager that does this. It literally does damn near everything that comes out of every HN discussion of package managers, ever, but very few seem to bother investigating it.
For my part, I've discovered since switching to Mac that I don't like having system packages managed alongside user packages.
It's one thing when I'm all-in on handling everything from the kernel up (it was fine back when I used Gentoo, for example) but the middle ground of massive meta-packages and such mingled with user-installed individual packages on e.g. Ubuntu is a train wreck. MacOS w/ homebrew's great. I can delete the entire directory with my brew installed crap in it and the system's unaffected. System upgrades are separate from my stuff. That plus the occasional drag+drop installed program cover my needs nicely, and better than everything's-a-bespoke-GUI-installer on Windows.
Lots of things. They're basically the same walled garden you get from Apple, Google, or Microsoft, only with a different gardener. They're also ludicrously inflexible in that they don't allow multiple versions of the same application to be installed, install applications to different disks, or often even get timely updates when a developer makes a change.
I agree that "installers" are not a good paradigm either, but a lot of Windows software can be had in portable (or at least non-installer) form.
> They're also ludicrously inflexible in that they don't allow multiple versions of the same application to be installed
They actually do allow multiple versions of the same application to be installed. It's just difficult to get it to work.
The package itself will install to (for example) /usr/bin. A new version of the package will have the same file names, so they get overwritten.
Most well-built packages will allow you to override the installation path though. This is perfect in case you want to install to your home directory (instead of globally) or some versioned directory.
For example you can relocate RPM installations given an example.rpm file:
Correct me if I'm wrong, but won't all of the dependencies also need to be installed to the same prefix? So I could end up with half the OS on there for one application?
When you stray from the beaten PATH (and LD_LIBRARY_PATH) then yes you can start encountering trouble. Most good tools provide the options for you to do that. If you want to start building from scratch then `configure` and `make` often have PREFIX (and flavors such as BINPREFIX, LIBPREFIX, etc). `cmake` has similar variables to control where it looks for dependencies. Other languages' tools undoubtedly do too. I highly recommend you to read the documentation of the tools you use.
In other words, I can get what I want, which is very simple on Windows and Mac, in Linux by compiling things from source or otherwise jumping through a bunch of containerization hoops.
Theoretically, yes, realistically, not really. For all its talk of openness and freedom for the user, the Linux community has a remarkable amount of disdain for anyone who isn't a C greybeard.
AppImage has already pretty much solved everything I want as far as application distribution. It is at least the third such implementation that has.
However, the community refuses to embrace it and so very little software is distributed that way, just like the last two times.
> the Linux community has a remarkable amount of disdain for anyone who isn't a C greybeard
I feel like you're the one providing disdain simply by saying "greybeard". Please don't disrespect people who have lots of C experience.
I've found plenty of non-C applications in Ubuntu's and Fedora's repositories though. I've found Python, Ruby, and Perl applications and developer tools (in addition to the obvious C and C++ ones). That's just off the top of my head. Erlang, Ocaml, R, and (maybe?) D also exist. Of course there's also the package-building tools.
Mailing lists and bug reports are usually open to the public too. And there's usually policies around offensive behavior. If you see a problem that hasn't been addressed satisfactorily then I highly encourage you to bring it forward to the maintainers.
> AppImage has already pretty much solved everything I want as far as application distribution. It is at least the third such implementation that has.
It looks like it supports every major Linux distribution I've come across. Then what's stopping you from using it?
> the community refuses to embrace it and so very little software is distributed that way, just like the last two times.
This sounds like you need to get application developers on board. Get software developers to provide an AppImage of their application.
Make sure to include source code for distributions to build their own version of the AppImage -- including any patches which might be necessary to tweak and support the distributions' different and unique environments.
> I've found plenty of non-C applications in Ubuntu's and Fedora's repositories though. [etc.]
You've missed the point. I could go on for quite a while about what the point was, but ultimately it comes down to this: "You can fix it yourself!" isn't really true. Professional programmers who work on Linux for a living haven't even managed to fix a lot of its problems.
> It looks like it supports every major Linux distribution I've come across. Then what's stopping you from using it?
Nothing, except that hardly any applications are available as AppImages and they're a 3rd class citizen at best on all major distributions.
> Make sure to include source code for distributions to build their own version of the AppImage -- including any patches which might be necessary to tweak and support the distributions' different and unique environments.
The whole point of AppImage is that that's completely unnecessary. The only thing that really aught to be done is that file managers should display AppImage icons and not harass the user about double clicking one to run it. My point isn't that such things are holding me back, it's that such things are indicative of the community's attitude towards AppImage.
> ultimately it comes down to this: "You can fix it yourself!" isn't really true. Professional programmers who work on Linux for a living haven't even managed to fix a lot of its problems.
I could say the same thing about every OS.
> The whole point of AppImage is that that's completely unnecessary.
Unnecessary to you perhaps. Being able to see the source code of applications I run is important to many Linux users.
> The only thing that really aught to be done is that file managers should display AppImage icons and not harass the user about double clicking one to run it.
I don't see why AppImage should be treated any differently than any other executable -- trusted or not. If that's not how it is now then you definitely have a point.
But I can't do these things on Windows 10 either, I tried to put all my installed software on D: and when I had an update I was unable to install it because Windows couldn't work out where my stuff was which required a reinstall to install the update. Now I have to manually enter 'D' prefix on all installations, and that's for software that actually lets me!
The Linux package managers aren't perfect, but overall offer more flexibility for what I personally need.
Unfortunately Microsoft has embraced hardcoding paths for their own software. However, I've found that the vast majority of Windows applications already are or can easily be made portable. Updating them is a bit more manual, but that's a tradeoff I'll make every time. After all, you never know when an update to your mail client will suddenly introduce ads.
Too off topic for this thread, also past experience has shown that all that will happen is a bunch of evangelists will come out of the woodwork to tell me variations of: you're using the wrong distro, you're using the wrong hardware, or your workflow is wrong and you should change how you do things.
Not the OP but my most recent attempt to ditch Windows on my gaming desktop ~6 months ago was tripped up when I discovered that core software like file managers and such still crash or wig out constantly (it's always been that way, I simply used to put up with stuff like that, just hoped that had changed since I'd been away) and generally basic interactions (e.g. drag & drop) are still inconsistent and buggy. This on Ubuntu which is allegedly relatively polished. Tried KDE and Gnome both (Gnome's performance was also terrible, incidentally).
And all of it looked pretty bad. I'm a fan of (though not insistent upon using) "ugly" old school interfaces, so it's not that it wasn't "flat" or "beautiful" enough or whatever the current trend is, just lacking consistency and the result of a whole bunch of small, poorly-chosen defaults adding up to a big stew of gross. I could overlook that if it weren't also crash-prone and janky as hell, but it is.
So I'm still stuck on Mac for anything serious, and my gaming desktop sits in storage because I can't be bothered these days to screw with either Windows post-10 or Linux. Sigh.
FWIW my favorite desktop I've ever used, as far as core software, stability, desktop UI/X, compared to its contemporaries, was BeOS.
I switched to Arch Linux when RedHat was swallowed by IBM. KDE Plasma 5 is my thing. Windows 10 is a VM guest in Arch host. 32GB workstation, multiple monitors - 2 24 inch Dell monitors, 1 Haeir 55 inch, SSDs and Western Digital Caviar Blacks and Transcend External HDDs... no issues to report.
Same experience with Fedora for more than 2 years.
FreeBSD with Xfce 4 and TWM on my laptop. No crashes, no issues.
Huh. I hit a ton just trying to manage some files and burn a cd or two. Workflow inconveniences, crashes in various applications, little "some KDE service you've never heard of has crashed" messages from the system tray. Firefox scrolling was jerky until I googled a bit and found the right setting to change which is apparently the stop-being-broken switch, which is a thing that exists for some reason. Stuff like that, and on what's supposedly one of the most just-works distros. Weird. Fits with my experience running Linux as my main desktop (and laptop) OS from ~2002-2008, though once I switched to Gentoo everything was less buggy at the cost of having to configure everything manually.
Consider FreeBSD and its ports model. You'll be building your own ports from sources but you needn't touch another package someone else built, if you don't want to.
No. I have no desire to have my primary means of installing software be to compile from source. That's retarded. Literally retarded, in that it is a backward concept from the before-times.
At some point aren't you going to have to choose between software requiring shared libraries and take the package-complexity or source build-hit, or have everything statically linked and take the program size hit?
Only if I insist on doing things the way UNIXs do, which is what is keeping me on Windows.
If Linux had a stable and consistent base system that was comprised of the most commonly used libraries, then developers would only have to statically link or include their own copy of less common libraries. I'm totally ok with that.
And let's be real for a minute, the program size hit is negligible. You can get a 1TB SSD for under $200. The largest part of almost any program that's in any way large are its assets, not its code.
Windows allows multiple instances of the same app to work via the WinSxS mechanism which keeps copies of basically every DLL ever installed, including system DLLs. So it is basically keeping duplicates of everything, regardless. This includes some executables as well. [1]
Before WinSxS you could achieve the same effect by keeping copies of the DLL dependencies in the app directory beside the executable. Windows would then choose the DLLs from the app current directory before looking at system directories and the path. [2]
Linux gives you the same choices and a couple more options.
Work with the system package manager [1].
Link dynamically and keep copies of differing DLLs in a different prefix and... take a size hit. [2]
Link statically and take a size hit.
Link dynamically and build everything yourself, minimizing your dependencies.
You can't do the last two with Windows, for the most part because you do not have source code.
> Linux gives you the same choices and a couple more options.
Only as the builder of an application. The application developers for Linux, on the whole, only ever chose option one, leaving everyone who uses their software no option that doesn't involve recompiling from source (which, as I've mentioned, is a ridiculous thing to expect people to do just to use your software).
I have several dozen applications on Windows that are not installed. Many of them aren't even designed to be portable, yet were easily made such just by extracting their files from the installer into a directory. I can put these applications on a thumb drive and take them to an entirely different Windows system and run them without installing them their either. These directories are exactly as they'd exist in C:\Program Files.
This is of course because they have copies of their dependent DLLs in their install directory, but only the ones that aren't part of the Windows distribution. Let's look at one: WinDirStat. It is 2.1MB total, 856KB for the executables, 844KB for dlls, and the rest is assets, docs, and configuration. This is a GUI application, but you'll notice it doesn't have to include a toolkit because it uses the one that comes with every Windows install. The DLLs, as far as I can tell, are unique to WinDirStat, so it wouldn't be unfair to say that this application actually doesn't have any dependencies other than the base OS.
That's what Linux can't get you. There is no base OS above syscalls. You can't depend on anything being there because there are hundreds of distributions and what's available to link with changes with configuration, distribution, and often version.
If I could, I'd have all my Linux applications as AppImages which follows pretty much the same model as portable applications do on Windows. As I said though, Linux developers rarely offer that as an option.
>> Only as the builder of an application. The application developers for Linux, on the whole, only ever chose option one,
Again, wrong.
The people that package distributions usually make these choices. Gentoo and distros like it give you a choice of how things are built.
Off the top of my head I cannot think of a popular library that does not have the option to be statically or dynamically linked.
>> I have several dozen applications on Windows that are not installed. Many of them aren't even designed to be portable, ....
>> ... The DLLs, as far as I can tell, are unique to WinDirStat, so it wouldn't be unfair to say that this application actually doesn't have any dependencies other than the base OS.
This works in Linux too, somewhat. You have to have compatible ABIs though.
>> ... dependent DLLs in their install directory........
And Windows has 8 tons of different versions of system DLLs and C library runtimes in the WinSxS directories. MSVC runtimes are not interchangeable. Linux in general does not go to that length since a large majority of the apps are source available.
> I hate package managers as an application installation paradigm
Flatpak has you covered, or at least, it's getting there pretty fast. You'll still need to manage debs for base system stuff but you can install most gui apps as flatpaks now.
If only it did. For one, its absolutely a package manager. You need a special tool to manage flatpaks.
Also, flatpak has a lot of problems in my opinion, not the least of which is that it has no mechanism for installing to a different disk without creating an entirely new "installation", which in its terminology means a completely new dependency chain too.
The "portals" idea is a pretty good one though, on the whole.
It's still a package manager, and it still has dependencies, but the dependency chain is "flat" such that it's always comprehensible and you'll never be stuck modifying large parts of your system to update one program. I see it as similar to Windows's Visual C++ and DirectX runtimes in that regard.
I hear you on installing to different disks, though. Appimages are also an option, but I've found they have a tendency to not actually work between different distros and system versions.
Unfortunately that's a problem with they way Linux is structured, or rather it's lack of structure. There's no consistent base system that can be relied upon like you have with other platforms. The community seems unwilling to deal with this issue, some outright deny that it is an issue.
One thing that would be interesting would be to use flatpak "runtimes" to provide a target base system for AppImages, but I don't think that level of cooperation is actually possible within the Linux community.
I don't really think it's an issue though, if you really want portability you shouldn't rely on any software provided by the distribution unless you want to depend on that specific distro.
Flatpak runtimes are basically interchangeable base systems, it seems silly to commit to any one in particular when you can use any number in parallel.
I suppose this is a consequence of making all future versions of Windows essentially free by adopting "Windows as a Service" release process. When everyone upgrades to Windows 10 there won't be anyone left to buy it anymore so they need a new source of income to replace that.
> I suppose this is a consequence of making all future versions of Windows essentially free by adopting "Windows as a Service" release process. When everyone upgrades to Windows 10 there won't be anyone left to buy it anymore so they need a new source of income to replace that.
OEM Licenses are bound to a single computer AFAIK. You change computers you need to pay for Windows again. So it's not like people will never buy Windows anymore as long as they buy computers installed with Windows on it.
Before one would get a separate Windows CD with his computer and be able to install it on many different machines, one activation at a time.
But it also used to be that new Windows versions came out around every three years so by the time you needed a new computer, you'd be paying the Windows tax again anyway. In practice, not a lot of difference.
Just another step towards the inevitable announcement of a "free," "promotion-supported" version of Windows. For decades, the beige-box industry was supported by doing things like including trialware antivirus, browser search bars, and shareware games on top of the OS. Why shouldn't Microsoft put the crapware on the machine directly, and eliminate the middle man?
I would hope that this would allow them to split the codebase, and make other, paid versions (professional, enterprise) that didn't have ANY of the tracking or advertising -- and didn't require extensive GPO manipulation to turn off -- but that's probably too much to ask.
What's funny is that there was definitely a point where Microsoft themselves realized this business model by OEMs was undermining their platform.
They actually promoted "signature" PCs which were supposed to come without crapware. I can recall buying a Windows 8 tablet from a brick-and-mortar Microsoft store because I knew it would come with a hassle-free clean install.
I only read personal mail in the web browser, its too much to maintain a mail app if you have to do more than enter much more than enter your username and password.
412 comments
[ 2.9 ms ] story [ 377 ms ] threadhttps://community.spiceworks.com/scripts/show/4378-windows-1...
Go to Settings -> Privacy, and slide many (most?) things to off. Including Diagnostic & Usage Data to Basic.
Settings > Personalization > Start and slide "Occasionally show suggestions in start" to off.
Settings -> Personalization -> Lock Screen and slide "Get fun facts, tips, and more from Windows and Cortana on your lock screen" to off.
Settings -> System -> Notifications & actions -> "Get tip, tricks, and suggestions as you use Windows" and "Show me the Windows welcome experience after updates and occasionally when I sign in to highlight what's new and suggested" to off.
Go to Cortana (Taskbar) -> Settings (Gear Icon) -> "Taskbar tidbits" to off
I wouldn't use the built in "Mail" app with or without ads, it is terrible. Just have him simply not use it.
Edit: Sorry. It's now a paid software. Checking for free alternatives. Found one: shutup10
Lastly, install Unchecky. It'll automatically uncheck options for additional unwanted programs, adware, toolbars... when installing software.
If you stay within the Settings UI then you are working within Windows 10's supported design. When you run a script that disables a system service (e.g. Cortana) things start to get quirky/broken.
Several of those break Windows Update for example.
Some people might consider that a feature /s.
8.1 brought back things like the closing button on metro apps, and some other fudging, to make it a bit more Win 7, and it just felt like even more of an abomination. The control panel/metro settings mixed UI - is the thing that killed Windows for me. In Windows 10, I can't find anything setting wise particularly easily. And on a recent new laptop start up, I cringed at all the options that I had to turn off, and at the apps I tried to remove.
To the best of my knowledge, the most recent complete leak was NT 4.0. That one did have a short-lived burst of retro affectionados under the name "OpenNT 4.0", though it's not entirely clear if it died on its own or if a cease & desist helped that along.
https://www.betaarchive.com/forum/viewtopic.php?f=61&t=33250
Open the PowerShell as admin and copy:
Get-AppxPackage * windowscommunicationsapps * | Remove-AppxPackage (remove space at the two asterisks, HN turns the text into italicized)
voilà, Mail and Calendar uninstalled. You can do this with all of the preinstalled stuff (Xbox as well), see the link above.
Having to do this has in the past been the main argument I've seen used against using Linux based OSes. If you have to do it anyway you might as well actually own your computer.
How did that get integrated? Is there some sort of Microsoft VP who decided to stake their career on Cortana and overrode everyone to merge that awful design decision into the installer, or do Microsoft employees like it as well?
Who ever wanted to run a Windows installer via voice? Has anyone ever actually used the voice installer?
For larger deployments, there's a registry key and other options:
https://docs.microsoft.com/en-us/windows-hardware/customize/...
For other cases, I'd consider filing feedback requesting an easier way to mute during the setup process (if there isn't one already; I don't remember looking for that option and instead just muting the device using a hardware switch).
You know who has demonstrated that they are their own worst enemy? Consumers.
People vote with their pocketbook. Now you have to live with that result.
My parents do not need $1.5k of laptop to browse the internet, email, and exchange photos of grandkids. Particularly in a world when a decent laptop doing the above should cost $750 or less out the store.
Plus... if you're fleeing Windows because of the advertising, you're gonna go to a google product? How on earth does that help?
Also, 5 year old Air for $1k? That's almost how much they cost new. I'm seeing half that on Ebay/Craigslist. My sister has a new $450 Toshiba and my 5 year old Air runs much better. I got a new battery for it and gave it to her.
Unfortunately you went way too far with your rant and buried your point about Windows' real competition being more expensive. But the low end of Windows laptops is pretty atrocious. Someone stuck buying those aren't doing it because Airs lost their SD card port. They're doing it because they can't afford anything other than a computer that serves them ads, has 3hrs battery life, and struggles to buffer 1080p video on Youtube like my sister's new Toshiba.
I'm going to guess none. $30 of bloatware is extremely tempting when your profit margin is 4%, but it doesn't represent any meaningful savings.
In this case, I don’t think that’s fair. A great number of people can’t devote the time to learning a new operating system or cannot afford a high priced OS X machine.
Beats me. Not like we couldn’t see this coming for two decades. Soon you’ll be the product.
Maybe next year will be the year of Linux.
Can't or won't? My experience has been that several people I know would rather stay in their comfort zone rather than learn something new. I'm not just talking about OSes here. I qualify this as a won't.
I suspect this reluctance to switch from the status quo might apply to a wider population than my acquaintances.
What's the use in learning an entirely new operating system on your desktop whenever most of your time (65%) is being spent on Mobile anyway? [2]
[0] http://www.pewinternet.org/2015/10/29/the-demographics-of-de... [1] http://www.pewinternet.org/2015/10/29/technology-device-owne... [2] "Mobile now accounts for about 65% of total digital media consumption" https://hackernoon.com/how-much-time-do-people-spend-on-thei...
There really isn't any learning unless you're a power user, and those are the people that can handle learning.
And even if there was, Microsoft moves their UI around often enough that people have had to "learn a new OS" at least three times by now.
Use Google, get everything for free subsidized by ads. Pay for Windows, still get ads?
Hmmmmm
Edit: You also don't get free updates if you don't have an Apple machine. Still not free.
[1] https://www.theverge.com/2018/5/7/17326184/firefox-ads-spons...
Recently Amazon has been playing ads at the beginning of shows that are offered for free via Prime. I feel like I even saw them when I had bought a season of a show, but it is now offered for free via Prime.
So far it seems like buying shows and movies through Apple is ad-free.
In practical terms, any software engineer knows how to work in a POSIX environment, so no issue here.
>setting up
If you mean "installing Linux", then any mainstream distro (e.g. Linux Mint) is ready to go out of the box. If you mean "configuring things up to your liking" then it is no more time-consuming than any other setup.
>operating
Don't know what you mean by this, but a POSIX environment is probably the most efficient/least painful way to do software development.
Gmail, Outlook.com/O365/etc, are designed for you to "keep everything." Problem is that all desktop email clients aren't designed that way, they want you to have a very small limited subset of email that they can index and constantly have in memory.
When these two worlds converge desktop clients choke and choke hard. Outlook is so terrible that it almost needs to be completely retired. They've barely updated it in ten years.
Best case scenario right now is full webmail (with anyone) and browser notifications for new email etc if you need that.
with large amounts of mail (~90k in one of my inboxes - and if someone asks just yesterday I had to go and fetch some info from a 2007 mail), GMail chokes much more than thunderbird, e.g. it takes 2/3 minutes and sometimes does not work to mark a large set of email as read/archived/whatever while this is almost instant with TB.
the default answer to op's question in enterprise land (where windows is still treading for relevancy) is Outlook.
thunderbird suffers the issues the comment mentions above -- barfs with large dbs, no support for exchange, not sure if it even supports push notifications, etc
it should be noted that mail on mac works well and the win 10 mail app this article is all about works pretty well too, which is why its a travesty MS is going to ruin it.
I have over 20 years of email stored locally in Postbox. I've kept every non-spam email since I first got online in 1997. I regularly use it to search for things people sent me (or I sent them) months or even years ago. It's really not a problem at all.
I've got Postbox open right now and it's using 625MB RAM. (That's a lot compared to the Eudora days, but I have 16GB RAM now.)
A desktop client on your laptop also means you can write emails & be able to search your email history while in flight mode, and queue the messages to be sent when you land & have WiFi again.
I use Mailmate. It’s everything I wanted Thunderbird to be.
I'd like an alternative that has close to tray functionality.
15 years later and 3 major windows iterations later and we still don't have it.
https://postbox-inc.com/
If you do feel that's a lot, try the demo first to see if you like it, and maybe wait for Black Friday in case they have a sale.
https://www.postbox-inc.com/store
I swear half of the clients on the market turn out to support only a small list of authentication methods, or don't let you specify port!
It is of course overkill though if you have one email account and only process a couple emails a day.
i still dont think it supports the iCalendar/calDAV standard either.
It's too easy to write a backhanded comment regarding this issue, but it's just really annoying.
https://github.com/Foundry376/Mailspring
https://drive.google.com/file/d/1yRBphYeo3uUl5jFFQJufLMKPmup...
Any recommendations for a touch-friendly (i.e. must be UWP) mail app for Windows?
The article suggests that the Mail app will never have ads for Office 365 users. Office 365 Home is often a good deal for a Surface owner anyway, as it boosts your OneDrive space, and gives you up to date copies of all the Office programs, and stuff like that.
Fuck Edge touch support - they don't support the event API used by Safari/Chrome, instead you need to use an MS specific touch API.
I support touch on IE11/Edge (because our business made a bet on Windows Phone) but I have wasted an obscene amount of time to do so (we have very few windows phone/laptop touch users).
I hate supporting Edge more than IE11 - weird bugs and continual problems with the debugger and touch support.
Edge supports both (and has for some time). IE supports only Pointer Events. (There is a polyfill for Touch Events on IE: https://github.com/WebReflection/ie-touch)
https://developer.mozilla.org/en-US/docs/Web/API/Touch_event...
https://caniuse.com/#feat=pointer
https://developer.mozilla.org/en-US/docs/Web/API/Pointer_eve...
Both are W3C recommended standards. Pointer Events has more capabilities, and an easier abstraction for multi-modal input (ie, mouse and touch, or touch and pen). So I often see Pointer Events recommended over Touch Events if you can use Pointer Events. (React recently moved to recommending Pointer Events, as one example.)
The big reason not to use Pointer Events is that Safari/WebKit has lagged far behind on Pointer Event adoption (it is finally in experimental preview in Safari, years after Firefox and Chrome/Blink, and seems to be part of the increasing Blink/WebKit split). There is a Pointer Event Polyfill for Safari (and other older browsers that support Touch Events but not Pointer Events): https://github.com/jquery/PEP
Also, I love Edge to death (it's the only browser I'll use when I'm on my Surface), but I forgot another reason why I don't want to deal with Gmail in the browser: Gmail doesn't let me have my personal and work email open at the same time.
Windows: Oh you want to rename a file that is opened by an application? Fuck you. We're not even capable of telling you what application. Have some ads instead!
Linux: Ooh you thought you could use $APPLICATION_NAME? Sorry, that isn't available, but you can try $MEDIOCRE_ALTERNATIVE instead.
I'm not a fanboy, I hate every major desktop OS.
If OSX wouldn't exist, I would likely try Linux, get sick of it, try Windows, get even more frustrated, and get back to Linux. It is very likely I'd be using ElementaryOS or Ubuntu Desktop in the end.
Now, that's not to say that X11 desktop environments are all sunshine and rainbows but at least a) if you like it the way it is, you can probably keep it exactly as it is with no consequences, and b) you can always choose something simpler, where there's less to go wrong.
I went with option b here, I've been using mostly the same version of the same lightweight window manager since 2012, it is just fine, it isn't ever sluggish on any machine, and it isn't "too little" for my workstation. I use it on OpenBSD as well, and when I tried out MINIX 3 in 2013, it wasn't any effort to use exactly the same window manager. There are no commits in it (mine is a fork, but it didn't really need to be) since February 2016, because it didn't need to change.
After the latest update, my hard drive wheezes like an asthmatic climbing his 25th flight of stairs on startup.
Just a long list of problems with Windows 10
Found your problem!
/har har har
But seriously, I still refer to even SSDs as hard drives. Spinning rust or memory chips. They're interchangable right now, until spinning rust phases out completely.
Oh, how the time has changed!
> although these ads are interest-based, they do not look at your emails to display ads based on data from your email.
I'm sorry, this snark is not really directed at you, but how exactly are they determining your interests if they're not reading your e-mails? Where does that data come from? I'd be happy to take back my /s if someone can tell me it was all voluntarily collected, but I suspect the answer is really "they just found a different way to spy on you other than reading your email."
I personally use Bing, but not Windows. Looking on my interests page for Bing, I see they have a decent chunk of info on me, nothing too crazy personal though.
Pretty much every ad platform out there does exactly that
0 - https://daringfireball.net/2017/03/the_deck_adieu
Is my money not good enough for them? This is getting absurd
The value of marketing to Windows users is in the network effect. If one Windows user can pay $1000 to avoid advertising and change nothing else, the other users will probably hear about this and take notice. But if every Windows user is subjected to ads, I very much doubt that Microsoft is making $1000 in advertising revenue from just one person.
Microsoft would never convert 100% of users to "ad-free" paid subscriptions, at any price. But they can make money directly from advertising to 100% of their users, and there is no "loss" except from attrition (which I would guess is still probably quite a bit smaller than their numbers for overall user-base growth.)
So people used the same argument with games and loot boxes. But even if videogames cost $500 a copy, there is so much money to be made with loot boxes that they would still exist in these games at that price.
I don't think this is true, but if it is, please provide a link to prove the assertion.
I doubt that companies are selling software at a loss with the intent to make it up in ad revenue. They're selling the software at a profit, and then getting ad revenue for additional profit.
All this just to "maximize revenue potential" for a bunch of faceless investment funds.
Let's be honest though, if a company can make money doing something they most likely will do it. It wouldn't matter if you paid $100,000 a month to use windows, if they could make an extra $1000 selling your personal information and pushing ads at you they will.
AFAIK a license is paid with every new computer coming WITH Windows, so it isn't like people get Windows for free. I'm not sure it is even possible to transfer Windows's license from computer to computer anymore, so why is Microsoft doing this? just because they can. Imagine tomorrow some f\*cking ads pop up each time you open the start menu, or on the background of the desktop, or on the login screen, or in Visual Studio Code because why not? Absolutely outrageous. "Pay for the product and you're no longer the product they say". and while we are at it, why not display ads on websites hosted on Azure even when you paid for hosting?
The default experience was garbage thrown in your face.
The leadership had to drive revenue growth and msn/ads were a money maker. That’s how they get promoted and we had to play their game.
Seeing that ladder chasing culture made me quit.
Very few companies actually legit give a shit about their users and the user experience. It’s so easy to blindly do things that annoy users but make a quick buck.
For anyone curious about vscode, there is a fork called "VSCodium" which is vscode without any telemetry shit in it.
https://github.com/VSCodium/vscodium
https://my.fsf.org/donate
I've been a member for a while now. And we even ran a domaition drive for a local food nonprofit, and our COO would match to a nonprofit org we said. More to the FSF.
Thanks for sharing the link :)
All this trouble so I can play modern PC games... Next system upgrade cycle I think I'll be moving to linux.
2018: I'd never connect a Windows box directly to the Internet. Microsoft might break my computer.
Blizzard has officially stated they won't boot you (permanently, apparently there are occasional false positives) for using other wine offerings. I've been tempted to try to install Overwatch using Lutris as well (productivity be damned!).
I'm lucky in that I tend to prefer indie games, which are more likely to support steamos, but I'm certainly feeling optimistic about gaming on linux. I've found that I need to boot into my win7 partition less and less in the last year for games reasons.
The good news is that if Microsoft refuses to let me buy it as an individual, then I have no moral/ethical qualms with not buying it per se.
I don't think that works, I'm pretty sure the networking stack bypasses the hosts file for Microsoft sites because it's a potential vulnerability.
source: configured windows10 manually once, decided i'd like to never experience that particular circle of hell again
As to linux gaming, a lot of effort has been made with Steam, another walled garden, but a nicer one. I just wish some of the UI/UX in linux would solidify a bit. I've been running elementary with some tweaks on my spare, but still not enough for my main desktop.
[1]: https://elementary.io
That being said, you really want money to go to developers making the software. Not to discredit the work of packagers and integrators (mostly at Debian) but actually writing the code at the end of the day is the deal breaker.
For that you can donate to Gnome, KDE, specific projects like Krita, or Software in the Public Interest that funds a bunch of the non-umbrella projects like LibreOffice and Postgres.
I still think it's an excellent option for the privacy conscious.
There are plenty of free software based operating systems that will take your money. Red Hat seems like the most obvious one.
I don't know about the parent; but, as an MBP owner for almost 15 years, for me "fiasco" means "best-in-class keyboard and trackpad have become second- to third-rate." (Most complaints are focussed on the keyboard, and the Touch Bar's insistence on interfering with touch typing muscle memory; but I've found the trackpad has its own weird set of problems, at least on my machine.)
May take a look at the Huawei Matebook X Pro, or it's successor next year. (Worries about chinese spying aside).
I can't even SSH into the box. The SSH daemon accepts connections, firewall is off, and login sharing is enabled. But the daemon refuses my key, my password. I can't even figure out how to get it to display the reasons for the denied access.
At least on Linux I can turn off the daemon, turn on logging, turn the daemon back on, and get somewhere with denied access.
Fixed it with `chsh` and now all that good stuff works \o/
https://www.reddit.com/r/macsysadmin/comments/9wt05m/cannot_...
Several of the latest OSX releases have progressively locked down some executables/directories to Ring -1 (requires more than just root/sudo privs). I see it as Apple locking down the OS against potential malware, as I've always been able to work around their changes with some Stack-Overflow-fu.
With macos, you need to buy the hardware, and the hardware locks you out of using Linux lately. You also have to spend like crazy on pre-upgrades since everything is soldered in. Then they fight tooth and nail to keep 3rd party repair from being viable, so the apple store says you need a fully new machine and yet the guy down the street can fix it for $700 (would be less but everything is soldered). The machines and os are great, but the other stuff is a buzz kill.
With red hat, I'm not sure its up to date enough for desktop use - dedoimedo likes to talk up his centos install but he doesn't actually recommend it. I'm going to look at red hat one day for production.
Anonymous description of a company/profile for the recruiter Anonymous profile of candidates looking for a job...
Swipe left...or swipe right...if both swipe connect them...
F. Why did I post this. Its such a good idea I could make money out of it.
Ads can be great in a (very) few areas: nobody is going to talk about a new great tampon or condom. Heck most people haven't even heard about the menstural cup, but other than things humans feel really embarrassed talking about, exceptional products don't need ads, and neither do exceptional programmers or companies.
Not sure about that part...at this point most companies that are growing at a significant rate use recruiters. They may be internal recruiters (but even then they often use a minimal amount of external recruiters too). All of the companies we all know (Facebook, Google, Microsoft, whatever) use recruiters, most VC startups do, and a portion of small to medium size companies also do.
The current process is inefficient and suboptimal for all concerned. Recruiters have trouble identifying good candidates. And on the other side we have trouble identifying good companies/projects. On both sides we dumb down the skills provided/needed to bullet points of technologies and tools, often completely ignoring the actual problem domains. While I don't think the suggestion here about gleaning insights from git repositories is a solution by any means, it might make a small improvement by providing some deeper insight into real demonstrated skills and experience. As someone else mentioned, of course this could lead to gaming the system, but I suspect most fraudulent stuff could be filtered out--there would be clear differences between real projects contributed to over the course of years and "resumé fodder".
>Is my money not good enough for them? This is getting absurd
Yeah, this is surreal.
Also, they (Apple, MS) don't want your money. They want your freedom, which is virtually priceless. I'm sure a fat monthly fee will suffice, but yeesh.
Enabling telemetry collection as the default setting needs to end - even if it is well intentioned.
[1] https://www.networkworld.com/article/2226648/opensource-subn...
[2] https://www.techradar.com/news/the-controversial-ubuntu-1804...
I wish the fsf distros could get more contributors. I understand why corporate contributions aren't aimed in that direction, but I wonder why Trisquel doesn't have more hobbyists pitching in.
edited to clear up opening sentence
(I agree that the Amazon search thing was bad.)
"This data collection is turned on by default"
I took this to mean that telemetry was being collected and, if someone desired, could be turned off after the fact. I have a problem with telemetry being opt-out (it should always be opt-in). If Ubuntu is prompting to allow or disallow the collection of data, then that is good news, as people are being asked for their informed consent.
There used to be a time where ads on the internet were, if not outright illegal, then at least considered like spam is today. I am not sure we wouldn't be better of if we didn't go back to to those classifications.
You could maybe make this argument about popups before browser vendors killed that behaviour off, but ad banners where still a thing even with those.
The idea that they were ever even close to “outright illegal” is almost certainly a stretch in the vast majority of legal systems.
It's the same direction they go when they desperately push subscription licenses (see: Office 365) They want a business model less dependent on people replacing their PCs every three years.
I don't doubt they see a rapidly approaching future of management saying "these computers are more than adequate for office work and web browsing, so we won't replace them until they physically fail in 5+ years." It's an easier and easier argument to make, especially when we're running out of easy, cheap across-the-board performance boosts. (Office and Chrome don't run that much better on a new 32-core Threadripper than on a second-generation i5)
Also, on prime video, occasionally, pressing the fast forward button won't skip the ad. I'm sure it's one of those "bugs" that just happens to force the user to do something the company wants but the user doesn't but I have a feeling it might just stop working all the time and we'll be forced to sit through the whole ad on a paid service.
Cable TV model isn't dead after all.
Similar things abound with Hulu. Even the more expensive "ad free" version has some ads for certain shows/movies due to "licensing agreements" (read: Hulu wanting more money).
Also, Cable is a bit of a mixed bag as the cable company you pay doesn't pay for ALL the content on 100+ channels. Even if you only want like 10 of them. That's where ads pick up the slack.
That said, the ever shrinking runtime is really annoying... late 70's hour long programs had around 48-50 minutes of content... today, it's inching under 42 minutes of content (and overlays on credits most of the time).
Slowly they became as ad infested as open channels.
Soon it will be two, then three, then.... The same thing happened to "previews" when you go to the movies.
My household was a subscriber in those early days and I remember the concern that ads were coming. I remember that they started by putting the ads between shows and only advertising upcoming shows on that channel. Then it moved to general ads about shows on that channel, then to ads about shows on other channels. At that point I knew the cause was lost and sure enough soon it was normal TV ads interrupting shows.
Here is a newspaper article from around that time asking how much longer cable TV will be ad free
https://www.nytimes.com/1981/07/26/arts/will-cable-tv-be-inv...
Netflix's above-the-fold, skippable "Now Showing" preview on the homepage can't be compared to the unskippable videos that Amazon and Hulu force you to watch. Hell, it even remembers that you muted it. And it only plays if you sit there staring at it.
Since almost all of Netflix's competitors force you to watch ads after you click play, I'm pretty happy with what Netflix does. Seems like a perfectly fair trade-off to me.
If you don't like it, just add ".billboard-row" to your content blocker. I bet you haven't. I bet you kinda like it. ;)
I've switched almost entirely to watching content I own, whether its iTunes purchases or ripped DVDs. I control the entire experience, I don't have to worry about content suddenly vanishing, I don't have to deal with anyone trying to steal my attention.
I feel the same way about dropping Netflix now as I did about dropping cable TV 5 years ago. I'm never going back.
This is why I won't subscribe to Hulu. If I'm paying for it, I'm not going to watch ads unless I absolutely have to. And "absolutely have to" is why I pay for satellite TV. Can't wait until I move to a place with OTA TV again. I don't mind watching ads on broadcast TV because it's free.
If Netflix introduces any mid-show ads, I suspect there'll be a sharp rise in illegal downloads...
Vote with your wallet.
And you could see that the ads are littered all over the place. And their ads don't really make sense anyway, as for example, the ads that I see are something that:
- I am not interested - Irrelevant as the products have no use for me - Unable to purchase as the products themselves are not available locally where I am located.
They should probably train their algorithms for ads to ensure that it is something interesting (or at least would compel the consumer to look). For example, I noticed that in Instagram, the ads are more targeted, and they usually remind you of stuff that you have recently searched.
[1] https://imgur.com/a/fcmyMMf
As a vocal critic of the Linux Desktop, your horrifically misguided attempts to "improve" Windows since Windows 10 have so degraded the experience of using it that I am now only 2 reasons[0] away from switching anyway, so at least I don't have to pay for the privilege of using crappy software.
I would ask you kindly to cease your endeavors to kill off the Windows Desktop and personal computing in general, but having dealt with you I know that it is your policy to ignore user feedback.
[0] 1. I hate package managers as an application installation paradigm, and 2. GPU drivers still perform worse for my card, if you must know.
Which GPU?
In addition to the performance degradation, nVidia and the kernel maintainers don't currently get along so I'd also have to deal with that.
I even tried the AMD pro drivers. Granted, the open source nvidia drivers couldn’t do it with a 1030 on 18.04 either, but the proprietary ones had no issue.
edit: I should note, the RX 560 runs fanless up to 60c on Windows, but the fans always spin on Linux.
I tried the latest proprietary amdgpu-pro drivers also.
I even bought the RX 560 because some other Linux users assured me the AMD Linux drivers were excellent. My post here is to help other people so they can make a more informed choice in the future.
https://www.phoronix.com/scan.php?page=news_item&px=MTg2NTY
Though if you follow the DP MST keyword there's still some corner cases being polished...
https://www.phoronix.com/scan.php?page=search&q=DP%20MST
https://www.phoronix.com/scan.php?page=news_item&px=DRM-Misc...
The Mesa drivers Ubuntu (and Debian, and many other distros) provide CAN be rather old. As you mentioned Ubuntu, follow the directions here to have a stack that's as updated as possible.
https://github.com/ValveSoftware/Proton/wiki/Requirements
I also tend to like checking Arhc Linux support page, but it doesn't seem to be helpful for this issue: https://wiki.archlinux.org/index.php/Radeon
wat
so you prefer go to vlc.com for vlc, 7-zip.com for 7-zip, update them by hand or add custom system services constantly running for every program to install themselves, etc etc instead of a `pacman -S 7z vlc krita libreoffice [every other software you need]`
There is not a single package manager in existence for Linux that gives me these things.
http://appfs.rkeene.org/
I don't need every program on my SSD. Makes perfect sense.
I knew a guy who had MS word 97, 2000, 03, and 07 all installed on the same machine, and he used all of them for various tasks. Insane? Most certainly, but welcome to humanity.
A better approach, if you care about this, is simply to have a lifecycle policy where more recently used blocks live on the SSD and less recently used blocks move to slower storage (bcache can do this for you), or are deleted (this is a planned feature: http://appfs.rkeene.org/web/tktview?name=930821d0fb ).
I really think your argument against package managers on Linux is incorrect. Many linux distributions provide "some recent version" of common tools, but of course you can get the latest package directly from the developers.
Nothing stops you from getting the latest packages directly from the developer (assuming the developer provides packages in the first place). Some developers even offer their own repository you could add to the package manager so you can stay up to date regardless of whatever version you don't like in the OS repository.
> I like having portable applications that don't even need installation.
You can build from source and run without installing. That's probably not quite what you mean though.
> I like being able to put applications on different disks.
No problem here. You can mount disks to wherever you want and you can install using a prefix path to those locations.
> I like being able to have multiple versions of an application.
Prefix path to the rescue here too.
Edit: answered more of your concerns
A package manager really should be able to do prefixed installs on any package, but that's far from a universal feature right now.
I've spent the past five years building a lot of things from source. I agree that there are definitely a lot of things that it's nontrivial for. However, I've found that the hardest part is satisfying dependencies. Once dependencies are satisfied, building is usually really easy. And the software that isn't easy to build after satisfying dependencies is usually not worth building in the first place.
Of course, that's just my experience as a C++ developer.
If you're willing to compile it from source, which I'm not since it isn't the 70s anymore.
>...
The rest of these are all the same thing: you can do these things as long as you compile from source. As I've already mentioned, I consider that a backwards and tedious way to deal with application installation. I mean ffs, getting build environments right is such a pain that many devs ship a docker container of a working environment. Not to mention it is only viable with OSS software.
Nix is the package manager that does this. It literally does damn near everything that comes out of every HN discussion of package managers, ever, but very few seem to bother investigating it.
Also, if I understand how nix works correctly, it still can't create portable applications.
For my part, I've discovered since switching to Mac that I don't like having system packages managed alongside user packages.
It's one thing when I'm all-in on handling everything from the kernel up (it was fine back when I used Gentoo, for example) but the middle ground of massive meta-packages and such mingled with user-installed individual packages on e.g. Ubuntu is a train wreck. MacOS w/ homebrew's great. I can delete the entire directory with my brew installed crap in it and the system's unaffected. System upgrades are separate from my stuff. That plus the occasional drag+drop installed program cover my needs nicely, and better than everything's-a-bespoke-GUI-installer on Windows.
I agree that "installers" are not a good paradigm either, but a lot of Windows software can be had in portable (or at least non-installer) form.
They actually do allow multiple versions of the same application to be installed. It's just difficult to get it to work.
The package itself will install to (for example) /usr/bin. A new version of the package will have the same file names, so they get overwritten.
Most well-built packages will allow you to override the installation path though. This is perfect in case you want to install to your home directory (instead of globally) or some versioned directory.
For example you can relocate RPM installations given an example.rpm file:
And when all else fails, there's docker.
No thanks.
One of the awesome things about Linux and it being open source is that if you don't like it then you can work to improve it.
AppImage has already pretty much solved everything I want as far as application distribution. It is at least the third such implementation that has.
However, the community refuses to embrace it and so very little software is distributed that way, just like the last two times.
I feel like you're the one providing disdain simply by saying "greybeard". Please don't disrespect people who have lots of C experience.
I've found plenty of non-C applications in Ubuntu's and Fedora's repositories though. I've found Python, Ruby, and Perl applications and developer tools (in addition to the obvious C and C++ ones). That's just off the top of my head. Erlang, Ocaml, R, and (maybe?) D also exist. Of course there's also the package-building tools.
Mailing lists and bug reports are usually open to the public too. And there's usually policies around offensive behavior. If you see a problem that hasn't been addressed satisfactorily then I highly encourage you to bring it forward to the maintainers.
> AppImage has already pretty much solved everything I want as far as application distribution. It is at least the third such implementation that has.
Search engine brought me here: https://appimage.org/
It looks like it supports every major Linux distribution I've come across. Then what's stopping you from using it?
> the community refuses to embrace it and so very little software is distributed that way, just like the last two times.
This sounds like you need to get application developers on board. Get software developers to provide an AppImage of their application.
Make sure to include source code for distributions to build their own version of the AppImage -- including any patches which might be necessary to tweak and support the distributions' different and unique environments.
You've missed the point. I could go on for quite a while about what the point was, but ultimately it comes down to this: "You can fix it yourself!" isn't really true. Professional programmers who work on Linux for a living haven't even managed to fix a lot of its problems.
> It looks like it supports every major Linux distribution I've come across. Then what's stopping you from using it?
Nothing, except that hardly any applications are available as AppImages and they're a 3rd class citizen at best on all major distributions.
> Make sure to include source code for distributions to build their own version of the AppImage -- including any patches which might be necessary to tweak and support the distributions' different and unique environments.
The whole point of AppImage is that that's completely unnecessary. The only thing that really aught to be done is that file managers should display AppImage icons and not harass the user about double clicking one to run it. My point isn't that such things are holding me back, it's that such things are indicative of the community's attitude towards AppImage.
Maybe.
> ultimately it comes down to this: "You can fix it yourself!" isn't really true. Professional programmers who work on Linux for a living haven't even managed to fix a lot of its problems.
I could say the same thing about every OS.
> The whole point of AppImage is that that's completely unnecessary.
Unnecessary to you perhaps. Being able to see the source code of applications I run is important to many Linux users.
> The only thing that really aught to be done is that file managers should display AppImage icons and not harass the user about double clicking one to run it.
I don't see why AppImage should be treated any differently than any other executable -- trusted or not. If that's not how it is now then you definitely have a point.
Have you built any AppImage-based programs?
The Linux package managers aren't perfect, but overall offer more flexibility for what I personally need.
well, go on then, elaborate...
And all of it looked pretty bad. I'm a fan of (though not insistent upon using) "ugly" old school interfaces, so it's not that it wasn't "flat" or "beautiful" enough or whatever the current trend is, just lacking consistency and the result of a whole bunch of small, poorly-chosen defaults adding up to a big stew of gross. I could overlook that if it weren't also crash-prone and janky as hell, but it is.
So I'm still stuck on Mac for anything serious, and my gaming desktop sits in storage because I can't be bothered these days to screw with either Windows post-10 or Linux. Sigh.
FWIW my favorite desktop I've ever used, as far as core software, stability, desktop UI/X, compared to its contemporaries, was BeOS.
As for linux, I went through a bunch until I found my love for fedora. It still requires some tinkering, but for the most part everything works.
I switched to Arch Linux when RedHat was swallowed by IBM. KDE Plasma 5 is my thing. Windows 10 is a VM guest in Arch host. 32GB workstation, multiple monitors - 2 24 inch Dell monitors, 1 Haeir 55 inch, SSDs and Western Digital Caviar Blacks and Transcend External HDDs... no issues to report.
Same experience with Fedora for more than 2 years. FreeBSD with Xfce 4 and TWM on my laptop. No crashes, no issues.
If Linux had a stable and consistent base system that was comprised of the most commonly used libraries, then developers would only have to statically link or include their own copy of less common libraries. I'm totally ok with that.
And let's be real for a minute, the program size hit is negligible. You can get a 1TB SSD for under $200. The largest part of almost any program that's in any way large are its assets, not its code.
Windows allows multiple instances of the same app to work via the WinSxS mechanism which keeps copies of basically every DLL ever installed, including system DLLs. So it is basically keeping duplicates of everything, regardless. This includes some executables as well. [1]
Before WinSxS you could achieve the same effect by keeping copies of the DLL dependencies in the app directory beside the executable. Windows would then choose the DLLs from the app current directory before looking at system directories and the path. [2]
Linux gives you the same choices and a couple more options.
You can't do the last two with Windows, for the most part because you do not have source code.So, which of those options do you choose?
You cannot have your cake and eat it too.
Only as the builder of an application. The application developers for Linux, on the whole, only ever chose option one, leaving everyone who uses their software no option that doesn't involve recompiling from source (which, as I've mentioned, is a ridiculous thing to expect people to do just to use your software).
I have several dozen applications on Windows that are not installed. Many of them aren't even designed to be portable, yet were easily made such just by extracting their files from the installer into a directory. I can put these applications on a thumb drive and take them to an entirely different Windows system and run them without installing them their either. These directories are exactly as they'd exist in C:\Program Files.
This is of course because they have copies of their dependent DLLs in their install directory, but only the ones that aren't part of the Windows distribution. Let's look at one: WinDirStat. It is 2.1MB total, 856KB for the executables, 844KB for dlls, and the rest is assets, docs, and configuration. This is a GUI application, but you'll notice it doesn't have to include a toolkit because it uses the one that comes with every Windows install. The DLLs, as far as I can tell, are unique to WinDirStat, so it wouldn't be unfair to say that this application actually doesn't have any dependencies other than the base OS.
That's what Linux can't get you. There is no base OS above syscalls. You can't depend on anything being there because there are hundreds of distributions and what's available to link with changes with configuration, distribution, and often version.
If I could, I'd have all my Linux applications as AppImages which follows pretty much the same model as portable applications do on Windows. As I said though, Linux developers rarely offer that as an option.
Again, wrong.
The people that package distributions usually make these choices. Gentoo and distros like it give you a choice of how things are built.
Off the top of my head I cannot think of a popular library that does not have the option to be statically or dynamically linked.
>> I have several dozen applications on Windows that are not installed. Many of them aren't even designed to be portable, ....
>> ... The DLLs, as far as I can tell, are unique to WinDirStat, so it wouldn't be unfair to say that this application actually doesn't have any dependencies other than the base OS.
This works in Linux too, somewhat. You have to have compatible ABIs though.
>> ... dependent DLLs in their install directory........
And Windows has 8 tons of different versions of system DLLs and C library runtimes in the WinSxS directories. MSVC runtimes are not interchangeable. Linux in general does not go to that length since a large majority of the apps are source available.
But afaik, this usually isn't an option in the Linux world. Unless, again, I'm compiling stuff myself.
Flatpak has you covered, or at least, it's getting there pretty fast. You'll still need to manage debs for base system stuff but you can install most gui apps as flatpaks now.
Also, flatpak has a lot of problems in my opinion, not the least of which is that it has no mechanism for installing to a different disk without creating an entirely new "installation", which in its terminology means a completely new dependency chain too.
The "portals" idea is a pretty good one though, on the whole.
I hear you on installing to different disks, though. Appimages are also an option, but I've found they have a tendency to not actually work between different distros and system versions.
One thing that would be interesting would be to use flatpak "runtimes" to provide a target base system for AppImages, but I don't think that level of cooperation is actually possible within the Linux community.
I don't really think it's an issue though, if you really want portability you shouldn't rely on any software provided by the distribution unless you want to depend on that specific distro.
Flatpak runtimes are basically interchangeable base systems, it seems silly to commit to any one in particular when you can use any number in parallel.
OEM Licenses are bound to a single computer AFAIK. You change computers you need to pay for Windows again. So it's not like people will never buy Windows anymore as long as they buy computers installed with Windows on it.
Before one would get a separate Windows CD with his computer and be able to install it on many different machines, one activation at a time.
I would hope that this would allow them to split the codebase, and make other, paid versions (professional, enterprise) that didn't have ANY of the tracking or advertising -- and didn't require extensive GPO manipulation to turn off -- but that's probably too much to ask.
They actually promoted "signature" PCs which were supposed to come without crapware. I can recall buying a Windows 8 tablet from a brick-and-mortar Microsoft store because I knew it would come with a hassle-free clean install.
I have to go to far more effort for my email accounts that are still web based, having to sign in every so often and do the 2fa dance.
Exactly how many dollars more would I have to pay for the Win10 license to not see ads? I'd like that thanks.