Found this a while back and thought it was an interesting perspective vis a vis some of the collective (mine included) perceptions of Microsoft based on their past behavior, illegal and unethical business practices and such. Not to necessarily excuse or erase that, but Nadella-era MSFT is very different than Gates- or Ballmer-era
There's an important lesson here: while reasonable skepticism is always good, it's important to remember that a policy of "never forgive, never forget" actually creates moral hazard. If people/companies think they will never be forgiven for current misdeeds even if they get better, then there's no incentive to get better and they might as well keep being horrible. Allowing people (or companies) to turn over a new leaf incentivizes good behavior.
I don't think it is reasonable to expect people completely disconnected from the negative effects of their actions to "turn over a new leaf", especially when they stand to make hundreds of thousands of dollars or more.
From all the examples I can see, "ask for forgiveness, not permission" is how many of the largest companies have become economically successful.
The point is, managers don't think in moral terms. They think in terms like risk vs. market opportunity. I've never heard a phrase like "But wouldn't it be immoral?" during a board meeting. So, if they feel no wrong was ever done, what should they ask forgiveness for?
This is flawed reasoning when you understand that some people and some companies are sociopathic. They will never 'turn over a new leaf' because they are incapable of doing so.
Nadella-era MSFT is different from Gates- or Ballmer-era, and this is very good, but, as a Linux user and free software contributor, I wouldn't say I feel "loved". I'd say I feel MSFT is trying to somehow get advantage of my work (which is fine, given that it is free software) while holding back what I would like to have from them (specifications for MSFT interfaces and file systems, if not code, so that better interoperability can be implemented). Which is also fine, of course, they do not owe me anything, but still I do not feel "loved".
(BTW, I already have more editors and code hosting platforms than I need, so they are not good points for me)
The claim that MS is "putting the Linux kernel at the heart of Windows 10 as part of WSL 2" is an overstatement. If they really wanted to put the Linux kernel at the heart of Windows, they would ditch the Windows kernel, which would be no great loss, and make Windows a GUI over the Linux kernel.
The only problem with this is that there are thousands of business-critical applications that are designed for the Windows NT kernel... Many of them have deep integrations with the operating system.
Just think of all the Steam games that are only available on Windows!
And yet programming languages ever since C were supposed to be cross platform. If you have to call on specific features of a particular OS when writing business application code, that's a problem with the OS.
> And yet programming languages ever since C were supposed to be cross platform.
Some were, plenty were not, and C is nothing special as a turning point in that.
> If you have to call on specific features of a particular OS when writing business application code, that's a problem with the OS.
No, insofar as it's a problem, it's a problem with the application language or libraries, in that they don't abstract sufficiently the deployment platform.
Now, if they are bundled with the OS, it may be the same product as the OS, but it's not really an OS. problem.
Post-C languages not made to be cross platform? Well, a whole host of application-bundled languages that were tied to apps that were their sole platform (and often those apps were OS-specific, too.) For starters.
Some features that Windows did better than Linux/Unix:
* IOCP is loads better than epoll
* SEH is a better model for synchronous processor instructions than Unix signals
* Windows allows some better/easier ways to introspect or modify other processes than Linux does. You can't change another processes memory map on Linux, whereas you can on Windows. And /proc is a pretty poor interface for getting process data, IMHO.
Not OP, but I do prefer how Windows handles high memory loads (by not becoming utterly unresponsive and forcing me to force reboot my system). It’s 2019. I don’t get how this kind of OS behavior is acceptable. Linux is the only environment where I have to micromanage my memory because I’m afraid I’ll hit that memory limit where everything stops working.
> The value of the NT kernel is absolutely enormous, probably on the order of a trillion dollars
No, the value of the Windows installed base is enormous. But that does not mean the NT kernel is the reason for that value. People, and more importantly large corporations that are the primary customers for Windows, don't buy it because it has a technically superior kernel. They buy it because MS has had lock-in on the PC OS market for several decades now, due not to technical superiority but to IBM making two bonehead decisions that handed that market to MS, and MS has used its market position to force PC manufacturers into bundling agreements that increase the cost of using any other OS to the point where corporations can't justify doing it.
In other words, the value of the Windows installed base is that, because of historical events that I would much rather had gone another way but that's water under the bridge now, to a first approximation, Windows is the interface that everybody knows how to use to use a PC. The point of my comment was that it would be an improvement to put that known interface that has huge market lock-in on top of an OS kernel that didn't have all the known security and other technical flaws of Windows.
I'm old enough to remember Bill Gates appearing as a borg on Slashdot and the infamous mutating linux ads MS ran when Win2K came out (https://images.app.goo.gl/Tqj5xPmgueJotQuN7) but I agree with Linus. The landscape has changed and the money today is in SaaS and Cloud.
MS missed the mobile revolution but they appear intent on ensuring their slice of the "cloud" pie through Azure. Making sure they support the linux ecosystem is a no brainer (and I think they're far too late to embrace and extend).
Having already lost browser, mobile and search; they'd be dead if they fucked up cloud too.
>I think they're far too late to embrace and extend
Isnt that exactly what their new android phone will become? Android that also runs Windows apps?
>Having already lost browser, mobile and search; they'd be dead if they fucked up cloud too.
In this case, Cloud is saas and iaas/paas. On the saas side, they have a huge industry lead with exchange/sharepoint, and transitioned with a similar playbook to Adobe. Theres too many customers that wont leave Outlook/Excel, and having hosted Outlook/Excel storage native to the programs themselves, is enough for many companies to make an overall vendor choice. Because of Ballmers licensing genius, once you buy Outlook/Excel, bundle pricing makes piecemealing with competitors products prohibitivly more expensive. Winning iaas is mostly just about how much money they dump into capital. Winning paas, if one can win and become the defacto api standard, isnt something Microsoft needs to survive. They could "lose" to aws, adopt all its conventions, and be a drop in replacement. I dont think that would happen, but winning the paas battle, as the only survivor isnt essential to Microsofts survival.
Microsoft also has one other push that few others have brought a challenge against, and that is corporate identity. With AD transitioning to AzureAD, and Office 365 being that gateway drug, companies will be invested into Microsoft's Identity services for decades to come. Transitioning to a higher priced product would net nearly zero gain. Okta might be nice if you dont already have anything, but it wont outgrow msft. At least second place and thrown into a bundle is the msft motto. Time will tell if that model translates to Azure paas pricing.
Microsoft wants to compete with Amazon and Google, and god bless them if they've been humbled by their recent failures and want to play fair now. I'd be more concerned about IBM, which through its acquisition of Red Hat now controls or funds most of the software in the "traditional" Linux environment.
Whether Microsoft will stay like this just depends largely on their current strategic positioning and how their organization is set up to make money now, rather than how they acted in the past.
If the incentives are aligned, they'll be well behaved. If not, then they'll do whatever they can.
I remember when everyone was a Google, Facebook and Apple fanboy, and those tunes sure changed. I also remember when the whole Open Source/Free Software community loathed Microsoft. Maybe the lesson in all this is, don't assume what a ginormous capitalist corporation will do based on your feelings about their brand.
Apple is still good. But their business model is aligned to be good stewards of the user unlike Facebook or Google. Microsoft and Apple products are irreplaceable today as well, while Google and Facebook offer convenience but no necessary value.
I think you miss my point... an entire company can't be "good". "Apple" is a brand. You may like a lot of their products, but they also have some products with shit features, and they sometimes do "things" that are not so much "good" for "the users" as they are good for the company.
Their MacBook Pro keyboards are the worst in the industry. They make it so a user can't repair a product that they own. They censor application platforms based on the country a user is in. The majority of their products are a "captive market", often impossible to use with anything but another Apple product. There's many things Apple does that are against the users. But here you are, enjoying some of their products, so you call their entire brand, and presumably their entire product line, "good".
How does one define "good", then? Good enough for all? For some? For you? Good in quality? Good morally or ethically, and if so, based on whose morals/ethics? Do the people and places they effect outside of their immediate users matter? I don't think it's useful to think of an entire brand in such a simplistic way.
BTW, Google and Facebook offer necessary value. Facebook is used to keep up with people's lives (without it, many grandparents would know next to nothing of the lives of their grandchildren, for example) and encompasses two of the largest instant messaging platforms. Google is probably the biggest provider of e-mail, an absolutely essential component of having an online life (which is increasingly necessary for more of regular life, from getting a ride somewhere, to accessing government services, to transferring funds), to say nothing of search, maps, docs. And many of these services would be pointless without the Apple/Google/Microsoft platforms, but many of those platforms would be pointless without the services. So what's "necessary" is a bit subjective.
Facebook/Google are both easily replaceable, that's why they have no intrinsic value they create. I can use Outlook.com or iCloud.com and I have the original social network (or, pickup the phone or write a letter). DuckDuckGo grants me search. Altavista was pretty good too.
I don't prefer all of Apple products, although I do like them all. I use a Thinkpad X1 Extreme for work, and also have a 2019 Samsung Notebook 9 Pen. We do use iPhones in this house. I thought "good" was a light-way of putting. Overall, all things considered I'd go up a notch to "best".
You may have been rattled by me calling Apple "good", but if I have to light a fire on all of these companies except one, I'd leave Apple intact. I could easily and happily use their entire lineup. The Macbooks aren't "that bad". ;) Hardly junk, and even their lowend isn't junk. I like the Macbook Air quite a bit and recommend it often. Quality across the entire line is what I expect from everyone, but it's why Apple is the best. They are even relatively privacy conscious. Even if I'm a best of breed person.
Apple is so dominant in the quality realm that I would say their products are best for everyone. You can't go wrong with a Mac, especially when sorting through all the junk on the market to get to my X1 Extreme and Notebook 9 Pen is honestly, fraught with peril. It requires good judgement and people often fail at that. They will go for the lower-lines because they think they're the "same" as the top of the line, and that's very much not true, outside of Apple.
I would recommend most people just go to the Apple Store. None of it is junk, and you have a store to take it to for repairs if you do have problems. Can't beat it. Better than "good"- best.
If I were dying, I would tell my wife to keep an eye on Consumer Reports, but that she can safely continue to invest in Apple products for the foreseeable future after my death. That's how strong their track record has been. I am one that believes an organization can earn goodwill, not just anger and bad will.
The article leaves out the minor detail that through sponsoring the Linux Foundation, Microsoft is paying Linus' paycheck.
That being said, I'm really excited about the apparent culture shift inside Microsoft. Let's forget about the ugly past and judge them by their recent actions.
For me? Absolutely. I hope no one is judging me based on my great grandfather's sins. That would be highly unproductive to treat people as if they were born having done humanity a great wrong.
63 comments
[ 2.9 ms ] story [ 130 ms ] threadhttps://www.hanselman.com/blog/MicrosoftKilledMyPappy.aspx
There's an important lesson here: while reasonable skepticism is always good, it's important to remember that a policy of "never forgive, never forget" actually creates moral hazard. If people/companies think they will never be forgiven for current misdeeds even if they get better, then there's no incentive to get better and they might as well keep being horrible. Allowing people (or companies) to turn over a new leaf incentivizes good behavior.
From all the examples I can see, "ask for forgiveness, not permission" is how many of the largest companies have become economically successful.
(BTW, I already have more editors and code hosting platforms than I need, so they are not good points for me)
Just think of all the Steam games that are only available on Windows!
Some were, plenty were not, and C is nothing special as a turning point in that.
> If you have to call on specific features of a particular OS when writing business application code, that's a problem with the OS.
No, insofar as it's a problem, it's a problem with the application language or libraries, in that they don't abstract sufficiently the deployment platform.
Now, if they are bundled with the OS, it may be the same product as the OS, but it's not really an OS. problem.
Which ones were not?
Ah, ok. I meant general purpose languages, not languages specialized to particular applications. I should have made that clear.
I am sure you mean "designed for Win32 API."
They mean that those apps are currently deeply integrated with NT and switching is very expensive for the developers. See most games as an example.
[see also https://news.ycombinator.com/item?id=11865760 for a lively discussion on this topic]
* IOCP is loads better than epoll
* SEH is a better model for synchronous processor instructions than Unix signals
* Windows allows some better/easier ways to introspect or modify other processes than Linux does. You can't change another processes memory map on Linux, whereas you can on Windows. And /proc is a pretty poor interface for getting process data, IMHO.
http://blog.zorinaq.com/i-contribute-to-the-windows-kernel-w...
You instantly lost all credibility by saying such an obvious Linux fanboy statement, ignoring reality and business constraints.
The value of the NT kernel is absolutely enormous, probably on the order of a trillion dollars.
Remind me: Which year was the year of the Linux desktop?
No, the value of the Windows installed base is enormous. But that does not mean the NT kernel is the reason for that value. People, and more importantly large corporations that are the primary customers for Windows, don't buy it because it has a technically superior kernel. They buy it because MS has had lock-in on the PC OS market for several decades now, due not to technical superiority but to IBM making two bonehead decisions that handed that market to MS, and MS has used its market position to force PC manufacturers into bundling agreements that increase the cost of using any other OS to the point where corporations can't justify doing it.
In other words, the value of the Windows installed base is that, because of historical events that I would much rather had gone another way but that's water under the bridge now, to a first approximation, Windows is the interface that everybody knows how to use to use a PC. The point of my comment was that it would be an improvement to put that known interface that has huge market lock-in on top of an OS kernel that didn't have all the known security and other technical flaws of Windows.
Of course linux (the kernel) also has a pretty huge phone marketshare, with android which uses the Linux Kernel.
I suppose that Linus might continue working to further a case he cares about.
MS missed the mobile revolution but they appear intent on ensuring their slice of the "cloud" pie through Azure. Making sure they support the linux ecosystem is a no brainer (and I think they're far too late to embrace and extend).
Having already lost browser, mobile and search; they'd be dead if they fucked up cloud too.
And XBill, don't forget about XBill!
https://en.wikipedia.org/wiki/XBill
- VSCode
- GitHub
As well. They are fighting for the hearts and minds of the next generation of developers. They might win.
Isnt that exactly what their new android phone will become? Android that also runs Windows apps?
>Having already lost browser, mobile and search; they'd be dead if they fucked up cloud too.
In this case, Cloud is saas and iaas/paas. On the saas side, they have a huge industry lead with exchange/sharepoint, and transitioned with a similar playbook to Adobe. Theres too many customers that wont leave Outlook/Excel, and having hosted Outlook/Excel storage native to the programs themselves, is enough for many companies to make an overall vendor choice. Because of Ballmers licensing genius, once you buy Outlook/Excel, bundle pricing makes piecemealing with competitors products prohibitivly more expensive. Winning iaas is mostly just about how much money they dump into capital. Winning paas, if one can win and become the defacto api standard, isnt something Microsoft needs to survive. They could "lose" to aws, adopt all its conventions, and be a drop in replacement. I dont think that would happen, but winning the paas battle, as the only survivor isnt essential to Microsofts survival.
Microsoft also has one other push that few others have brought a challenge against, and that is corporate identity. With AD transitioning to AzureAD, and Office 365 being that gateway drug, companies will be invested into Microsoft's Identity services for decades to come. Transitioning to a higher priced product would net nearly zero gain. Okta might be nice if you dont already have anything, but it wont outgrow msft. At least second place and thrown into a bundle is the msft motto. Time will tell if that model translates to Azure paas pricing.
You can actually buy RedHat support from them also.
https://www.suse.com/c/know-suse-supporting-rhat-years/
If the incentives are aligned, they'll be well behaved. If not, then they'll do whatever they can.
Whether Microsoft will stay like this depends on how they respond to the changing landscape moving forward.
I think you miss my point... an entire company can't be "good". "Apple" is a brand. You may like a lot of their products, but they also have some products with shit features, and they sometimes do "things" that are not so much "good" for "the users" as they are good for the company.
Their MacBook Pro keyboards are the worst in the industry. They make it so a user can't repair a product that they own. They censor application platforms based on the country a user is in. The majority of their products are a "captive market", often impossible to use with anything but another Apple product. There's many things Apple does that are against the users. But here you are, enjoying some of their products, so you call their entire brand, and presumably their entire product line, "good".
How does one define "good", then? Good enough for all? For some? For you? Good in quality? Good morally or ethically, and if so, based on whose morals/ethics? Do the people and places they effect outside of their immediate users matter? I don't think it's useful to think of an entire brand in such a simplistic way.
BTW, Google and Facebook offer necessary value. Facebook is used to keep up with people's lives (without it, many grandparents would know next to nothing of the lives of their grandchildren, for example) and encompasses two of the largest instant messaging platforms. Google is probably the biggest provider of e-mail, an absolutely essential component of having an online life (which is increasingly necessary for more of regular life, from getting a ride somewhere, to accessing government services, to transferring funds), to say nothing of search, maps, docs. And many of these services would be pointless without the Apple/Google/Microsoft platforms, but many of those platforms would be pointless without the services. So what's "necessary" is a bit subjective.
(Caveat to all of the above: Fuck Oracle.)
I don't prefer all of Apple products, although I do like them all. I use a Thinkpad X1 Extreme for work, and also have a 2019 Samsung Notebook 9 Pen. We do use iPhones in this house. I thought "good" was a light-way of putting. Overall, all things considered I'd go up a notch to "best".
You may have been rattled by me calling Apple "good", but if I have to light a fire on all of these companies except one, I'd leave Apple intact. I could easily and happily use their entire lineup. The Macbooks aren't "that bad". ;) Hardly junk, and even their lowend isn't junk. I like the Macbook Air quite a bit and recommend it often. Quality across the entire line is what I expect from everyone, but it's why Apple is the best. They are even relatively privacy conscious. Even if I'm a best of breed person.
Apple is so dominant in the quality realm that I would say their products are best for everyone. You can't go wrong with a Mac, especially when sorting through all the junk on the market to get to my X1 Extreme and Notebook 9 Pen is honestly, fraught with peril. It requires good judgement and people often fail at that. They will go for the lower-lines because they think they're the "same" as the top of the line, and that's very much not true, outside of Apple.
I would recommend most people just go to the Apple Store. None of it is junk, and you have a store to take it to for repairs if you do have problems. Can't beat it. Better than "good"- best.
If I were dying, I would tell my wife to keep an eye on Consumer Reports, but that she can safely continue to invest in Apple products for the foreseeable future after my death. That's how strong their track record has been. I am one that believes an organization can earn goodwill, not just anger and bad will.
[1] https://www.zdnet.com/article/linus-torvalds-isnt-worried-ab...
That being said, I'm really excited about the apparent culture shift inside Microsoft. Let's forget about the ugly past and judge them by their recent actions.
They're arguably worse than they were 10-20 years ago IMO... throwing some Linux support on top of stuff just makes me scoff.
The relatively more purist views of the FSF are of value, as well.
A spectrum of opinion keeps any Towers of Babel from getting too tall.