"There are a lot of people who’d be a lot happier if they stopped worrying about other people’s 800 pound gorillas."
I wonder how this sentiment would apply to linux DEs. Consider the design decisions which would be made if the designers of gnome and kde thought about what would make them most productive instead of thinking about what will get people to switch from windows to linux. Perhaps, we will get a lot less dumbing-down of interfaces and an end to blind adherence to the standard windows UI and finally see some genuine innovation in this field again.
This is not to say that all DEs are stuck in a cycle of continuously copying windows. FVWM (my wm of choice) splits completely. Similarly, from an eye-candy perspective, beryl and compiz bring new things to the table. Still, beryl and compiz are relatively small projects and the vast majority of coding effort goes into make nice GUIs for systems configuration and things like emulating MS Office with OOo.
If you want to go mainstream, anybody's gorilla is your gorilla.
There's no reason why you can't innovate along the way, but it's pretty clear to me that if Linux is ever to make any inroads into non-geek circles, it'll need a familiar interface.
But that is my point exactly. Why is making inroads into the non-geek circles such a high priority? Most of the cash comes from voluntary donations and some very rare instances of funding. Most of the work performed on Linux kernel and associated software is not done for profit. What if the devs working on kde right now instead started to contribute to integrating reiser4 into the kernel or building up btrfs or suspend support or any of the dozens of possible improvements to the linux kernel directly beneficial to the devs themselves? It could have some interesting effects.
That's at least partially because you can make Windows and Linux look like Mac OS X, but there's no practical way of making Mac OS X look like Windows or Linux.
I think that if you made something that made Mac OS X look like one or the other, very few people would bite.
OS X has the most beautiful aesthetic of any computer I know. But it's not just the looks that matter. It's the functionality. You don't have a taskbar and an in-window menu bar. You have a dock and a top menu. And that's the best system I've ever worked with. You can't copy that over to Windows or Linux effectively, either.
If a Linux distro was smart, it would emulate OS X rather than Windows, because fundamentally, the design is more efficient and more usable. As long as they copy Windows they're at a disadvantage.
>I think that if you made something that made Mac OS X look like one or the other, very few people would bite.
Only the absolute minority of people make their Windows or Linux machines to look like Mac OS X either - you're looking at a minority of enthusiasts using the fringes of operating system functionality and you're pointing to that as definitive proof that something is better. To me, that's outright flawed, especially when there are core software limitations that prevent you from imitating the Windows or GNOME UIs in Mac OS X.
You can have a top-menu and a dock in Linux - there's pre-built tools that let you do that (e.g. https://wiki.ubuntu.com/global_menu, activate the MacStyle menu setting in KDE, combine them with a dock app like Avant Window Navigator). Only an small minority of people use them and it's not very well supported because of it, but it's certainly possible. The Linux desktop is incredibly configurable.
Note that I haven't actually stated my personal opinion on the matter at all so far. I think you're suffering from a case of self-confirmation bias.
Did I act like you stated your personal opinion? Sorry if it seemed that way.
You can have that in Linux, but it's one thing for it to be possible and another for it to be the out-of-the-box default. It's the problem with choice: it's confusing and it takes time and other programs don't configure to the same standards.
I was trying to contrast your point-of-view, that you prefer OS X, in contrast to the fact that there's little empirical evidence that OS X is an outright better model or that there's little evidence that lots of people are going out of their way to imitate it. Sorry if that didn't come across very well.
My personal opinion is that for every advantage that OS X has via its menubar and dock layout it has an equal disadvantage, so both styles are equal.
For example, for the menus at the top gives you the advantage of Fitts' law but makes accessing menus while multitasking and switching between applications and using menus more difficult. Moving from looking at the top of the screen to the window is a further distance, which may make viewing slower. You do get the advantage of a more intuitive MDI applications though.
The dock is better for frequently used applications, but it uses more screen space, the locations of applications in the dock shift as you use it and it uses dramatically more space than the quick-launch bar. It also stylistically makes maximisation look rather 'off'.
If you solely rely on keyboard navigation, Windows-style is probably better - you can easily load an application without the mouse (especially in Vista), plus you primarily use alt+tab/cmd+tab which means that the menu is largely wasted space. If you use many applications, the Start Menu is likely quicker than using Finder, especially on Vista with its built-in search. Of course, there's Quicksilver and their ilk but that's not standard and similar applications are available cross-platform.
Finally, if you believe they're approximately equal, if you're developing a new OS from scratch with no existing users and you have to choose between the two it's probably better to lean to lean towards a Windows-style layout as more people are familiar with it to narrow the learning curve, which would make choosing Windows-style windows as the default a reasonable decision.
I'm certainly not saying this was the thought process at all - let's not forget Mac OS X was years from release when the GNOME and KDE projects were started and they were building on top of the paradigms of existing X applications - but it certainly shows that if you were making the decision today choosing the Windows model would still be a reasonable one, and that's excluding the factors involved in changing like alienating the current user-base.
Can't speak for KDE, but your statement regarding GNOME is totally incorrect. We follow our own UI guidelines, we do NOT blindly follow the Windows guidelines. The focus is on our thoughts on usability, not on being different. Further, people work on whatever they want.
However, Gnome takes a large part of its interface layout directly from Windows. The taskbar and the window layout always seemed to be direct copies of that system. Instead of crating something new and more functional, Gnome and KDE (though I haven't tried 4) seem to just lift Windows' ideas wholesale at times, and merge them with the Linux ideas of app and system management.
For one, GNOME is usually compared with Mac OS X. I find your comparison with Windows pretty funny. Secondly, look at the gnome-shell discussion on planet.gnome.org. Further, just because Windows has things looking like GNOME doesn't mean one copied from the other. Especially not that GNOME copied from Windows. Click e.g. on the clock in GNOME. IIRC that is now in Vista.
If Microsoft had anything worth pirating, I'd pirate it, just to get even close to paying myself a reasonable hourly rate for the time I've had to spend debugging css for IE6.
This isn't some badass act of defiance against The Establishment like Gruber seems to want you to think, given his last paragraph: 4% of his viewers use IE.
If that went up overnight to 40%, you know whose gorilla it would become.
the point is, the likelihood of that happening is pretty close to zero. gruber has aligned himself with the tech elite, and the early adopters.
suppose some new version of firefox was released that sucked, and broke everybody's css layouts. gruber's crowd would almost certainly not adopt that particular version. they value technical superiority above almost everything else.
"It’s easier for me to ignore IE because the DF audience predominantly uses Safari, Firefox, and MobileSafari (roughly 53, 25, and 8 percent, respectively, with IE coming in around 4 percent). I have no idea whether the DF Paraphernalia store is even legible under IE, because I didn’t even bother to check."
Perhaps if the site weren't broken in IE, half of his visitors would be from that browser.
To the people who downmodded this, is it just that you hate IE a lot, or do you have an actual reason why I'm wrong?
I wasn't being sarcastic -- don't you think he's turning away his IE visitors by presenting a site that doesn't look right, and then turning around and claiming that IE is 4%?
Even for sites with a highly technical audience, that number is hard to believe.
I downmodded you because what you said was asinine. That I hate IE a lot and have an actual reason why you're wrong is mere coincidence.
It's probable that the decision to stop caring about IE was made in light of similar stats. Note that he said he no longer concerns himself with IE, and that his stylesheet still contains IE-specific hacks. He's done nothing to intentionally dissuade any hypothetical IE users that might come along, and if IE's rendering gets worse over time, that only further justifies his decision to stop caring. On the other hand, if IE's rendering gets better over time, then not caring about IE won't be a problem because IE will render more like modern browsers.
It's also important to be clear that the site does look right. If simply not catering to one particular browser causes your site to be completely unusable in that browser, that's a strong indication that the browser vendor is incompetent.
Given that his site focuses almost exclusively on news for the early-adopter Apple enthusiast, I would guess that he would naturally have a much lower proportion of IE users in his audience than other sites, even if his site worked flawlessly in IE.
Note, for example, that he tests his site against both Safari and Firefox, so we can assume he is not turning away those users. If his site attracted the "average" web surfer, one would expect roughly three times as many Firefox users as Safari users. But he actually attracting twice as many Safari users as Firefox. Since his non-IE traffic is so heavily skewed to Safari, it's not unimaginable that his IE traffic would be similarly skewed, regardless of any IE rendering problems.
That isn't the case, because his site isn't broken in IE. All he said was that he doesn't care to check any more. I used to access it when I was at school and IE was the only option, and the only thing that didn't work was the unicode character he uses to return from a footnote to where it is referenced. The reason that IE's share is so low is the nature of the web site and the type of people who read it.
So he's saying that if your core audience is in Windows, and your product works best in Windows, you should ignore the Linux people clamouring for a port?
It goes both ways - you ignore them today, they'll ignore you tomorrow.
That's only if you want an OS that's not windows but acts like windows in most manners of using. If you want something different, it still takes more then an hour or two to set up.
Note that the author of the original "800-pound gorilla" posting was talking about an open-source product he maintains. He said he was happy to accept patches from Windows developers who had figured out ways to make the product work on Windows.
I certainly wouldn't criticize any maintainer of an open-source Windows application for saying the same thing in the other direction.
You know, by that logic anybody developing an application for a single operating system is being a fool. And yet: most of the applications I use on OS X are developed exclusively for my operating system. See: anything Panic has ever made. See: Delicious Library. And they're obviously made by developers that love my system, and they're making their developers a ton of money, so why should they bother porting it over?
Considering Hacker News talks a lot about appealing to a niche rather than trying to appeal to everybody, I find this comment a little irritating.
So much for cross browser accessibility. At the end of the day its his site and he can choose which browsers he wants to ignore, but I think even ignoring that 4% who use IE is a big mistake.
And its not like 'daring fireball' would be all that difficult to make IE render correctly anyways.
Daring Fireball does render almost correctly. And why should Gruber care? He makes a living off of his web site alone as it is. Worrying about 4 percent when you're writing for your own site and it's self-sufficient, when worrying about that slim percentage means a lot of pain, isn't exactly a waste of time, but: why bother?
jamis has the right to support whatever he wants, it's his project that he designed for his purposes. if windows developers need to tweak for windows, they should fork capistrano. i guess jamis is saying, fork you! (ha)
4% of his users are IE, let's say, he's running a multi million dollar business. It'll cost him 5 million a year to keep IE working, and the business from the 4% equates to 15 million. What would you do? Beat your chest and scream "look at my website I'm awesome!".
Jamis had a legitimate point. DaringFireball is hijacking this to pound his chest. Who _cares_. For some people and companies, supporting IE in some incarnation is legitimate. For others, it's not. That's all very reasonable.
I for one am sick of these fallacious arguments masquerading as rational and revolutionary thought. Peh. Who cares.
Microsoft is an 800 lb gorilla, but one located in another part of town, or even another country. If you do web development, and you're not using .net/CLR/C#, then you're probably not developing or deploying on Windows.
What proportion of startups run Windows? It's a good deal lower than those running Linux or OS X -- so much so that Microsoft have launched a "BizSpark" service where startups can use their software without charge (see http://www.includipedia.com/blog/2008/11/05/microsoft-bizspa... )
So if you want to develop with PHP, MySQL, Ruby, Rails, Django, etc, then you should bear in mind that Windows support is probably not going to be as good as Linux or OS X support.
39 comments
[ 2.3 ms ] story [ 78.0 ms ] threadI wonder how this sentiment would apply to linux DEs. Consider the design decisions which would be made if the designers of gnome and kde thought about what would make them most productive instead of thinking about what will get people to switch from windows to linux. Perhaps, we will get a lot less dumbing-down of interfaces and an end to blind adherence to the standard windows UI and finally see some genuine innovation in this field again.
This is not to say that all DEs are stuck in a cycle of continuously copying windows. FVWM (my wm of choice) splits completely. Similarly, from an eye-candy perspective, beryl and compiz bring new things to the table. Still, beryl and compiz are relatively small projects and the vast majority of coding effort goes into make nice GUIs for systems configuration and things like emulating MS Office with OOo.
There's no reason why you can't innovate along the way, but it's pretty clear to me that if Linux is ever to make any inroads into non-geek circles, it'll need a familiar interface.
i think the beauty of open source is you can work on what you love, even if it's not efficient
i see a lot of "make Windows looks like Mac OS X" from google or even "make Linux look like Mac OS X"
i believe I've never seen the reverse (OSX->xp or even OSX->linux)
to me linux is copying the wrong model, her role model should be osx
OS X has the most beautiful aesthetic of any computer I know. But it's not just the looks that matter. It's the functionality. You don't have a taskbar and an in-window menu bar. You have a dock and a top menu. And that's the best system I've ever worked with. You can't copy that over to Windows or Linux effectively, either.
If a Linux distro was smart, it would emulate OS X rather than Windows, because fundamentally, the design is more efficient and more usable. As long as they copy Windows they're at a disadvantage.
Only the absolute minority of people make their Windows or Linux machines to look like Mac OS X either - you're looking at a minority of enthusiasts using the fringes of operating system functionality and you're pointing to that as definitive proof that something is better. To me, that's outright flawed, especially when there are core software limitations that prevent you from imitating the Windows or GNOME UIs in Mac OS X.
You can have a top-menu and a dock in Linux - there's pre-built tools that let you do that (e.g. https://wiki.ubuntu.com/global_menu, activate the MacStyle menu setting in KDE, combine them with a dock app like Avant Window Navigator). Only an small minority of people use them and it's not very well supported because of it, but it's certainly possible. The Linux desktop is incredibly configurable.
Note that I haven't actually stated my personal opinion on the matter at all so far. I think you're suffering from a case of self-confirmation bias.
You can have that in Linux, but it's one thing for it to be possible and another for it to be the out-of-the-box default. It's the problem with choice: it's confusing and it takes time and other programs don't configure to the same standards.
My personal opinion is that for every advantage that OS X has via its menubar and dock layout it has an equal disadvantage, so both styles are equal.
For example, for the menus at the top gives you the advantage of Fitts' law but makes accessing menus while multitasking and switching between applications and using menus more difficult. Moving from looking at the top of the screen to the window is a further distance, which may make viewing slower. You do get the advantage of a more intuitive MDI applications though.
The dock is better for frequently used applications, but it uses more screen space, the locations of applications in the dock shift as you use it and it uses dramatically more space than the quick-launch bar. It also stylistically makes maximisation look rather 'off'.
If you solely rely on keyboard navigation, Windows-style is probably better - you can easily load an application without the mouse (especially in Vista), plus you primarily use alt+tab/cmd+tab which means that the menu is largely wasted space. If you use many applications, the Start Menu is likely quicker than using Finder, especially on Vista with its built-in search. Of course, there's Quicksilver and their ilk but that's not standard and similar applications are available cross-platform.
Finally, if you believe they're approximately equal, if you're developing a new OS from scratch with no existing users and you have to choose between the two it's probably better to lean to lean towards a Windows-style layout as more people are familiar with it to narrow the learning curve, which would make choosing Windows-style windows as the default a reasonable decision.
I'm certainly not saying this was the thought process at all - let's not forget Mac OS X was years from release when the GNOME and KDE projects were started and they were building on top of the paradigms of existing X applications - but it certainly shows that if you were making the decision today choosing the Windows model would still be a reasonable one, and that's excluding the factors involved in changing like alienating the current user-base.
For example: the GTK file picker dialog.
For a long time it was probably the only profitable Microsoft product that wasn't Basic, Office, or Operating Systems.
If that went up overnight to 40%, you know whose gorilla it would become.
suppose some new version of firefox was released that sucked, and broke everybody's css layouts. gruber's crowd would almost certainly not adopt that particular version. they value technical superiority above almost everything else.
Perhaps if the site weren't broken in IE, half of his visitors would be from that browser.
I wasn't being sarcastic -- don't you think he's turning away his IE visitors by presenting a site that doesn't look right, and then turning around and claiming that IE is 4%?
Even for sites with a highly technical audience, that number is hard to believe.
more than 1 in 25 people on a site like HN using IE6 would seem odd to me.. I'm not sure of anyone that would like this site and is still running ie6.
(on an assumption here that we're talking organic traffic - not SE results)
PG - want to give us the browser stats?
As time goes on, IE market share continues to slip. This process is vastly accelerated in technical discussion and news sites.
It's probable that the decision to stop caring about IE was made in light of similar stats. Note that he said he no longer concerns himself with IE, and that his stylesheet still contains IE-specific hacks. He's done nothing to intentionally dissuade any hypothetical IE users that might come along, and if IE's rendering gets worse over time, that only further justifies his decision to stop caring. On the other hand, if IE's rendering gets better over time, then not caring about IE won't be a problem because IE will render more like modern browsers.
It's also important to be clear that the site does look right. If simply not catering to one particular browser causes your site to be completely unusable in that browser, that's a strong indication that the browser vendor is incompetent.
Note, for example, that he tests his site against both Safari and Firefox, so we can assume he is not turning away those users. If his site attracted the "average" web surfer, one would expect roughly three times as many Firefox users as Safari users. But he actually attracting twice as many Safari users as Firefox. Since his non-IE traffic is so heavily skewed to Safari, it's not unimaginable that his IE traffic would be similarly skewed, regardless of any IE rendering problems.
It goes both ways - you ignore them today, they'll ignore you tomorrow.
I certainly wouldn't criticize any maintainer of an open-source Windows application for saying the same thing in the other direction.
Considering Hacker News talks a lot about appealing to a niche rather than trying to appeal to everybody, I find this comment a little irritating.
And its not like 'daring fireball' would be all that difficult to make IE render correctly anyways.
Jamis had a legitimate point. DaringFireball is hijacking this to pound his chest. Who _cares_. For some people and companies, supporting IE in some incarnation is legitimate. For others, it's not. That's all very reasonable.
I for one am sick of these fallacious arguments masquerading as rational and revolutionary thought. Peh. Who cares.
What proportion of startups run Windows? It's a good deal lower than those running Linux or OS X -- so much so that Microsoft have launched a "BizSpark" service where startups can use their software without charge (see http://www.includipedia.com/blog/2008/11/05/microsoft-bizspa... )
So if you want to develop with PHP, MySQL, Ruby, Rails, Django, etc, then you should bear in mind that Windows support is probably not going to be as good as Linux or OS X support.