Comments on: Cool Stuff in Foreign Realms http://blog.mozilla.com/tglek/2008/10/09/cool-stuff-in-foreign-realms/ Just another Blog.mozilla.com weblog Sun, 12 Feb 2012 07:38:09 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Mysterious Andy http://blog.mozilla.com/tglek/2008/10/09/cool-stuff-in-foreign-realms/comment-page-1/#comment-20084 Mysterious Andy Fri, 10 Oct 2008 11:18:50 +0000 http://blog.mozilla.com/tglek/?p=93#comment-20084 It hasn't been pointed out yet, so I thought I'd mention that Ubuntu and Fedora both have releases coming within 3-5 weeks. I wouldn't expect the Fedora Project or Canonical to make any big decisions about future plans until they are out of crunch time. I also note from the article: "At the conference roundup Friday, speaker Kyle McMartin announced that both Fedora and Ubuntu have fixed some delays in their boot process, and there was much applause." That sounds promising to me. Optimizing boot time for a netbook with limited and known hardware is always going to be easier than making a widely-used (desktop, workstation, laptop, netbook, ...) distro dance faster, but it sounds to me like ideas were exchanged and new things were tried by all. I dig that about Linux geeks. I agree wholeheartedly that I want a system that boots in less time than it takes me to drum my fingers; I just don't expect to see any announcements until the bigger players get a change to play around with these new ideas a bit. On a different note, didn't Ubuntu and Fedora both move from init to Upstart to speed up boot times by parallelizing the process? Kinda funny that these Intel guys did all their heavy lifting with init... It hasn’t been pointed out yet, so I thought I’d mention that Ubuntu and Fedora both have releases coming within 3-5 weeks. I wouldn’t expect the Fedora Project or Canonical to make any big decisions about future plans until they are out of crunch time.

I also note from the article:

“At the conference roundup Friday, speaker Kyle McMartin announced that both Fedora and Ubuntu have fixed some delays in their boot process, and there was much applause.”

That sounds promising to me.

Optimizing boot time for a netbook with limited and known hardware is always going to be easier than making a widely-used (desktop, workstation, laptop, netbook, …) distro dance faster, but it sounds to me like ideas were exchanged and new things were tried by all. I dig that about Linux geeks.

I agree wholeheartedly that I want a system that boots in less time than it takes me to drum my fingers; I just don’t expect to see any announcements until the bigger players get a change to play around with these new ideas a bit.

On a different note, didn’t Ubuntu and Fedora both move from init to Upstart to speed up boot times by parallelizing the process? Kinda funny that these Intel guys did all their heavy lifting with init…

]]> By: Arpad Borsos http://blog.mozilla.com/tglek/2008/10/09/cool-stuff-in-foreign-realms/comment-page-1/#comment-20083 Arpad Borsos Fri, 10 Oct 2008 09:13:24 +0000 http://blog.mozilla.com/tglek/?p=93#comment-20083 In my opinion the problem is that the distros don't see this as their priority. They even don't have a "no regressions" policy on boot time. As my Boot time regressed by 10 seconds when switching from hardy to intrepid. Unfortunately there is no real "upstream" for boot time. Only half of the work that Arjan did is going into projects shared by the distros. Like sReadahead or kernel work. Distros have to put more resources into removing their bloat. Not paying the price for something you don't need is the first step in the right direction. And well if you want to customize the kernel and everything else on your system, you can still use gentoo. :D But then do not complain if something breaks. In my opinion the problem is that the distros don’t see this as their priority. They even don’t have a “no regressions” policy on boot time. As my Boot time regressed by 10 seconds when switching from hardy to intrepid.
Unfortunately there is no real “upstream” for boot time. Only half of the work that Arjan did is going into projects shared by the distros. Like sReadahead or kernel work.
Distros have to put more resources into removing their bloat. Not paying the price for something you don’t need is the first step in the right direction.

And well if you want to customize the kernel and everything else on your system, you can still use gentoo. :D But then do not complain if something breaks.

]]> By: Preed http://blog.mozilla.com/tglek/2008/10/09/cool-stuff-in-foreign-realms/comment-page-1/#comment-20078 Preed Thu, 09 Oct 2008 21:27:24 +0000 http://blog.mozilla.com/tglek/?p=93#comment-20078 <i>Unfortunately my view of Linux ways is unrealistic.</i> It seems to me that you describe a difference in requirements. Most distros exist to serve a particular market segment (RedHat the enterprise, Ubuntu desktop users), so I imagine you see a chilly reception to the things *you* specifically listed because while it's interesting, it doesn't sell more units of RHEL or get grandmothers to switch away from Win32. In those cases, rock-solid stability and API compatibility, a la RHEL, and a better desktop experience, a la Ubuntu, are what people get paid to work on. Mozilla is no different; for instance, I think the rewriting stuff you're working on is pretty cool, and is going to provide a real benefit in terms of speed and code reduction. So why doesn't Mozilla "bite the bullet" and hold Firefox 3.1 until it's all landed and ready to go? As you ask, "Don't you like to be challenged?" We're not seeing this because because we know there are other requirements that need to be taken into account, and many of those, we may not personally care much about. The beauty of open source is not that projects are "required" to do what others think they should; it's that one has the ability *and* the freedom to do it themselves, and to defend the value of their plan and/or work to the community of developers and users who care about that particular software product. If you tell RedHat you'll buy 2 million units of RHEL if only it booted faster, or promised Ubuntu 2 million users for life, if only it booted faster, their answers might be different. ;-) Unfortunately my view of Linux ways is unrealistic.

It seems to me that you describe a difference in requirements.

Most distros exist to serve a particular market segment (RedHat the enterprise, Ubuntu desktop users), so I imagine you see a chilly reception to the things *you* specifically listed because while it’s interesting, it doesn’t sell more units of RHEL or get grandmothers to switch away from Win32.

In those cases, rock-solid stability and API compatibility, a la RHEL, and a better desktop experience, a la Ubuntu, are what people get paid to work on.

Mozilla is no different; for instance, I think the rewriting stuff you’re working on is pretty cool, and is going to provide a real benefit in terms of speed and code reduction.

So why doesn’t Mozilla “bite the bullet” and hold Firefox 3.1 until it’s all landed and ready to go? As you ask, “Don’t you like to be challenged?”

We’re not seeing this because because we know there are other requirements that need to be taken into account, and many of those, we may not personally care much about.

The beauty of open source is not that projects are “required” to do what others think they should; it’s that one has the ability *and* the freedom to do it themselves, and to defend the value of their plan and/or work to the community of developers and users who care about that particular software product.

If you tell RedHat you’ll buy 2 million units of RHEL if only it booted faster, or promised Ubuntu 2 million users for life, if only it booted faster, their answers might be different. ;-)

]]> By: Anders http://blog.mozilla.com/tglek/2008/10/09/cool-stuff-in-foreign-realms/comment-page-1/#comment-20076 Anders Thu, 09 Oct 2008 19:25:23 +0000 http://blog.mozilla.com/tglek/?p=93#comment-20076 I agree completely. However, in general I think that the distributors do a great job. I am more angry with large corporate FOSS projects that "don't get it". Plugins are great, but insisting that the projects themselves must distribute them as binaries, ranther than the linux distributions, is a disease. Yes, it patches the general suckyness of windows software management. But it also blocks the full benefit of FOSS. I am looking at you: GoogleGears - http://groups.google.com/group/gears-users/browse_thread/thread/74c541a4496964f/5034303542d231a5 OpenOffice - http://lists.go-oo.org/pipermail/dev-go-oo.org/2008-October/000816.html Eclipse - OMG everything... -Anders PS Maybe you'll like Gentoo? I agree completely. However, in general I think that the distributors do a great job. I am more angry with large corporate FOSS projects that “don’t get it”.

Plugins are great, but insisting that the projects themselves must distribute them as binaries, ranther than the linux distributions, is a disease. Yes, it patches the general suckyness of windows software management. But it also blocks the full benefit of FOSS.

I am looking at you:
GoogleGears – http://groups.google.com/group/gears-users/browse_thread/thread/74c541a4496964f/5034303542d231a5
OpenOffice – http://lists.go-oo.org/pipermail/dev-go-oo.org/2008-October/000816.html
Eclipse – OMG everything…

-Anders

PS Maybe you’ll like Gentoo?

]]>