[Xpert]Re: Re: Is the XFree development stuck in a dead end?
On Tue, 16 Jul 2002, Andrew P. Lentvorski wrote: >Date: Tue, 16 Jul 2002 10:52:22 -0700 (PDT) >From: Andrew P. Lentvorski <[EMAIL PROTECTED]> >To: [EMAIL PROTECTED] >Content-Type: TEXT/PLAIN; charset=X-UNKNOWN >List-Id: General X Discussion >Subject: Re: Re: Is the XFree development stuck in a dead end? > >On Tue, 16 Jul 2002, [iso-8859-1] José Fonseca wrote: > >> My interest in the XFree86 development is infact rather specific: the >> RandR extension will be a crucial brick for a proper 3D support on lower >> end cards such as Mach64 which have little onboard memory. > >My question to you is: why do you want to spend your time on this when you >can buy a *much* higher end card than a Mach for <$40 US? Your time is >worth than that, even if you are a student. > >One of the primary time-wasters in open-source is that developers spend a >lot of time on support for hardware which is obsolete. When obsolete >hardware was much cheaper than brand new hardware, this support for was >important. Now, however, that difference is pretty minimal. Buy a new >card and save yourself lots of pain. The primary motivating factor behind such an effort is laptops. You can't generally replace the video hardware in /most/ laptops, and as such, if you have a Mach64 based laptop, and no reason to shell out $2000-3000 because the machine is suitable still for what you require, then having 3D work on it may be worth the effort. -- Mike A. Harris Shipping/mailing address: OS Systems Engineer 190 Pittsburgh Ave., Sault Ste. Marie, XFree86 maintainer Ontario, Canada, P6C 5B3 Red Hat Inc. http://www.redhat.com ftp://people.redhat.com/mharris ___ Xpert mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xpert
[Xpert]Re: Re: Is the XFree development stuck in a dead end?
On Tue, 16 Jul 2002, Geoffrey wrote: >Date: Tue, 16 Jul 2002 09:29:02 -0400 >From: Geoffrey <[EMAIL PROTECTED]> >To: [EMAIL PROTECTED] >Content-Type: text/plain; charset=us-ascii; format=flowed >List-Id: General X Discussion >Subject: Re: Re: Is the XFree development stuck in a dead end? > >Mike A. Harris wrote: >> On Mon, 15 Jul 2002, Christian Berger wrote: >> >> Netscape is much faster than Mozilla. I think it's just that some design decisions in the X version of Mozilla, which is probably much different than the Window's version, are suboptimal. Having seen enough >>>Well I doubt Mozilla for Windows is faster than Mozilla for Linux. >>> >> >> You've not tried it then. Mozilla for Windows running on my box >> on one processor runs faster than Mozilla in Linux running on >> both processors. (Win98SE). Application startup time is faster >> for Mozilla in Windows, as is runtime execution. Not measured or >> benchmarked mind you. It is visibly noticeable. > >I'm really hoping this thread dies soon, but I've got to chime in here. > There is absolutely no comparison between the windows gui and X. I >routinely have 10-15 windows open, which would include at least 2 >mozilla mail windows and minimally 3-4 browser windows. I also have 2 >desktops with 3 virtual desktops each. You just don't get that kind of >functionality with a Windows gui and if you tried to have that many >windows open on win95, win98, nt, win2k, it would purely meltdown. I >have no experience with winxp and don't plan to, but I suspect it's no >different. The above argument appears to revolve around Mozilla, not X. I think perhaps you've misinterpreted what I've said. My above claims did not really have anything at all to do with comparing X to Windows. What I was saying was that Mozilla is slower in Linux than it is in Windows, and that the reason for that is a Mozilla issue, not an X issue or a Windows issue. Mozilla shares code between Windows and X11, however the Windows specific Mozilla code is more optimized than the X specific Mozilla code. Basically, the fact Mozilla is slower in Linux/X has nothing to do with X, and is no indicator that X is slower than Windows. The entire Mozilla discussion is a sidetrack from the $topic actually. I just wanted to set the record straight that Mozilla *is* faster in Windows contrary to the given hypothesis that it would not be faster in Windows. I mostly use Mozilla in Linux, however any time I get stuck in a Windows environment and need a browser, that browser is Mozilla, and since I use it all the time, seeing it start up and run several times faster than I'm used to in Linux, is a noticeable thing. I've never assumed that this was due to Windows being faster. IMHO, any slow GUI code running in X, is more likely to be a poorly written *application* or an unoptimized app, or unoptimized toolkit or similar. In the end however, only true profiling can be the judge in any particular case of a slow app. -- Mike A. Harris Shipping/mailing address: OS Systems Engineer 190 Pittsburgh Ave., Sault Ste. Marie, XFree86 maintainer Ontario, Canada, P6C 5B3 Red Hat Inc. http://www.redhat.com ftp://people.redhat.com/mharris ___ Xpert mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xpert
[Xpert]Re: Re: Is the XFree development stuck in a dead end?
On Tue, 16 Jul 2002 [EMAIL PROTECTED] wrote: >> >Is that a joke ? Did you ever try to set up a second gfx card and >> >monitor under Mac OS ? It's a breeze, just point'n'click. Whereas in >> X, >> >you have to hunt for the Xinerama HOWTO and mess with the config >> file. >> >> xf86cfg has multihead configuration built in, although it isn't >> what I personally consider user friendly. This is something that >> will become more friendly in the future though as multihead >> becomes much more popular. > > :-) > > I wan't to stop someday and write code for a better multihead configuration >interface in the textmode (currently it just adds new screens to the left of >the last one). For the graphics interface, I plan to write a wizard mode, >similar to the text interface. And also, make this wizard mode allow configuring >everything without the need of a working mouse; curently if the mouse does >not work, it is required to use xkb mousekeys, and this really is not user >friendly :-) > > Anyway, the code is there, and I don't mind if someone uses part (or all) of >the xf86cfg code in a Gtk or Kde interface. I just think that any such code >should use libxf86config, so that if it reads an existing XF86Config file, >when writing a new one, does not miss any information from the previous >configuration file (in the current libxf86config, comments may be rewritten out >of order, but are not lost). Our current tool under development is using libxf86config exactly for the reasons you suggest. ;o) It made no sense to reinvent the wheel. Those wishing to try the tool out, and comment about it, please do! redhat-config-xfree86 is the name of the tool/package and it is currently in rawhide. It is written in python/C and uses GTK+. All feedback and suggestions would be greatly appreciated. I'd like to see the tool enhanced over time to be a very user friendly tool for both GNOME and KDE users, and potentially users not using GNOME or KDE as well. Any requests for enhancement, etc. should be filed in bugzilla preferably also. Thanks for libxf86config! -- Mike A. Harris Shipping/mailing address: OS Systems Engineer 190 Pittsburgh Ave., Sault Ste. Marie, XFree86 maintainer Ontario, Canada, P6C 5B3 Red Hat Inc. http://www.redhat.com ftp://people.redhat.com/mharris ___ Xpert mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xpert
