On Sun, 18 Jan 2004, Pedro M. wrote: >> the sources first before committing to such an effort however, >> and make sure you've got a barf bag handy. ;o) > >This is an interesting idea. A sourceforge project would be a good thing >( with Xconfigurator or a new tool). Ah, so you haven't looked at the Xconfigurator source code yet. ;o) >> Nonetheless, this is the open source community, and anyone >> sufficiently motivated enough, can take either utility and >> revitalize them for their own purposes. >> >There is user´s interest about it . See >http://wiki.debian.net/index.cgi?DebianDesktop I see goals of having a friendly useable desktop there, not interest in someone bringing XF86Setup or Xconfigurator back from the dead. ;o) >Now, my screen is black, without Xwindow. Why ??. When I could >see it in Knoppix LiveCD without problems ??. :? Insufficient information to draw any conclusions... ;o) >I think the problem is the monitor frequency, but not sure... I >would like to try to change the frequency ( with a help tool >that automatically restore to the last value if I don´t press >the OK key in a popup window in a certain time after the changes >). > >This is newbie-proof, isn´t it ?? ( one uses it in Windows too ;) Some people or distributions might be interested in following that approach perhaps. That isn't user friendly however. A user friendly system would have monitors that just worked, requiring zero user intervention or configuration. Any monitor metadata needed for configuration would be supplied by the system (or manufacturers) and included with the OS (or driver updates), just as it is done in Windows. You (and others) may want to have a dialog for punching in this information indefinitely perhaps. Personally, I want to see the problem solved on a higher level, and be automatic, thus having users not need to know anything about their monitor other than how to turn it on. >> >Why when i go to Windows the first time, it goes in a low >> >resolution and frequency and no in XFree ??. >> >> Windows starts up generally in 800x600 nowadays (XP). It uses a >> special driver Microsoft supplies called "VGASAVE". > >Good thing to imitate ;) I agree. >> I'm not >> completely at understanding at what this driver is doing >> specifically, however I believe that it is more or less a VESA >> BIOS driver, which can fallback to VGA mode if need be, and might >> possibly have special code in it for certain other hardware. >> It's more or less a "failsafe" driver which is "good enough" in >> order for a proper native driver to be installed for the video >> hardware in the system. > >That´s what I need for my first installation. Later, I can use another >driver... Indeed, that is what I'd like to see in the future. The problem is a complex one however to solve. There is so much video hardware out there, that it is impossible to test it all. All a vendor such as us can do, is make changes to our infrastructure and defaults, release beta OS releases, get feedback from users and try to fix things in time for the final OS release. That isn't good enough though, because what's more likely to happen, is that once the OS is released, 100 times as many people use it, find problems that didn't turn up in beta testing or internal testing, and then they wonder why it doesn't work. So I'd like to see a failsafe driver in the future, but I'm not sure what the best approach is to create one in an evolutionary manner, rather than switchover overnight. Also, such a driver would need to exist for x86, ia64, AMD64, ppc, ppc64 for us, and also Alpha, sparc and other systems to be useful across the board. That complicates things a bit. I'd focus on x86 and AMD64 at first, and add bits to other architectures over time, in order to make the problem sanely approachable. >> I believe it would be possible to fork the "vesa" driver into a >> new driver named "failsafe" or somesuch, which uses VESA VBE by >> default, but gets special hacks put in place for systems that are >> reported to not work, but workarounds can be found, and detected >> by PCI ID and other mechanisms. I'd like to at least experiment >> with this in the future to see if it is a crazy idea or not. > >Go ahed with it. You are taking Linux to a next step forwards. Well, it wouldn't be Linux specific per se. It'd be OS neutral more or less, but it would have some architectural ties likely. I'd like to see some kind of driver emerge that is a superset of the "vesa" driver. I might fiddle with it myself over time as people report problems in the vesa driver to us, although I dont think such hacks belong in the vesa driver itself per se. as that would clutter the existing driver and go against it's purpose IMHO. But a driver based upon it, which gets hacked up over time might suffice. ;o) Take care, TTYL -- Mike A. Harris _______________________________________________ XFree86 mailing list XFree86@xxxxxxxxxxx http://XFree86.Org/mailman/listinfo/xfree86
