Go Back   nV News Forums > Linux Support Forums > NVIDIA Linux

Newegg Daily Deals

Reply
 
Thread Tools
Old 01-07-03, 06:39 PM   #1
Lethal Weapon
Registered User
 
Join Date: Aug 2002
Posts: 14
Send a message via ICQ to Lethal Weapon
Angry nvidia is aware of bugs, but after almost 1 month it's still not fixed?

Nvidia is annoying.. They are aware of some bugs, but don't fix them very fast

Be aware of it before going for nvidia!

----

From: Andy Ritger
To: Lethal Weapon
Cc: linux-bugs
Subject: Re: "loosing the display with keyboard freeze"
Date: 15 Dec 2002 02:50:03 -0500

Thanks for the bug report,

Yes, we're aware of some problems with running multiple X servers,
and hope to have this resolved for a future release.

I certainly hope the next release doesn't take that long...

I apologize for the inconvenience,
- Andy Ritger


On 15 Dec 2002, Lethal Weapon wrote:

> hi,
>
> I use redhat linux 8 + kernel 2.4.21pre1 + nvidia drivers 4191 on a
> gef4ti4200 sparke 128MB vivo (compiled with the src.rpm)
>
>
> The problem is the following:
> i often use X on a crt screen and start a second X :1 -layout tv on tv
> output (using then mplayer to redirect output on tv)
>
> But now i have installed the latest drivers, under some conditions i
> haven't discovered yet, after the startup of the "tv X", i'm used to
> swap back to X with ctrl alt F7, do the mplayer command, and go back to
> the TV X with ctrl alt f8
>
> on that swap, it hangs very often (like 3 times on 5): the screen goes
> black (monitor doesn't go in standbye), tv isn't back initialised, and i
> loose the control of the keyboard (numlock doesn't work). ctrl alt del
> is still catched by init and reboots the computer.
>
>
> Do you want me to try to find a better trigger of this bug? I suppose
> it's the latest driver since i never had this problem with the 3123.
>
> And more important: will i have to wait 5 months again before a new
> driver release fixes this?
>
>
>
>
> I have to say stability improved a lot with 4xxx drivers in opengl
> games, but i've lost the X -swapping-...
>
>
>
> Attached: my XConfig file, use 100Hz on crt, layout -tv for tv output
>
>
>
> greetings,
> Ltwp
>
__________________
Greetz,
Ltwp
Lethal Weapon is offline   Reply With Quote
Old 01-09-03, 09:38 AM   #2
misleb
Registered User
 
Join Date: Dec 2002
Posts: 12
Default

Isn't this how all commercial software is? I mean, short of a critical security issue, you jusually have to wait for the next release of the software or at least a service pack. Are you hoping for 4191sp1?

Perhaps you have been usijng Linux too long and are used to downloading the latests Alan Cox kernel patch when things are not working quite right. You're just spoiled. :P

What *I* would like to see from nVidia is not faster driver updates, but a GPL release. I never really understood why drivers are so secret.

-matthew
misleb is offline   Reply With Quote
Old 01-09-03, 10:24 AM   #3
nutball
International God of Sex
 
Join Date: Jul 2002
Location: en.gb.uk
Posts: 655
Default

Quote:
Originally posted by misleb

What *I* would like to see from nVidia is not faster driver updates, but a GPL release. I never really understood why drivers are so secret.
Because to write good drivers you have to know the gory details about the secret technology on the GPU.

So you either

a) get drivers written by people who don't know anything about the chip, or

b) the GPU manufacturer has to release lots of details about the hardware that they'd really rather keep confidential for commercial reasons

I don't think anyone would want (a), as for (b), well that's intellectual property for you.
__________________
Got nuts?
nutball is offline   Reply With Quote
Old 01-09-03, 10:30 AM   #4
Richard999
Registered User
 
Join Date: Aug 2002
Posts: 22
Default

Why is it that people keep thinking that GPLing the drivers is the answer? What on earth makes you think the open source community is any more capable than software engineers that have direct access to the resources NVIDIA's programmers have?

Explain to me why most hardware with GPL drivers is poorly supported. Don't get me wrong, I'm not unappreciative of the efforts of the guys that do this stuff for free.

It's just that its totally unrealistic to think that if and when NVIDIA opened the source all the problems would magically disappear.
Richard999 is offline   Reply With Quote
Old 01-09-03, 10:52 AM   #5
misleb
Registered User
 
Join Date: Dec 2002
Posts: 12
Default

Quote:
[i]So you either

a) get drivers written by people who don't know anything about the chip, or

b) the GPU manufacturer has to release lots of details about the hardware that they'd really rather keep confidential for commercial reasons

I don't think anyone would want (a), as for (b), well that's intellectual property for you. [/b]
Well, technically, the manufacturer doesn't have to explicitly release any details. Just the source code for the drivers. Sure, you COULD infer the details of the hardware from the source, but how realistic is that? What I want to know is how real is the risk of releasing the the source? Are there any real cases of competitors using driver source code to copy or or clone the hardware itself? Or is it all just a bunch of corporate paranoia?

-matthew
misleb is offline   Reply With Quote
Old 01-09-03, 11:02 AM   #6
nutball
International God of Sex
 
Join Date: Jul 2002
Location: en.gb.uk
Posts: 655
Default

Quote:
Originally posted by misleb
Well, technically, the manufacturer doesn't have to explicitly release any details. Just the source code for the drivers. Sure, you COULD infer the details of the hardware from the source, but how realistic is that? What I want to know is how real is the risk of releasing the the source? Are there any real cases of competitors using driver source code to copy or or clone the hardware itself? Or is it all just a bunch of corporate paranoia?

-matthew
Well I don't know for sure, but I'd guess that a talented driver person could glean quite a lot of information about a GPU architecture from the driver source code. Especially if it's properly commented!

It's probably corporate paranoia too
__________________
Got nuts?
nutball is offline   Reply With Quote
Old 01-09-03, 11:15 AM   #7
misleb
Registered User
 
Join Date: Dec 2002
Posts: 12
Default

Quote:
Originally posted by Richard999
Why is it that people keep thinking that GPLing the drivers is the answer? What on earth makes you think the open source community is any more capable than software engineers that have direct access to the resources NVIDIA's programmers have?
Excuse me? I never said releasing hte source code would necessarily solve any prombles with the drivers themselves. All releasing the source means is that users have the OPTION of fixing it themselves if they want to try rather than being completely dependant on nVidia for binary releases.

Also, the idea is that nVidia would be the main contributers to the driver source. I NEVER said open souce developers could do better than nVidia engineers. Geez.

Quote:
Explain to me why most hardware with GPL drivers is poorly supported. Don't get me wrong, I'm not unappreciative of the efforts of the guys that do this stuff for free.
Again, excuse me? The Linux kernel is full of GPL drivers that are very well supported. As an example of vendor supported GPL drivers, take a look at Intel's e100/e1000 (not eepro100) drivers. Well, I dunno if the Intel drivers are actually GPL, but the source is there. Or IBM's ips (ServeRAID) drivers. I've used both and they work quite well. nVidia could easily be among this list.

Quote:

It's just that its totally unrealistic to think that if and when NVIDIA opened the source all the problems would magically disappear.
Magically? No, it would take work. It would take coordination from nVidia.

-matthew
misleb is offline   Reply With Quote
Old 01-09-03, 06:23 PM   #8
Noth
Registered User
 
Join Date: Nov 2002
Posts: 44
Default

Quote:
Explain to me why most hardware with GPL drivers is poorly supported.
I don't share the same view. My nVidia card is the only piece of hardware I use that doesn't have a really good GPL'd driver.

But this has been beaten over and over. nVidia doesn't own all the code in their drivers, they can't release what they don't own under a totally different license.
Noth is offline   Reply With Quote

Old 01-11-03, 05:17 PM   #9
puterguy
Registered User
 
Join Date: Jan 2003
Posts: 76
Default

Why can't there be more of a compromise, middle-of-the-road position?

It seems that the key objections to source code release are that some portions of the code are owned by 3rd parties and that other portions would reveal hw trade secrets.

Why can't the truly proprietary portions be kept to a minimum and released as binary with a documented API to call it. The rest could then be released as open source.

At least this solution would minimize the portions of code that would not benefit from the help of the open source community.

In a sense, this would be a natural extension of the current state where nVidea release a binary driver file and a kernel source file. I am just saying put as much as possible in the kernel source files.
__________________
Fedora Core6/Linux 2.6.x
nVidea driver: 9631
Asus P4PE motherboard
Intel P4 2.8GHz
1.5GB PC2700 DDR Kingston
Leadtek Ultra A250 GF4 4600 128MB with VIVO/TVO
WinFast TV2000XP Deluxe
Dell 1901FP 19" LCD
Antec 400W Power Supply
puterguy is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


Similar Threads
Thread Thread Starter Forum Replies Last Post
Need Help Installing NVIDIA Tesla M2070Q in Linux RHEL5 Ferianto85 NVIDIA Linux 0 05-18-12 08:35 PM
Rumor regarding lack of 680 availability ViN86 Rumor Mill 6 05-09-12 04:48 PM
Radeon 9700 not all that? sancheuz Other Desktop Graphics Cards 200 10-12-02 09:31 PM

All times are GMT -5. The time now is 10:46 AM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.