Re: [9fans] What a poor VGA support !

Re: [9fans] What a poor VGA support !

[andrey mirtchovski]

>> I will say, though, that Plan 9 would be a lot more widely useful
>> if it contained a "generic" VESA driver in order to support new
>> video cards as soon as they appear.  Xfree86 has one that perhaps
>> could serve as the starting point for a Plan 9 driver.

i had a look at the X86 vesa driver -- at the time it seemed like this
is the only way to get the SiS630 on-board vga going...

waht stopped me from investigating further was the lack of hardware
cursor...  i simply gave up...

Re: [9fans] What a poor VGA support !

[Agnelo de la Crotche]

Digby Tarvin wrote:

> > Sorry folks, but I'm in a particularly pedantic mood this morning ...
> >
> > > 6 DOSez : MS-DOS, PC DOS, Open DOS, DR DOS, PTS DOS, FreeDOS
> >
> > None of these have VGA support. They use the BIOS to write to the
> > screen as if it was a 24x80 tty.
> >
> > > 6 Windozes : 95, 98 , ME, NT, 2K, XP
> >
> > Make that two Windows: 95 and NT. The others are just different
> > releases of 95/NT.
>
> I'd also disqualify any Microsoft operating system from such comparison
> on the basis that hardware support is invariably supplied by the
> hardwave vendor, and no alternative OS vendor can hope to compete
> with that sort of monopoly advantage...
>
> Unix/Linux and other system that use XFree86 has a much larger developer
> base, and so is bound to have better hardware support.
>
> The remaining systems in you list make more interesting comparisons, and
> I would be interested in seeing any 'supported hardware' lists for them.
> How many of them provide source code for the drivers?
>

That OS listing was just an answer to the question 'what Os and how I boot them'.

Speaking of VGA support wasn't right and not really what I meant, allthough most
of them have graphic mode capabilities (directly or indirectly).
But all these OS have ( at least im my own experience ) something in common,
that I didn't find in Plan9 : I was able to install them !
Either they provide a graphic mode installation, which recognized and supported
my hardware or they are console oriented and don't care about the graphic card,
at least during the installation process.
I managed to install Plan9 too, but after having spent a lots of time and finally

broken a monitor. The worse is that before you can figure out  how to edit
Plan9.ini
to boot Plan9 in console mode, you have to get it installed.
I believe that the first choice should be given to a console mode installation.
The
setup could later switch to graphic mode if it detects a supported chipset ( but
remember : the problem with some ATI chipsets is that they are more or less
correctly identified but still unsupported, which has unpredictable (or rather
predictable) effects.

The best example of friendly compatible OS I can think is AtheOS.  I is
intended to be a graphic mode only operating system.  So it runs its
installation in a reasonnable GUI. It provides a S3 driver and tries to
initialize a S3
graphic card.  In fact, it scans the video adapters according to the drivers
present
in the video drivers directory.
I found a Mach64 driver + sources (http://www.kamidake.org). But I didn't need it

for the installation.

Another pleasant  one is Oberon, (http://www.oberon.ethz.ch/native/), which has
VGA,
Vesa and some drivers. Sources for the OS are available ( I don't know for the
drivers, but why not?)

Beos is dead and it's a bummer ! It used to have a pretty wide hardware support.

Amoeba and Hurd have ports of XFree86 I haven't tested yet..

I don't think that the QNX sources are available and you would probably say that
it's just
another Unix.  However hardware support is great.


> For Plan9 it is probably a better use of resources to have well developed
> drivers for products from the few responsible hardware vendors
> that see fit to adequately document the interface to their hardware,
> especially as that usually selects for the better quality hardware
> in any case. So long as this is able to encompass a range that includes
> some examples of high end capabilities, and some that are low cost, so that
> buying Plan9 capable hardware doesn't involve having to compromise too
> much.
>

Are people involved in OS projects in such a concurrence like hardware
manufacturers ?
If not they should establish a new standard, so that hardware products would
simply become usefull and the label "requires Microsoft Windows" would
disappear. It's an utopia, I know.


>
> What we really need is some sort of industry acreditation that can
> be awarded to manufacturers of non-Microsft specific hardware, so
> that people who want a choice can avoid winding up with a lemon
> without needing to carry around a long list of supported hardware
> for every OS they may ever want to run - as well as providing an
> incentive for manufacturers to play the game. I don't mind so much if my
> hardware is not currently supported - it is when it *can't* be
> supported that annoys me...
>

Iike the damned winmodems !

>
> Unix support is often a good sign, but I know of some cases where
> drivers are provided by manufacturers, or under a NDA, such that
> source is not available, or where drivers have had to be reverse
> engineered and developed in spite of the manufacturer...
>
> Unfortunately I suspect the vast majority of the market will have
> already bought hardware 'bundles' before realising there is a problem,
> making intelligent hardware selection no longer an option,
> and as a result will be stuck with Windows...
>
> Regards,
> DigbyT
> --
> Digby R. S. Tarvin                                              digbyt@acm.org
> http://www.cthulhu.dircon.co.uk

Re: [9fans] What a poor VGA support !

[Douglas A. Gwyn]

Lyndon Nerenberg wrote:
> FreeBSD has a compatability layer for the Linux VESA bits, but it's use
> is thin (some games, IIRC). Anyone running a graphical interface is
> running X.

And X needs a framebuffer driver, e.g. VESA.
In fact that's how I'm using an Nvidia card under Solaris.

Re: [9fans] What a poor VGA support !

[Agnelo de la Crotche]

Lyndon Nerenberg wrote:

> >>>>> "Agnelo" == Agnelo de la Crotche <agnelo.geo@yahoo.com> writes:
>
> Sorry folks, but I'm in a particularly pedantic mood this morning ...
>
> > 6 DOSez : MS-DOS, PC DOS, Open DOS, DR DOS, PTS DOS, FreeDOS
>
> None of these have VGA support. They use the BIOS to write to the
> screen as if it was a 24x80 tty.
>

Whatever does the support ,  there are plenty of applications, web
browsers, disk utilities,
graphic shells running  with 16 colors and a reasonnable resolution,
while Plan9
produces only garbage with unsupported/half supported video adapters.


>
> > 6 Windozes : 95, 98 , ME, NT, 2K, XP
>
> Make that two Windows: 95 and NT. The others are just different
> releases of 95/NT.
>

That's true and not true, but  pretty irrelevant here.
If you  insist (but you do not) I can add other OS I've tested but not
installed ,
like Oberon (http://www.oberon.ethz.ch/native/) or MenuetOS
(http://www.menuetos.org/),
providing a GUI which has not broken any monitor yet..


>
> > 3 BSD : FreeBSD, Open BSD, Net BSD
> > 2 Linux : Debian, Suse
> > 2 Unix : SCO, Solaris
>
> As with DOS, none of these operating systems treats the VGA
> device as anything more than a cursor-addressable tty device.
> All the VGA magic is handled by a user-space application: the X
> server.
>

>
> --lyndon

That's true. But at least you can install them and configure the X
server
whenever you like (even if some newest Linux distributions like to
play Windoze and identify everything for you, including video adapter,
but usually they do it well).

At whatever level the VGA magic occures,  there are only two or
three user issues :
- stuff on screen (all OS I've mentionned)
- garbage (Plan9 - speaking for me )
- broken monitor
Do you mind taking these facts into consideration ?



agnelo

Re: [9fans] What a poor VGA support !

[Digby Tarvin]

> Sorry folks, but I'm in a particularly pedantic mood this morning ...
>
> > 6 DOSez : MS-DOS, PC DOS, Open DOS, DR DOS, PTS DOS, FreeDOS
>
> None of these have VGA support. They use the BIOS to write to the
> screen as if it was a 24x80 tty.
>
> > 6 Windozes : 95, 98 , ME, NT, 2K, XP
>
> Make that two Windows: 95 and NT. The others are just different
> releases of 95/NT.

I'd also disqualify any Microsoft operating system from such comparison
on the basis that hardware support is invariably supplied by the
hardwave vendor, and no alternative OS vendor can hope to compete
with that sort of monopoly advantage...

Unix/Linux and other system that use XFree86 has a much larger developer
base, and so is bound to have better hardware support.

The remaining systems in you list make more interesting comparisons, and
I would be interested in seeing any 'supported hardware' lists for them.
How many of them provide source code for the drivers?

For Plan9 it is probably a better use of resources to have well developed
drivers for products from the few responsible hardware vendors
that see fit to adequately document the interface to their hardware,
especially as that usually selects for the better quality hardware
in any case. So long as this is able to encompass a range that includes
some examples of high end capabilities, and some that are low cost, so that
buying Plan9 capable hardware doesn't involve having to compromise too
much.

What we really need is some sort of industry acreditation that can
be awarded to manufacturers of non-Microsft specific hardware, so
that people who want a choice can avoid winding up with a lemon
without needing to carry around a long list of supported hardware
for every OS they may ever want to run - as well as providing an
incentive for manufacturers to play the game. I don't mind so much if my
hardware is not currently supported - it is when it *can't* be
supported that annoys me...

Unix support is often a good sign, but I know of some cases where
drivers are provided by manufacturers, or under a NDA, such that
source is not available, or where drivers have had to be reverse
engineered and developed in spite of the manufacturer...

Unfortunately I suspect the vast majority of the market will have
already bought hardware 'bundles' before realising there is a problem,
making intelligent hardware selection no longer an option,
and as a result will be stuck with Windows...

Regards,
DigbyT
--
Digby R. S. Tarvin                                              digbyt@acm.org
http://www.cthulhu.dircon.co.uk

Re: [9fans] What a poor VGA support !

[Lyndon Nerenberg]

>>>>> "Jason" == Jason Gurtz <jason@tommyk.com> writes:

    Jason> There is the VESA Framebuffer driver in the Linux kernel
    Jason> which will do 16 colors at a reasonable resolution in the
    Jason> console.  Not sure if the BSD's have similar support

FreeBSD has a compatability layer for the Linux VESA bits, but it's use
is thin (some games, IIRC). Anyone running a graphical interface is
running X.

--lyndon

RE: [9fans] What a poor VGA support !

[Jason Gurtz]

> > 3 BSD : FreeBSD, Open BSD, Net BSD
> > 2 Linux : Debian, Suse
...
>
> As with DOS, none of these operating systems treats the VGA
> device as anything more than a cursor-addressable tty device.
> All the VGA magic is handled by a user-space application: the X
> server.

There is the VESA Framebuffer driver in the Linux kernel which will do
16 colors at a reasonable resolution in the console.  Not sure if the
BSD's have similar support

~Jason

--

Re: [9fans] What a poor VGA support !

[Lyndon Nerenberg]

>>>>> "Agnelo" == Agnelo de la Crotche <agnelo.geo@yahoo.com> writes:

Sorry folks, but I'm in a particularly pedantic mood this morning ...

> 6 DOSez : MS-DOS, PC DOS, Open DOS, DR DOS, PTS DOS, FreeDOS

None of these have VGA support. They use the BIOS to write to the
screen as if it was a 24x80 tty.

> 6 Windozes : 95, 98 , ME, NT, 2K, XP

Make that two Windows: 95 and NT. The others are just different
releases of 95/NT.

> 3 BSD : FreeBSD, Open BSD, Net BSD
> 2 Linux : Debian, Suse
> 2 Unix : SCO, Solaris

As with DOS, none of these operating systems treats the VGA
device as anything more than a cursor-addressable tty device.
All the VGA magic is handled by a user-space application: the X
server.

--lyndon

Re: [9fans] What a poor VGA support !

[Russ Cox]

> I will say, though, that Plan 9 would be a lot more widely useful
> if it contained a "generic" VESA driver in order to support new
> video cards as soon as they appear.  Xfree86 has one that perhaps
> could serve as the starting point for a Plan 9 driver.

don't let me stop you!

russ

Re: [9fans] What a poor VGA support !

[Ronald G Minnich]

On Fri, 14 Jun 2002, Douglas A. Gwyn wrote:

> I will say, though, that Plan 9 would be a lot more widely useful
> if it contained a "generic" VESA driver in order to support new
> video cards as soon as they appear.  Xfree86 has one that perhaps
> could serve as the starting point for a Plan 9 driver.


you're right. I probably should not have blown up quite so spectacularly
...

Plan 9 needs a lot of things, I just wish we all had time to do them!

ron

Re: [9fans] What a poor VGA support !

[Agnelo de la Crotche]

Agnelo de la Crotche wrote:

> presotto@plan9.bell-labs.com wrote:
>
> > > Simple, because it is supported by the 28 other
> > > operating systems I managed to install on that machine.
> >
> > My interest is clearly piqued.  What are those 28 OS's and
> > what to you use to boot them all?
> >
>
> As long as you're asking 'how?' and not 'why?' ....
>
> - 6 DOSez : MS-DOS, PC DOS, Open DOS, DR DOS,  PTS DOS, Free DOS
>     - 6 Windozes : 95, 98 , ME, NT, 2K, XP
>     - 3 BSD : Free BSD, Open BSD, Net BSD
>     - 2 Linux : Debian, Suse
>     - 2 Unix : SCO, Solaris
>     - 1 Minix
>     - 8 others : Hurd, BeOS, QNX, OS/2, AtheOS, VsTa, Netware, Amoeba
>     - 1 Plan9
>
> on 2 IBM IDE 40 GB harddisks,  a single AMD K6-2 500 MHz CPU and 512 MB Ram.
>
> Basically I use 3 bootmanagers :
>     - xfdisk ( http://www.mecronome.de/xfdisk/ ) in the MBR of the first HD
>     - the XP bootloader in the bootsector of a primary FAt 16 partition ( <=
> 256 MB )
>     - Grub in a bootsector of some logical partition.
>
>     and some small shell scripts and batchfiles calling utilities (like dd)
> to rewrite
>     the partition tables when necessary.
>
> As a rule, I never use a floppy, emulator or virtual machine. If I cannot
> boot an
> operating system the simple way from the harddisk, I don't install it.
>
> It is not as complicated  as writing video drivers and probably much more
> useless, but
> having spent the last couple years playing around with bootsectors, hard disk
>
> partitions and 'ennemy' operating systems, I would probably have enough stuff
> to
> dedicate a website to the subject, although there are already plenty of
> those, but
> I've never heard of anybody getting as many bootable OS.  The only problem is
>
> that I need at least  3 hours of sleep.
>
> You can easily figure out that the greatest difficulty has been to get the
> widest compatible
> hardware components for all these OS.   Replacing a video card is a great big
> deal
> for me.
>
> >
> > > But , please, tell me, because I'm curious : Why is it so difficult, so
> > > impossible to add some kind of minimum VGA support  ?
> >
> > We used to have minimum support, i.e., 640x480x{1,2,4}.  Those modes are
> > standard and normally grey scale.  You can use a color map with them
> > but 16 or fewer colors doesn't make for great viewing, red letters on
> > a blue backround was my favorite headache producer.
> >
>
> Even under DOS I can get some graphic based applications working
> in a reasonnable resolution. ( which doesn't mean that I use them !)
>
> >
> > We nuked support for the standard modes because
> > a) it was very odd to implement especially after we changed
> >   graphics models.  It doesn't really fit the
> >   model of a frame buffer.  Instead you had to do odd things
> >   to write the different bits to diferent planes.
> > b) it was made clear to us that supporting 640x480x4 grey scale
> >   was so inferior to what people wanted that it wasn't worth the
> >   effort.
> >
> > All other modes are non-standard, i.e., different from card type
> > to card type.  There are a smaller number of VGA chips than cards
> > though even that number is unwieldy.  How the differenty
> > card manufacturers map the chips and plug them together varies.
> > Often the only extra info we have to know about a card is how to
> > recognize it.  PCI is helping there though in the past we had to
> > look at the onboard BIOS and look for patterns; half of /lib/vgadb
> > is just trying to figure out what's there.  If you wc /sys/src/cmd/aux/vga
> > you'll see that the ammount of code per chip type is pretty small
> > (often with an equivalent amount in the kernel to support
> > cursors and such).  However figuring out that code is hard:
> >
>
> I've already tried, read, compared, posted, modified all that ( without
> knowing what
> I was doing ) last year, with the only result of destroying a monitor.  I'm
> not
> complaining about that, I was aware of the risk.
>
> >
> > - documentation is usually poor or non-existant.  Drivers from
> >  more popular OS's (or X) are often the only documentation
> > - Some registers bits lock out other registers; just reading the
> >  settings used by windows and rewriting them back is usually
> >  impossible.
> > - setup requires writing values into MANY registers, most of
> >  which are different from chip to chip.  The values are dependent
> >  on clock speed, bit depth, screen size, and chip version.  Just
> >  getting them working on one machine doesn't ensure they'll work
> >  on another.  In some cases it takes futzing with a number of
> >  machines and board revs to get a `stable' implementation.
>
> What I found particularly annoying and frustrating is that it almost works.
> Almost !  I have already discussed that issue and probalby other people
> did as well.  You know : the problem with the overlapped screen !
> The chipset is identified,  because different ATI Xpert use the same
> chipset and so have the same BIOS signature. I was told that they run at
> different clock speed.
> Even if what I got to  see, about 2/3 of  windows is sufficient for the
> installation,
> you cannot think of starting anything  with it.
>
> >
> > Our most popular laptop is the T23 (because of a company deal with
> > IBM) and we set it up by being careful not to touch too many registers
> > since we have no idea what the LCD extension registers do.
> >
> > In our group, we have one person, jmk, who sepnds a considerable amount
> > of time just getting the cards we have working.  However, that's
> > in addition to what he's supposed to be doing, not what he's paid
> > for.  Some of the community is also nice enough to send us the /lib/vgadb
> > info + drivers to get other cards working.
>
> So did I.
>
> >  However, because the
> > Plan 9 distribution is a 'free time' effort at the Labs we can't
> > aim one person to doing full time VGA support.  If the community
> > were larger, we'ld probably get enough contributions to cover a
> > larger percentage of the cards or if Plan 9 were actually popular
> > we might be able to get manufacturers to write drivers.
> >
> > It would be useful if the manufacturers could get together and
> > come up with another VGA standard so that we would waste less
> > time and angst.  However, there's no real motivation for this.
> > To first order, there's only one OS for PC's, and to second order
> > only two.  It's not that hard for a manufacturer to put together
> > drivers for their new board for 2 OS's.
>
> Obviously if they were humanists, they won't be manufacturers.
>
> >  I'm pretty impressed that
> > there are 26 others that manage to support all cards.
>
> As I said Plan9 is the 'only' OS which doesn't work like it should.
> I've been coming so far with that operating system to write a
> bootmenu and set the default to 'no vga'. So at least, it boots
> and opens a console and I can count it as the 29th.
>
> >
>
> > Besides, if you ever got your ATI Xpert (Rage Pro) working,
> > we'ld have to go through this conversation again explaining why
> > we don't have a web browser worth using.
>
> Since I (almost) never get up, I keep trying regurarly ...

What a funny mistake ! You would have read 'give up' instead.

>
>
> agnelo.

Re: [9fans] What a poor VGA support !

[Agnelo de la Crotche]

"Fco.J.Ballesteros" wrote:

> I think we all may agree with you, but putting Ronald message in a
> different format, I'd say that it would be better if we all tried to
> add the missing pieces or help to add them instead of complaining
> about them not being there.  In that sense, I couldn't agree more with
> Ronald.
>
>   ------------------------------------------------------------------------
>

Plus : in one's native language, they may be hundreds of manners of expressing
how you feel about something which doesn't work, without giving others the
impression that you're complaining.  In my poor english I found only one.
Sorry about that.
Agnelo




>
> Subject: Re: [9fans] What a poor VGA support !
> Date: Fri, 14 Jun 2002 08:45:43 GMT
> From: "Douglas A. Gwyn" <DAGwyn@null.net>
> Reply-To: 9fans@cse.psu.edu
> Organization: U.S. Army Research Laboratory
> To: 9fans@cse.psu.edu
> References: <3D079A78.895ACEAE@yahoo.com>, <Pine.LNX.4.33.0206130816370.24315-100000@snaresland.acl.lanl.gov>
>
> Ronald G Minnich wrote:
> > I hate it when people complain like this about stuff that costs them
> > nothing.
>
> I will say, though, that Plan 9 would be a lot more widely useful
> if it contained a "generic" VESA driver in order to support new
> video cards as soon as they appear.  Xfree86 has one that perhaps
> could serve as the starting point for a Plan 9 driver.

Re: [9fans] What a poor VGA support !

[Agnelo de la Crotche]

presotto@plan9.bell-labs.com wrote:

> > Simple, because it is supported by the 28 other
> > operating systems I managed to install on that machine.
>
> My interest is clearly piqued.  What are those 28 OS's and
> what to you use to boot them all?
>

As long as you're asking 'how?' and not 'why?' ....

- 6 DOSez : MS-DOS, PC DOS, Open DOS, DR DOS,  PTS DOS, Free DOS
    - 6 Windozes : 95, 98 , ME, NT, 2K, XP
    - 3 BSD : Free BSD, Open BSD, Net BSD
    - 2 Linux : Debian, Suse
    - 2 Unix : SCO, Solaris
    - 1 Minix
    - 8 others : Hurd, BeOS, QNX, OS/2, AtheOS, VsTa, Netware, Amoeba
    - 1 Plan9

on 2 IBM IDE 40 GB harddisks,  a single AMD K6-2 500 MHz CPU and 512 MB Ram.

Basically I use 3 bootmanagers :
    - xfdisk ( http://www.mecronome.de/xfdisk/ ) in the MBR of the first HD
    - the XP bootloader in the bootsector of a primary FAt 16 partition ( <=
256 MB )
    - Grub in a bootsector of some logical partition.

    and some small shell scripts and batchfiles calling utilities (like dd)
to rewrite
    the partition tables when necessary.

As a rule, I never use a floppy, emulator or virtual machine. If I cannot
boot an
operating system the simple way from the harddisk, I don't install it.

It is not as complicated  as writing video drivers and probably much more
useless, but
having spent the last couple years playing around with bootsectors, hard disk

partitions and 'ennemy' operating systems, I would probably have enough stuff
to
dedicate a website to the subject, although there are already plenty of
those, but
I've never heard of anybody getting as many bootable OS.  The only problem is

that I need at least  3 hours of sleep.

You can easily figure out that the greatest difficulty has been to get the
widest compatible
hardware components for all these OS.   Replacing a video card is a great big
deal
for me.


>
> > But , please, tell me, because I'm curious : Why is it so difficult, so
> > impossible to add some kind of minimum VGA support  ?
>
> We used to have minimum support, i.e., 640x480x{1,2,4}.  Those modes are
> standard and normally grey scale.  You can use a color map with them
> but 16 or fewer colors doesn't make for great viewing, red letters on
> a blue backround was my favorite headache producer.
>

Even under DOS I can get some graphic based applications working
in a reasonnable resolution. ( which doesn't mean that I use them !)

>
> We nuked support for the standard modes because
> a) it was very odd to implement especially after we changed
>   graphics models.  It doesn't really fit the
>   model of a frame buffer.  Instead you had to do odd things
>   to write the different bits to diferent planes.
> b) it was made clear to us that supporting 640x480x4 grey scale
>   was so inferior to what people wanted that it wasn't worth the
>   effort.
>
> All other modes are non-standard, i.e., different from card type
> to card type.  There are a smaller number of VGA chips than cards
> though even that number is unwieldy.  How the differenty
> card manufacturers map the chips and plug them together varies.
> Often the only extra info we have to know about a card is how to
> recognize it.  PCI is helping there though in the past we had to
> look at the onboard BIOS and look for patterns; half of /lib/vgadb
> is just trying to figure out what's there.  If you wc /sys/src/cmd/aux/vga
> you'll see that the ammount of code per chip type is pretty small
> (often with an equivalent amount in the kernel to support
> cursors and such).  However figuring out that code is hard:
>

I've already tried, read, compared, posted, modified all that ( without
knowing what
I was doing ) last year, with the only result of destroying a monitor.  I'm
not
complaining about that, I was aware of the risk.


>
> - documentation is usually poor or non-existant.  Drivers from
>  more popular OS's (or X) are often the only documentation
> - Some registers bits lock out other registers; just reading the
>  settings used by windows and rewriting them back is usually
>  impossible.
> - setup requires writing values into MANY registers, most of
>  which are different from chip to chip.  The values are dependent
>  on clock speed, bit depth, screen size, and chip version.  Just
>  getting them working on one machine doesn't ensure they'll work
>  on another.  In some cases it takes futzing with a number of
>  machines and board revs to get a `stable' implementation.

What I found particularly annoying and frustrating is that it almost works.
Almost !  I have already discussed that issue and probalby other people
did as well.  You know : the problem with the overlapped screen !
The chipset is identified,  because different ATI Xpert use the same
chipset and so have the same BIOS signature. I was told that they run at
different clock speed.
Even if what I got to  see, about 2/3 of  windows is sufficient for the
installation,
you cannot think of starting anything  with it.


>
> Our most popular laptop is the T23 (because of a company deal with
> IBM) and we set it up by being careful not to touch too many registers
> since we have no idea what the LCD extension registers do.
>
> In our group, we have one person, jmk, who sepnds a considerable amount
> of time just getting the cards we have working.  However, that's
> in addition to what he's supposed to be doing, not what he's paid
> for.  Some of the community is also nice enough to send us the /lib/vgadb
> info + drivers to get other cards working.

So did I.


>  However, because the
> Plan 9 distribution is a 'free time' effort at the Labs we can't
> aim one person to doing full time VGA support.  If the community
> were larger, we'ld probably get enough contributions to cover a
> larger percentage of the cards or if Plan 9 were actually popular
> we might be able to get manufacturers to write drivers.
>
> It would be useful if the manufacturers could get together and
> come up with another VGA standard so that we would waste less
> time and angst.  However, there's no real motivation for this.
> To first order, there's only one OS for PC's, and to second order
> only two.  It's not that hard for a manufacturer to put together
> drivers for their new board for 2 OS's.

Obviously if they were humanists, they won't be manufacturers.


>  I'm pretty impressed that
> there are 26 others that manage to support all cards.

As I said Plan9 is the 'only' OS which doesn't work like it should.
I've been coming so far with that operating system to write a
bootmenu and set the default to 'no vga'. So at least, it boots
and opens a console and I can count it as the 29th.

>

> Besides, if you ever got your ATI Xpert (Rage Pro) working,
> we'ld have to go through this conversation again explaining why
> we don't have a web browser worth using.

Since I (almost) never get up, I keep trying regurarly ...

agnelo.

Re: [9fans] What a poor VGA support !

[Fco.J.Ballesteros]

[-- Attachment #1: Type: text/plain, Size: 290 bytes --]

I think we all may agree with you, but putting Ronald message in a
different format, I'd say that it would be better if we all tried to
add the missing pieces or help to add them instead of complaining
about them not being there.  In that sense, I couldn't agree more with
Ronald.



[-- Attachment #2: Type: message/rfc822, Size: 1839 bytes --]

From: "Douglas A. Gwyn" <DAGwyn@null.net>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] What a poor VGA support !
Date: Fri, 14 Jun 2002 08:45:43 GMT
Message-ID: <3D08DC19.9731EF3E@null.net>

Ronald G Minnich wrote:
> I hate it when people complain like this about stuff that costs them
> nothing.

I will say, though, that Plan 9 would be a lot more widely useful
if it contained a "generic" VESA driver in order to support new
video cards as soon as they appear.  Xfree86 has one that perhaps
could serve as the starting point for a Plan 9 driver.

Re: [9fans] What a poor VGA support !

[Agnelo de la Crotche]

Ronald G Minnich wrote:

> On Thu, 13 Jun 2002, Agnelo de la Crotche wrote:
>
> > But , please, tell me, because I'm curious : Why is it so difficult, so
> > impossible to add some kind of minimum VGA support  ?
>
> because vga cards are a bitch, and pc compatibility is a bitch, and
> because PC hardware sucks, and the vendors suck, and you can't get
> information on the cards, and overall .... graphics cards are a major pain
> the ass. I can state this with certainty as I'm dealing with it every day
> in LinuxBIOS. Most of the people who design PC hardware should be shot.
>
> > I 'm not a video  expert  (definitely not) but as much as I can see when
> > booting Plan9  with video auto detection
>
> look, the specs are out there just as much for you as anyone else. If you
> don't have support, and you want support, write it. You can't write it
> because you are not capable, don't complain here.
>
> I hate it when people complain like this about stuff that costs them
> nothing.
>

As well as I do when people answer in newsgroups to say nothing !
The fact that I'm not capable of doing something doesn't mean that I don't
have the right to complain.
Why not ?  Following this logic leads to a nonsense ( or a dictature ).

I'm not asking if it's moral or allowed to complain about  stuff which costs
nothing or whatever.
And I wasn't comparing myself with You guys (?)  but simply an operating
system with other operating
systems (and of course it implied their designers).

While telling that PC hardware sucks, and vga cards sucks, and vendors suck,
you do nothing but complain,
and you do worse because you say nothing new.  (And you do not take any risk,
because everybody must agree)
You may or may not hate me. I will continue to complain whenever it is
necessary.
(In what world do you think you would live if people had never complained ?)

sincerely,

agnelo.

Re: [9fans] What a poor VGA support !

[Douglas A. Gwyn]

Ronald G Minnich wrote:
> I hate it when people complain like this about stuff that costs them
> nothing.

I will say, though, that Plan 9 would be a lot more widely useful
if it contained a "generic" VESA driver in order to support new
video cards as soon as they appear.  Xfree86 has one that perhaps
could serve as the starting point for a Plan 9 driver.

Re: [9fans] What a poor VGA support !

[Dan Cross]

> > But , please, tell me, because I'm curious : Why is it so difficult, so
> > impossible to add some kind of minimum VGA support  ?
>
> because vga cards are a bitch, and pc compatibility is a bitch, and
> because PC hardware sucks, and the vendors suck, and you can't get
> information on the cards, and overall .... graphics cards are a major pain
> the ass. I can state this with certainty as I'm dealing with it every day
> in LinuxBIOS. Most of the people who design PC hardware should be shot.

Don't hold back, Ron, tell us how you *really* feel....

	- Dan C.

:-)

Re: [9fans] What a poor VGA support !

[plan9]

On Thu, Jun 13, 2002 at 02:16:15PM +0100, forsyth@vitanuova.com wrote:
>
> there's an amusing essay somewhere by the person at Be who
> had to write the card support.

http://david.weekly.org/writings/be-adventure.php3

Re: [9fans] What a poor VGA support !

[Russ Cox]

> Our most popular laptop is the T23 (because of a company deal with
> IBM) and we set it up by being careful not to touch too many registers
> since we have no idea what the LCD extension registers do.

And the initial T23 support got written by guessing that
the card was like the other Savage-based cards and then
playing with one register at a time to figure out the rest.

Re: [9fans] What a poor VGA support !

[presotto]

> Simple, because it is supported by the 28 other
> operating systems I managed to install on that machine.

My interest is clearly piqued.  What are those 28 OS's and
what to you use to boot them all?

> But , please, tell me, because I'm curious : Why is it so difficult, so
> impossible to add some kind of minimum VGA support  ?

We used to have minimum support, i.e., 640x480x{1,2,4}.  Those modes are
standard and normally grey scale.  You can use a color map with them
but 16 or fewer colors doesn't make for great viewing, red letters on
a blue backround was my favorite headache producer.

We nuked support for the standard modes because
a) it was very odd to implement especially after we changed
  graphics models.  It doesn't really fit the
  model of a frame buffer.  Instead you had to do odd things
  to write the different bits to diferent planes.
b) it was made clear to us that supporting 640x480x4 grey scale
  was so inferior to what people wanted that it wasn't worth the
  effort.

All other modes are non-standard, i.e., different from card type
to card type.  There are a smaller number of VGA chips than cards
though even that number is unwieldy.  How the differenty
card manufacturers map the chips and plug them together varies.
Often the only extra info we have to know about a card is how to
recognize it.  PCI is helping there though in the past we had to
look at the onboard BIOS and look for patterns; half of /lib/vgadb
is just trying to figure out what's there.  If you wc /sys/src/cmd/aux/vga
you'll see that the ammount of code per chip type is pretty small
(often with an equivalent amount in the kernel to support
cursors and such).  However figuring out that code is hard:

- documentation is usually poor or non-existant.  Drivers from
 more popular OS's (or X) are often the only documentation
- Some registers bits lock out other registers; just reading the
 settings used by windows and rewriting them back is usually
 impossible.
- setup requires writing values into MANY registers, most of
 which are different from chip to chip.  The values are dependent
 on clock speed, bit depth, screen size, and chip version.  Just
 getting them working on one machine doesn't ensure they'll work
 on another.  In some cases it takes futzing with a number of
 machines and board revs to get a `stable' implementation.

Our most popular laptop is the T23 (because of a company deal with
IBM) and we set it up by being careful not to touch too many registers
since we have no idea what the LCD extension registers do.

In our group, we have one person, jmk, who sepnds a considerable amount
of time just getting the cards we have working.  However, that's
in addition to what he's supposed to be doing, not what he's paid
for.  Some of the community is also nice enough to send us the /lib/vgadb
info + drivers to get other cards working.  However, because the
Plan 9 distribution is a 'free time' effort at the Labs we can't
aim one person to doing full time VGA support.  If the community
were larger, we'ld probably get enough contributions to cover a
larger percentage of the cards or if Plan 9 were actually popular
we might be able to get manufacturers to write drivers.

It would be useful if the manufacturers could get together and
come up with another VGA standard so that we would waste less
time and angst.  However, there's no real motivation for this.
To first order, there's only one OS for PC's, and to second order
only two.  It's not that hard for a manufacturer to put together
drivers for their new board for 2 OS's.  I'm pretty impressed that
there are 26 others that manage to support all cards.

Besides, if you ever got your ATI Xpert (Rage Pro) working,
we'ld have to go through this conversation again explaining why
we don't have a web browser worth using.

Re: [9fans] What a poor VGA support !

[Ronald G Minnich]

On Thu, 13 Jun 2002, Agnelo de la Crotche wrote:

> But , please, tell me, because I'm curious : Why is it so difficult, so
> impossible to add some kind of minimum VGA support  ?

because vga cards are a bitch, and pc compatibility is a bitch, and
because PC hardware sucks, and the vendors suck, and you can't get
information on the cards, and overall .... graphics cards are a major pain
the ass. I can state this with certainty as I'm dealing with it every day
in LinuxBIOS. Most of the people who design PC hardware should be shot.

> I 'm not a video  expert  (definitely not) but as much as I can see when
> booting Plan9  with video auto detection

look, the specs are out there just as much for you as anyone else. If you
don't have support, and you want support, write it. You can't write it
because you are not capable, don't complain here.

I hate it when people complain like this about stuff that costs them
nothing.

ron

Re: [9fans] What a poor VGA support !

[forsyth]

>>But , please, tell me, because I'm curious : Why is it so difficult, so
>>impossible to add some kind of minimum VGA support  ?

because the only bit that is standard at the VGA register level
doesn't give you much, if you'd like some colour
(even 640x480x8 uses non-standard extensions).

there is a VESA bios interface that could be used to provide a card-independent
way of implementing a flat map but even several years ago
many cards didn't implement the 32-bit interface, and it could
only be activated from real mode.  things might have changed.
roger peppe had a version of (native) Inferno that tried the
real mode call (it wasn't easy), with some card-specific code as well,
and that worked reasonably well.  we've made that code available.

there's an amusing essay somewhere by the person at Be who
had to write the card support.   it wasn't straightforward.
how did they manage?  they had money (at the time) to throw at it,
at least one person, a fair number of machines and lots of cards
as samples.

[9fans] What a poor VGA support !

[Agnelo de la Crotche]

I don't know very much about Plan9. All I can tell is that it has the
poorest ( and highlest dangerous) VGA support I have never seen for an
operating system. Now somebody is going to ask ( or maybe even not )
: What kind of video adapter do you have ? Have you try to change this
or that in plan9.ini ? Do you see a red screen ?  ... as I was already
asked one year ago, and two years ago.
OK ! It's a bummer. I was starting to hope that something would have
changed with that last release. But you (?) definitely do not want to
support my video card.
But , please, tell me, because I'm curious : Why is it so difficult, so
impossible to add some kind of minimum VGA support  ?
Guess,  I have an ATI Xpert (Rage Pro). Why that one and why do I not
intend to replace it ? Simple, because it is supported by the 28 other
operating systems I managed to install on that machine. Nobody but Plan9
has trouble (and makes trouble ! Don't mention my old monitor ! ) with
it.
I 'm not a video  expert  (definitely not) but as much as I can see when
booting Plan9  with video auto detection (  At least,  I can boot in
console mode but to do what ? ),  something  is wrong or missing in the
Mach64 driver, which all other OS (i.e BeOS, QNX, AtheOS)  have solved.
Is there "definitely" no hope of a solution ?

agnelo