Haig-ACPI driver question

[Brent]

I see that the new driver 5.20 is MS and PC 99 certified for the G400 and G200. Will it support S3 suspend to RAM ACPI state for the G200. My G400 with 5.15 and 5.25 supports S3 but my Mystique G200 with 4 series driver resumes to a black screen and with 5.15 or 5.25 resumes to a garbled and frozen display.
Before I reformat maybe you could advise?

[Brent]

Responses-anyone?

[Haig]

Hi Brent,

We only support D0 and D3 modes. I don't know if S3 is part of these 2 modes.

If it is, let me know and I'll have it reproduced here and enter it into the bug data base.

Haig

[Brent]

S3 is the Suspend To Ram state as specified by MS for Win98/SE for a "/p j" ACPI install and enabled by Asus through the BIOS on the P3bf under a Win98/SE ACPI install. I am not concerned at this stage about S4 which is suspend to disk as it will only be officially supported by Win 2000.
I am sure you know about all this so forgive me for my explanation.
I have tried 3 different ACPI compliant mobo's with the G200 and G400 and as first mentioned in my post, the G200 does not recover from resume. Some mobo web sites have test reports on card compatibility and the G200 always fails the ACPI tests.
This is also a common topic on mobo forums with users trying to implement ACPI with a G200.
I wonder if it is a driver issue and what will happen with Win 2000 because I believe that if Win 2000 detects an ACPI compliant mobo (as almost all mobos now are) , it will BY DEFAULT install an ACPI BIOS and not APM BIOS as Win 98 currently does. Surely this will mean that G200 users will have a frozen system every time the computer goes into standby?

[Haig]

Hi Brent,

I will get a P3B-F tomorrow and let you knwo what the outcome is.

PS: Is it me or is something up with the submit? This is the 4th time I answer this thread and it hasn't registered eventhough I get the "thank you for posting....."

Haig

[This message has been edited by Haig (edited 10-01-1999).]

[kjliew]

Brent,

Can you please tell me which sites have tested and posted results for various motherboards ACPI compliance. When I first got my G200, I also formated and reinstall Win98 using /p j to enable ACPI on my new mobo. And as you said, the system never wake up again from suspend/standby. At that time, I didn't even suspect that it's G200 that causes the problem. My guess would be mobo BIOS that is not fully ACPI compliant.

So in the end, I format again and reinstall Win98 for APM BIOS.... too bad.

[Brent]

The Aopen site is quite good with the "test reports" you can click on for the motherboards. Try clicking on a newish board like the AX63Pro and then in the side menu click on "test reports" and then on the section that tests vga cards and sound cards for compatibility with resume from suspend.
If you also hang out at your mobos forum for long enough, you will see a pattern of users success or lack thereof with ACPI and various video cards/ sound cards. BTW, the SB Live has been the equivalent of the G200 as far as sound cards and problems with ACPI goes but after much bitching by users, a few beta bioses later and the latest SB drivers, I got ACPI and my G400/ SB Live working perfectly.

[kjliew]

Brent,

I have SBLive/G200 combination. Can you tell me which Powerdesk version and which LiveWare version you're using that do not have problems in ACPI. I'll soon reformat my computer again, for some experiments purposes. Hopefully this time, I get ACPI work out flawlessly. Do you use the latest LiveWare 2.1 drivers? I'm think of purchasing the LiveWare 3 CD when it is available. I'm still stuck with LiveWare 1.0 which comes in the CD when I got my SBLive.

Thanks.

[Brent]

If you have a P3bf, first make sure you have at least beta bios 1003.006 or later.
With the G400, PD 5.15 (for RRG only and 5.25 were fine).
For the SB Live, I recommend you purchase the Liveware 2 CD or 3 (no driver diff just new software). There is a problem using the web Liveware 2/ 2.1 upgrade over Liveware 1. After running 2 off the CD, run the 2.1 patch.
Win 98 SE also has new ACPI support over 98 and once again an upgrade might not work perfectly.
Your BIOS settings are also an important factor.
The whole ACPI thing is very flaky and the way I see it is that because 99% of people will run a standard install and as long as their system swithes the monitor of etc., they will not look any further so hardware and software suppliers are delaying implementation (probably due to costs) until Win 2000 makes it a necessity. Only notebooks are working right and there you are looking at handpicked components and battery life as a major selling point (along with a price). I just believe that if most people applied the average attitude of switching a light or other appliance off when not using it to their computers and demanded a fix, something would be done now about it. Also, how many people would put up with a TV that took over a minute to come to life instead of seconds. Would they accept that the way for a quick switch on was a standby mode that used as much electricity as a 25W light bulb and also kept something running in the TV that made a noise and sucked dust in 24 hours a day? ACPI is the greatest advance towards making a PC like a white goods in operation for the unitiated masses.
Under APM for years, I left my system running all the time and put up with the downsides. I wished there was a solution at the time and now nearly 18 months later after the "solution" was released, it still hardly works. I wonder where the "greenies" are and whether they have worked out the cost to the planet of this wastage in not just non renewable resources but also money and time.
So endeth my rant and to many of you calling me a total whacker, horses for courses and keep your minds open.

[kjliew]

Thanks a lot Brent.

Power management is very important to me. I use my computer very often, more then 6 hours per day. And, sometimes, it is just left turning ON for the whole night. I have been using APM for a long time. The new ACPI standards offers much better features than APM, but it's too lame that it takes so long to become matured. The first chipsets, Intel 430TX which claims to support ACPI, was simply a joke, none of the Intel430TX BIOS implemented ACPI. Being able to suspend to disk is another features that I would love very much. Well, that's not to save a few penny on my elec bill, but just doing something for the mother earth.

It's too shamed for Microsoft that they totally ignored Power Management on Windows NT, until Windows 2000. I wish they could have implemented at least APM support on Windows NT.

[Brent]

Thanks Kjiliew,
I think that the general level of support and integrity in the computer industry must be the lowest of any sector. Even in the second hand car industry which is the butt of endless jokes, you can get warranties with some substance and successfully sue the dealer if you are sold something that does not deliver what was promised. In the retail consumer industry that I have a lot of experience in (electronic appliances other than computers), the customer could come back weeks later even with a well used product and successfully argue that they have bought a lemon or been sold the wrong thing for their needs. With computers, the buck is always passed or the consumer is told they are at fault due to user error. Very often this is true but it is because in the first place the consumer could not get a straight answer that covered ALL the bases. The poor salesperson even if honest can not rely on the manufacturers and publishers of the products to support him to satisfy the customer.
It is no longer an excuse to blame the infancy of the industry. It is time that someone successfully puts together the biggest class action in history to ensure that the manufacturer of the next computer product to hit the shelves thinks twice before knowingly ripping the consumer off. If this leads to a slowing down of technology or price increase I will not object as a stitch in time saves nine. I do not expect the caution of a space shuttle launch to be applied, but by comparison, the computer industry is currently fueled with nitro glycerine.

[Haig]

Hi Brent,

we ran tests on a P3B-F with a G200 and a G400 using the 5.25 display drivers. Both cards worked fine.

The only changes we had to make were in the Asus bios:

Suspend to Ram - Auto
Pwr button <4sec - Suspend
Pwr up on PS2 KB/Mouse - Auto

Haig

[Brent]

Thanks Haig for trying it out and being the only "inside track" person I have ever come across that is dependable.
I am sure you did get it to work with your set up and this proves my theory about the flakiness of ACPI. I bet you did not even have a sound card in place. After all, you would want to test YOUR product without the vagaries of other drivers and apps??
All I know is that with my same set up (RRG, SB Live, Adaptec 2910C, 3DFX V2 SLI, Intel PRO 100 NIC, Fasttrack 66 and various drives etc etc), swopping the G400 for the G200 and doing the same clean install with the same drivers etc, I lose resume from STR.
On two other ACPI compliant BX boards (not STR featured), the above setup with a clean ACPI install would not even work with resume from STANDBY (S1 or S2) with the G200. At least the P3Bf and G400 combo works with S3 (only the power supply fan remains on but that might be fixed with Win 2000 and S4).
HAIG-could you let me in on some info? Is a MS certified driver, be it for a video card or anything else, a guaranteed way to build up a system that should work with all features of the OS? My brain is ticking away that whilst Matrox releases certified drivers, I have yet to see certified drivers for much of my hardware.
Also, if a driver is contained on the Win 98 SE CD, does this mean that it is certified?
I would appreciate you hanging on to this thread a while longer.
Cheers-Brent

[Haig]

Hi Brent,

You're right, we did not fill up the slots with cards. We were mainly concerned if our card was actually going to sleep and waking up. Both cards did. We even took a meter to our cards to make sure that they were going to sleep and waking up.

The only other card that was in their was a live card using liveware 2.1.

"Is a MS certified driver, be it for a video card or anything else, a guaranteed way to build up a system that should work with all features of the OS?"

Yes it will work with all features of the OS as long as that feature is in the WHQL spec.

Certified drivers are by no means any kind of optimized driver. It's just there to garantee no problems with any other device that also has certified drivers.

From a technical support poit of view, these are the easiest drivers to support. Since MS has certified them, then any bugs found gets re-directed back to them.

The reason why alot of manufacturers don't have certified drivers is because it's too expensive. They have to send an employee down to MS for a few weeks. When the employee gets to MS, the company has to pay MS for rental of space (one desk and one PC!!) while the employee runs the WHQL tests. There are other costs also involved but you get the picture.

The only positive side for having drivers certified is that the OEM's want it or else you may lose the deal.

"Also, if a driver is contained on the Win 98 SE CD, does this mean that it is certified?"

No they aren't certified. Usually, these drivers are developed by MS and has the bare minimum support of the card. In rare instances, manufacturers will send a scaled down version of the drivers to MS so that they can include them into the OS.

I think that if any major OEM yells loud enough for ACPI to work properly with all devices, then you will see no more ACPI problems.

Haig

[Brent]

Thanks Haig,
I will now retreat back to the real world and make do with what works. Hopefully time will sort ACPI out and maybe one day I will do a new installation and everything will be hunky dory out of the box.
See ya later.
Cheers-Brent