Who do you think is to blame

[The PIT]

Interesting discussion going on here,

http://bbs.pcstats.com/viahardware/messageview.cfm?catid=25&threadid=5225

Roj the moderator seems to think that Via shouldn't test major products with their chipset and the via 686b bug isn't Vias problem.

[ayoub_ibrahim]

I think Roj is the weakest link in there!

[Kosh Naranek]

I second that !

[Greebe]

Sure glad I don't advise like that!

(what a maroon

[Dr Mordrid]

Roj has some very valid points. So do those on the other side.

VIA does have a history of bugs in the cache pipeline and the bus controllers that affect bus and memory throughput. These bugs go all the way back to the MVP3 chipsets used on the Super7 boards. Since VIA also makes Apollo those same bugs were on those chipsets too.

These cache and bus control problems cause contentions between high bandwidth devices that are running at the same time (ex: SBLive, video capture &/or SCSI or RAID controllers). Back with the MVP3's they also could cause some AGP cards to crash out when fault data was sent back to the card from the cache.

The 686b bugs Roj mentions are real and another maifestation of VIA's poor design quality control.

On the SBLive side of things it's common knowledge that their drivers are poorly written and the card seems designed to hog the bus. Not good for a card that may have to share with capture cards, SCSI and RAID controllers. One reason why is that it constantly hits the system interrupt timer for God knows what reason.

Such a card paired with a chipset that has a faulty Southbridge is a poison pill. Otherwise why would VIA and Creative have acknowledged their collective problems?

Dr. Mordrid

[Greebe]

Doc

[The PIT]

I agree with you Doc.
My only problem with Roj is that he won't admit Via should have picked this up during testing. He's just saying Via released a new chip its up to Creative to redesign their soundcard or anyone else for that matter. He also calls Creative a small company Mmm. I think he hates Creative that much that it clouds his judgement somewhat.
He also calls the orginal poster a fool changing his motherboard which I think is a bit off.
Just about every post he replies in he says dump the live card which isn't helpful to the person whos just bought a new motherboard. When it isn't gaurenteed to fix the problem.
Other soundcards are effected and in even worse cases theres data corruption without a soundcard fitted.
This is ignoring problems like Zip drives not working CD writers having problems when udma enabled plus usb problems. These pop up all the time in the viahardware forums.
By the way I haven't forgotten Intels big blunders of the past either.

[Dr Mordrid]

Yup, plenty of blame to go around. In this case Creative AND VIA both have the hip-high piles though. Everyone else is just ankle deep in s***

I'm waiting to see what the reports are on Creatives new DSP card say. Supposedly they learned something, but I'll believe it when I see it.

Dr. Mordrid

[The PIT]

Roj is still willfully missing my point about testing. Thats the only point I'm disscussing. Still it is a viahardware forum and they're going to be slightly biased.
I guess his sig should say Blind faith manages.
When the 686b bug first reared it's ugly test the blame went like this.

1) Creatives fault buy Santa cruz instead.
2) Lunix works so it's Microsofts fault.
3) Winxp works so it's win2k fault.
4) Oh both lunix and winxp sometimes have problems so it's creative fault.
5) It's abits fault buy msi instead and santa cruz
6) It's acpi fault.
7) Oh Via say they have a bug but it's only with sblive.

The truth is it's Via's fault (poor QA as already started) Creatives some motherboards manufacturers again not applying proper QA. Ooops forgot Microshaft leaving the debugging code in DX8 which produces crackling etc setting the debug level to zero has a wonderful effect.