Mobo Compatability

Aleksandar Milivojevic amilivojevic at pbl.ca
Tue Jan 4 16:02:08 UTC 2005


Michael A. Peters wrote:
> On 01/03/2005 05:44:11 PM, Felix Miata wrote:
> 
>> http://www.mozillaquest.com/Linux04/Asus_Sucks_Story-01.html
> 
> 
> I've never had any problems whatsover with Asus boards and Linux.
> Talk is cheap.

[snip]

> When the IEEE 1394 didn't work (RH8 days) I didn't expect Asus to fix  
> it, I expected either OSS or nVidia to fix it. OSS did.
> 
> When the onboard 3Com nic didn't work - I didn't blame Asus, I  
> downloaded the 2 line kernel patch that made it work. When the nVidia  
> network adapter didn't work, I had the option of using a closed source  
> driver from nvidia. Instead I chose to not use that adapter.

I kind of agree with you.  If we go that route, than we should not buy 
any hardware at all, because most hardware manufacturers are just like 
Asus.  If they can get away by not providing support, they'll do it.

My favorite example was when Intel refused to provide support for one of 
their's motherboards.  The motherboard would not power up if more than 
two Intel dual-gigabit network adapters were plugged into the 
motherboard (actually, it would power-up, but everything would freeze 
before there was any output to the monitor, so the screen would stay 
pitch dark).  Both motherboard and network adapters were original Intel 
(entire board, not just the chipset, boxed by Intel (not OEM) and so 
on).  The reason behind buying is that since it's all Intel, if anything 
dosn't work, it would be all supported.  Wrong.  The official reason 
support was denied?  I mixed components with "desktop" label in name, 
and components with "server" label in name.  I mean, it's a standard 
motherboard and standard PCI card, no matter what some silly marketing 
label says.  That is what specs for both components say "supports all 
PCI cards" and "works in every motherboard with standard PCI slots". 
They should work together no matter what.  If they don't, one of them is 
broken.

The only answer I could get from Intel is to buy much more expensive 
quad-CPU Xeon motherboard (if I want motherboard that they would support 
with more than 2 dual-gigabit cards plugged in).  Not likely.  I'd 
rather buy another manufacturer mobo and cards that do work together for 
fraction of price of that quad-CPU mobo.

Apperently, "desktop" and "server" products are supported by two 
different departments inside Intel that do not talk to each other, so 
they refused to even make a test configuration in their lab.  They never 
responded to my last couple of attempts to contact them about the issue. 
  Later, I found out that that mobo had exactly same issues (gets stuck 
on power-up) with some other cards (in more standard and simple 
configurations).  Some of them were fixed by BIOS updates, but not this 
one.  Another story of badly designed product, and abandoned customers. 
  I kind of didn't expected that from brand name like Intel.

So, it is not just about support for particular OS.  It is about support 
in general, and the way manufacturers are handling it.  My guess is that 
Asus from original story would try to avoid providing support even if 
the issue was Windows related (as long as it isn't something that 
would/could hit average home mail-reading/web-browsing user), or not OS 
related at all (by reading the article about Asus story, it wasn't OS 
related, NIC worked under Windows by pure luck, but it still had 
hardware flaw).  Just as the Intel does, and just as most of them do. 
Only if you are really lucky to get to the right person somewhere above 
the lower levels of support chain, than you get "great support" stories 
(I wasn't lucky to fight my way to that "right person" in my Intel 
incident).

-- 
Aleksandar Milivojevic <amilivojevic at pbl.ca>    Pollard Banknote Limited
Systems Administrator                           1499 Buffalo Place
Tel: (204) 474-2323 ext 276                     Winnipeg, MB  R3T 1L7




More information about the fedora-list mailing list