2.3.51 tulip broken

[Jeff Garzik]

David Ford wrote:
> Jeff Garzik wrote:
> > Changes are easier to isolate with incremental changes.  One big update
> > means things work, or they don't -- with the fallback (before the
> > update) typically being a far inferior, and more buggy version of the
> > software.
> 
> If the tulip driver was such a simple thing as a=b=c then that would work
> nicely.  It's not.  One change for one chipset often breaks another chipset.
> Most of the time it is subtle and sometimes it is undocumented.
> 
> The way Donald does changes is right IMO.  It's in a change constrained
> environment and he has a much better grasp on the effects.  When he's happy with
> a group of changes all getting settled neatly in place then out comes your big
> patch...with the majority of cards working.

This is not the case at all.

Network cards, and tulip in particular, are such a mess that ANY large
driver update is likely to break more cases than it fixes, generally. 
It is simply impossible for Donald to have enough feedback in isolation,
compared to the amount of feedback and testing given to kernels in the
regular Linus tree -- even the development series.

Further, if there is ANY problem after the huge driver update, it is
much more difficult to isolate problems.

	Jeff


-- 
Jeff Garzik              | My to-do list is a function
Building 1024            | which approaches infinity.
MandrakeSoft, Inc.       |
-------------------------------------------------------------------
To unsubscribe send a message body containing "unsubscribe"
to linux-tulip-request@beowulf.org