On Wed, May 16, 2001 at 08:46:48PM -0400, Dan Streetman wrote:
> I didn't mean to personally offend you or any of the other usb-uhci authors with
> my opinion. I was only saying that, for both uhci and usb-uhci, I tried to
I wasn't offended by the opinion itself (you may like what you want), it's
just the argument and I wanted to give a statement why usb-uhci looks like
it looks. And besides the clarity of the code (and yes, usb-uhci has a few
complex and not very easily understandable parts) there's another (for me
more important) point: It works (in most cases...).
> follow the flow of a URB (from submit to completion), and I thought that uhci
> was easier to follow. It had less conditionals, different execution paths based
> on URB info, etc. In my opinion, less complicated is usually better. It very
> well may be that usb-uhci performs better, I really can't comment on that. But
If you would have said "uhci is 10% faster/more stable", no problem...
> I think in many cases, code that is difficult to follow, is difficult to modify.
> And code that's difficult to modify is difficult to maintain, and it winds up
> that very few people really understand what it's doing.
>
> I certainly did not mean to imply that it is bad; I only was trying to say, if
> the time comes that for whatever reason one of the UHCI drivers goes away, I
> recommend the one that's easier to follow, because more people will be able to
> understand exactly what it's doing.
That isn't true for the UHCI-HCDs, most people don't _want_ to understand
what's going on, they take is as god given. With the exception of a few
patches, noone besides the original authors really worked on the internals
of the HCDs. There aren't even questions about how things internally work,
AFAIR there were no questions eg. about the bulk queuing method in usb-uhci
or why the interrupt-TDs in init_skel() are distributed with that weird
for-loop.
As some severe host controller crashes with usb-uhci appeared in january
2000, and we weren't immediately able to fix that, also nobody tried to find
out what's going on, the only reaction was a long list thread "did we make
the wrong choice?". Now it's fixed, but does anybody care (except me...),
what exactly the cause was and which additional lines in usb-uhci actually
take care for that? There are a lot of comments in the code, but I can't
write everything...
> Anyway, that really was a side note to my email. Do you have an opinion on
> the patch? If you could take a look at what I did in usb-uhci I'd appreciate
> it.
I still don't see why the patches are actually needed, I have to think about
that again.
--
Georg Acher, [EMAIL PROTECTED]
http://www.in.tum.de/~acher/
"Oh no, not again !" The bowl of petunias
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
http://lists.sourceforge.net/lists/listinfo/linux-usb-devel
