Re: Linux-2.1.62..

James Mastros (root@jennifer-unix.dyn.ml.org)
Tue, 4 Nov 1997 15:40:20 -0500 (EST)


On Mon, 3 Nov 1997, Linus Torvalds wrote:
> On Mon, 3 Nov 1997, Matthew Kirkwood wrote:
> > I'm happy to collate all of these into one big patch if it
> > would make a useful difference...
>
> Thanks. I don't want a big patch that collates these changes: it's
> actually easier for me to apply 2 small patches than one large one.
>
> But what I _would_ like to see is patches in my mailbox that have been
> tested - I follow the kernel mailing lists very spottily, so sending the
> patches to the mailing list does not mean they reach me. I still encourage
> people to do that (that's what the mailing list is there for), but please
> consider that a "testing phase", to then later send the patches to me
> personally...
Absolutely... Test, then send 'em up to the big man... In fact, when you
post patches to the list, it would be really nice if you said
1) How well tested it is (looks nice; compiled OK; kernel ran but I don't
have the hw; worked for me; tried on a bunch of machines....)
2) What the dangers are (might eat your hdd alive...)
3) What testing we should do (do you want us to check it with big I/O
loads? Low mem avaible?)
4) Do you want positive reports (It fixed my problem), null reports (no
change - I didn't have a problem and still don't), negitive reports (I
had a problem, and I still do), or really negitive reports (I had no
problem, but now I do)?

> Also, final comment: if you _have_ sent patches to me, and they didn't
> show up in a kernel and you heard nothing back about why they didn't show
> up, please don't despair - re-send them instead (possibly with a longer
> explanation of why you want them in the standard kernel).
When you reject a patch (or for that matter, when you don't), it would
probably be nice to write a quick note saying why (user-space problem, messy
solution, "OK, but there's this really cool thing that I've been meaning to
do that would fit nicely with this..."), so the submitter knows how to
rework their patch (or eventualy wether to re-submit if if they don't get a
message from you in a bit or eight)...

> I'm afraid that
> some of my mail does tend to disappear: not because of unreliable email
> per se, but simply because I get a lot of it and if I'm busy doing
> something else I often overlook email I shouldn't overlook (but once it's
> been overlooked I -very- seldom have time to go back to old email).
You might want to make a folder of kernel-patches to take a look at
when you get a half-second (such as when you are waiting for a compile...
Assuming that you're still using a computer that takes non-negitive time to
compile.)

>
> Linus
>

-=- James Mastros
---
When the annals of distributed computing are written, and the name 'Bovine'
appears in there, I can say "Hey, I was a part of that, I checked .0012% of
the keyspace".
-=- Brian Wilson <wilsonb@mindspring.net>
Go to www.distributed.net before I come make you!