[Ksummit-2009-discuss] Meeting userspace requirements

[Ingo Molnar]

* Greg KH <greg at kroah.com> wrote:

> On Sat, Jul 11, 2009 at 02:07:16PM -0600, Matthew Wilcox wrote:
> > On Sat, Jul 11, 2009 at 09:34:55PM +0200, Ingo Molnar wrote:
> > > and the thing is, the most difficult technical problems in Linux 
> > > tend to be cross-discipline - so them touching a generic list like 
> > > lkml is an obvious bonus. Noise is a problem but can also be 
> > > filtered out - on the other hand the lack of cohesion and the lack 
> > > of a broader picture can cause irreversible loss.
> > 
> > Can you share your filtering technique with those of us who just can't
> > handle the lkml firehose any more?
> 
> I use mutt's filters to handle this.

i use that too, yes.

> I've seen others use gmail to filter quite well also, in fact, I 
> do both now (mutt and gmail) and it seems to catch everything that 
> I'm interested in.

Yeah, i know several people who do that, that method works as well. 
procmail works too.

There's a technical solution i've suggested for this before: adding 
specialistic 'alias' / 'mirror' lists on vger.kernel.org, which 
emits to lkml but allows specific receive. That would allow people 
with a general interest to still see all activity, without having to 
subscribe to hundreds of splintered email lists.

Semantics: say we'd create linux-tracing at vger.kernel.org. Currently 
all development is done on lkml, but this list would offer 
interested people to subscribe to that list in isolation - without 
getting the rest of lkml traffic (which they often perceive as noise 
and overhead).

The difference to other lists is that all mails from linux-tracing 
also get posted to lkml. So i dont _have to_ subscribe to that list, 
and i can get the whole picture from lkml alone. But people can 
subscribe to that list and can read it and post to it, without 
getting the rest of lkml.

Anyone someone posts something tracing related - or a thread gets a 
not previously expected tracing angle, it can be Cc:-ed to 
linux-tracing at vger.kernel.org.

This would allow isolation/filter, but would still maintain the 
general picture as well.

	Ingo