Re: [gentoo-dev] Too many mailing lists

[Simon Stelling]

Ivan Yosifov wrote:

> I agree that no one wants to get too much irrelevant mail. However if
> there is a single dev list where improvements (and not user problems)
> are discussed there will usually be several threads that just keep
> growing. AFAIK Thunderbird (which you appear to be using) supports
> message folding and threading. So if there are a dozen messages under
> the XOrg message tree you can quickly tell that they are not for you. I
> mean that just because there are 100 messages more , does not mean you
> will need more than 10 seconds to filter them all.

Yes, it does, and I use it. But the problem is, that many problems 
(especially those you are talking about) are not that clearly assigned 
to one herd/project. What about a user that finds an error in our 
amd64-specific documentation? Depending on his subject (i.e. "Found 
error in doc" or "amd64 technotes contain errors") I will miss it or 
not. To ensure I don't miss them (as I don't want users feel ignored)
I'd have to read them, and that's the problem. You say: "You can't 
expect form a user to see the amd64 TCP/IP stack bug when all he knows 
is that gaim can not connect." Of course we can't. But the user that 
can't figure that out won't write a subject like "amd64 tcp/ip stack is 
buggy", he'd write "why is gaim unable to connect?". I'd completely 
ignore this thread, because I really don't bother about gaim. Perhaps 
the gaim maintainer would figure out that this is amd64-specific, and he 
would say: "amd64-guys, could you have a look at that?". Very likely in 
the same thread, with the same (boring) subject, and I still would 
ignore it. The user would feel ignored, and that's not what anybody want.

> I agree. What I meant was that sometimes users have ideas about
> improving Gentoo (apart from fix bug #####). And such ideas (i think)
> are for gentoo-dev.

I thought you thought so. ;)

> Mind your own example with app-foo/bar-1.0 on amd64. Most bugs (and
> problems) are arch independent. Especially problems like "How do I use
> this app" , or "where is this in the gnome menu". And these are the
> problems a user is likely to ask help for. You can't expect form a user
> to see the amd64 TCP/IP stack bug when all he knows is that gaim can not
> connect. I believe that ppc,amd64,x86,etc users (and lists ) have more
> experience to share than arch specific stuff.

see above. This is a very good example: The error could be a 
configuration error (interfaces, firewalls, even gaim), a bug in gaim or 
a bug in any other part of the OS related to internet connection. 
Someone will figure out where the error really is, but the responsible 
dev most likely will miss it. It's like spam: The more spam (unwanted 
mails) you want, the higher is the risk of missing an important information.


> I understand. However something currently going on the java list may
> very well have to do with amd64 , and you may never know about it.

Right, but if a java dev finds out that it has something to do with 
amd64, he will contact us. And he will do that on #gentoo-amd64, 
amd64@g.o or gentoo-amd64@g.o, or even in a bug, but not 
on gentoo-java@g.o, as he knows that most of amd64 devs don't 
read the list (i dont know if this is true or not, at least i don't ;).
If the whole thread would be on -dev, he probably would ask us to have a 
look at it in the same thread, because we're actually receiving this list.

Another aspect could be bandwith: I know there are a few devs that are 
not reading/receiving -dev because it has such a high amount of traffic. 
Not everybody has a 1MB flatrate, there are still some people that have 
to use a 32kbit dialup.

Greetings,

blubb