[CentOS] IRC telling you to go to mailing list - it's for good reason ....

Sun May 30 15:22:46 UTC 2010
R P Herrold <herrold at centos.org>

On Sun, 30 May 2010, Gabriel - IP Guys wrote:

> Just to answer the aspect of why people in IRC tell you to 
> go to a mailing list – because mailing list answers are more 
> permanent that IRC, and also, it spreads the knowledge of 
> the question, and the answer over a much wider time frame 
> rather than just the two to three sentences that are 
> involved in fixing your situation.

You omit mentioning that (from tracking latencies to 
appearing) the CentOS site, wiki, MLs, CentOS' planet are all 
regularly trawled by major search engines.  The Forum and the 
bug tracker are also indexed but less often. So an answer 
there in these non ephemeral 'gateway into being readily 
re-findable ... thus some of the extended posts I make

> These days people treat IRC like a street market for info, 
> there are no rules other than do not paste more than 3 
> lines! (pastebin is your friend) and you can say what you 
> like to who you like, and log off, without so much as 
> contributing to the community

Drive-bys and externalities are a problem and under the old 
design the CentOS wiki was not such a pigsty; some other team 
members lobbied for a less OCD approach on what is in and off 
topic, and how 'spoon-feeding' was treated. The jury is still 
out on this new approach in my mind; I know I get push back 
from old-timers who preferred the higher signal to noise 
ratio, and want 'quality' back as a preferred metric over 
'friendliness' ...

> IRC has been around for a long time, but unless the irc 
> server is irc.(project domain name here).org then expect it 
> to be staffed by volunteers and people who may not have the 
> same focus as yourself

The CentOS project has cloaks both for team members, and 
trusted sources in IRC nics to address the issue (think of the 
clerk at the local hardware store, wearing a distinctive 
over-vest to mark them ...)  A quick: /whois (nick) <CR> can 
tell a lot

-- Russ herrold