| 1 |
Mart Raudsepp <leio@g.o> posted 1241633816.25192.2.camel@localhost, |
| 2 |
excerpted below, on Wed, 06 May 2009 21:16:56 +0300: |
| 3 |
|
| 4 |
> On Wed, 2009-05-06 at 17:25 +0000, Duncan wrote: |
| 5 |
>> Christian Faulhammer posted: |
| 6 |
>> |
| 7 |
>> > recruiters do quite long IRC sessions with the applicant. Apart from |
| 8 |
>> > questions not found in the quiz they want people to elaborate on |
| 9 |
>> > answers they made. |
| 10 |
>> |
| 11 |
>> As others have occasionally noted, the assumption seems to be that |
| 12 |
>> developers "do" IRC. |
| 13 |
> |
| 14 |
> Are you suggesting that recruiters should do long e-mail exchanges with |
| 15 |
> the applicants instead, having no real time conversations, leading to no |
| 16 |
> idea about the applicants real knowledge (when there is not much time to |
| 17 |
> do research after a question is posed), attitude and so on? |
| 18 |
|
| 19 |
Noting that mail turnaround can be seconds or minutes, particularly when |
| 20 |
needed and arranged beforehand... like say an IRC session might be... |
| 21 |
|
| 22 |
There's a saying that the mark of an expert isn't that he knows |
| 23 |
everything, but that he knows where to look to find it when he needs it. |
| 24 |
|
| 25 |
I'd suggest that applies here. |
| 26 |
|
| 27 |
In the Gentoo development environment, what's so pressing that a few |
| 28 |
hours' delay checking a reference to make sure something's done correctly |
| 29 |
is a problem? If it's not a problem "on the job", why make it a problem |
| 30 |
"for the interview"? |
| 31 |
|
| 32 |
Basically, I'd argue there's no vital information obtainable by an IRC |
| 33 |
interview that can't be obtained in an email exchange. There's |
| 34 |
certainly /some/ additional information available in the instant format, |
| 35 |
but I'd argue it isn't vital information given a longer view and a |
| 36 |
history to work with, and in fact, that said additional information is |
| 37 |
quite trivial. |
| 38 |
|
| 39 |
I'd argue that real knowledge and attitude should easily be apparent by |
| 40 |
the time of a serious interview, whether it's via IRC or mail, in any |
| 41 |
case. As the developer's handbook points out, the first step in the |
| 42 |
process is simply "helping out". What's the bug submission history? |
| 43 |
Patches? Sunrise and/or project overlay and/or AT record? Whether on |
| 44 |
the various devel and project lists or IRC channels, what has been the |
| 45 |
candidate's attitude? Do they take well suggestions from others, yet are |
| 46 |
able to take their own positions and defend them technically when |
| 47 |
necessary? Are they able to discern when it's a big deal and when it's |
| 48 |
not? When they screw up, do they apologize, then work to fix it |
| 49 |
themselves, asking for help when needed, or are they either slacking off |
| 50 |
waiting for someone else to fix their mess, or refusing offered and |
| 51 |
needed (time-wise or technically) help? |
| 52 |
|
| 53 |
Are you saying that history, generally of months if not years[1], is |
| 54 |
easily faked, and that an IRC session is better at detecting such fakes |
| 55 |
than an email exchange of approximate equal depth would be? If you are, |
| 56 |
I must say I disagree. I just don't see it. |
| 57 |
|
| 58 |
[1] There's a place for shorter term commitments, see the SoC, bugday, |
| 59 |
simply pitching in with patches or testing, etc, but those aren't Gentoo |
| 60 |
devhood. |
| 61 |
|
| 62 |
-- |
| 63 |
Duncan - List replies preferred. No HTML msgs. |
| 64 |
"Every nonfree program has a lord, a master -- |
| 65 |
and if you use the program, he is your master." Richard Stallman |
Archives