On Sun, Jan 02, 2005 at 09:29:37PM -0500, Benj. Mako Hill wrote: > <quote who="Helen Faulkner" date="2004-12-30 01:46:38 +0000"> > > It strikes me that testing whether or not I can spend 40 hours in a week > > doing all my T&S stuff in one big hit is not very useful to Debian. My > > impression, after living with a DD for years, is that what is more useful > > to Debian are people who are prepared to put in a long commitment that will > > mainly involve small amounts of work that are distributed in time, but > > which ideally require a quick response as they come up (ie to fix a bug > > that appears in your package, but there might not be a bug for weeks or > > longer at a time). > > I think that having that applicant put in the small amount of time > over a long period of time so that their skills and ability to follow > through on the Debian tasks they intend to do as developers become > apparent should be tasks and skill test enough. > > Early on in the NM process, few AMs asked any T&S questions at > all. They expected NMs to demonstrate that they had the skills by > doing the tasks. I still think that a history of doing good work, an > audit of packages, and a few good recommendations say as much or more > than any number of questions broken up in any number of chunks. In > most cases, they make the questions themselves (and the 10-40 hours of > work that answering them entails) of questionable use. :)
I agree. Basing the NM process on lots of questions does a really good job of demonstrating how well an applicant can answer lots of questions, and that's about it IMO. There isn't a very strong correlation between answering questions well and actually being a good developer. Note that I went to a university that focused very heavily on test taking, and very little on practical skills. As a result, I have a nice shiny mechanical engineering degree but can barely assemble a typical ballpoint pen (damn that spring). So, my opinion is probably a little biased... -- For every sprinkle I find, I shall kill you!
