On 2007-02-08, csant <[EMAIL PROTECTED]> wrote:
> leaving this work to the community can lift away some  
> workload from you: but might add the additionl supervising/clean-up task -  
> which, of course, could be distributed among some core ion users or  
> volunteers. 

I wouldn't mind cleaning things up that much, if I didn't have to do
it through a poor editing box in a Web browser... and fight spam all
the time.

> Speaking of which: you ask people to send in patches  
> via darcs, but occasionally some contributors don't manage to get their  
> configuration right. There aren't that many external contributions, but  
> how often do you have to take over those patches sent in via some other  
> way? 

I reject patches sent in other ways... nobody has sent an important
enough patch yet to not be able to reject it. It's too much work
applying a plain-old-diff, figuring out what new files need to
be added under version control, thinking of a log message, whether
the author of the patch should be listed somewhere, and so on.
With darcs patches, all I have to do is type '|darcs apply' in
mutt, in the working directory with the email selected, 'darcs push' 
to the "local central repository" and then run my 'darcs_sync' script
to push the changes online and generate the feeds. And it is this
latter script that should be improved to actually generated and push
some content as well, if a poor man's wiki were created. Additionally
for regular contributors, perhaps some GnuPG key and .procmailrc 
automation should be created...

-- 
Tuomo

Reply via email to