Re: project infrastructure

On Fri 02 Dec 2011 01:05:57 PM EST, seanh wrote:
>> 1. Move from gitorious to github.  In doing this, we gain post-commit 
>> hooks some of which are wildly useful, an issue tracker, inline comments 
>> for pull requests, probably an order of magnitude more people who already 
>> have github accounts and who are used to drive-by-fixes, and probably some 
>> other github features I haven't yet discovered.
>
> Github can host the project website for you as well, and a wiki.
>
> The one disadvantage is that it's not open source whereas gitorious is.
> Hosting your git repo on github doesn't tie you down to anything as you
> can always move it somewhere else if you want to. But the more you come
> to depend on github's extra features like the issue tracker, the wiki,
> etc., the more you're becoming dependent on a proprietary platform.

I dig what you're saying.  I definitely don't want to repeat another SF 
situation--that sucked.

I currently plan to use github for code, the issue tracker, and patch 
workflow.  We can trivially move the code elsewhere.  The patch 
workflow is transient, so if we move elsewhere we'll lose that and 
change workflows.

That brings us to the issue tracker.  I'm pretty ok with ditching 
issues in one system and moving to a new system.  More often than not, 
issues that languish are either fixed, uninteresting, not fleshed out 
or in some state where ditching it isn't a big deal.  I don't want this 
project to accrue task-debt.  We either work on things we're interested 
in, or we don't.  The set of issues we're interested in working on 
changes.  If I had my druthers, I'd have our bug system auto-expire 
bugs that haven't been touched in a couple of months.  Given that, 
switching to the github issue tracker is fine and doesn't lock us in.  
I'm game for hosting our own bug tracker, but I don't want to use Trac, 
Roundup or a non-Python-based tracker.  That really limits the options. 
 Essentially it means I should write my own bug tracker that meets my 
needs.  I may do that at some point, but not any time soon.

The website (currently at http://pyblosxom.bluesock.org/) will stay 
where it is and act as an anchor for the project allowing us to switch 
infrastructure without losing people.

That's what I'm currently thinking.

/will

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure 
contains a definitive record of customers, application performance, 
security threats, fraudulent activity, and more. Splunk takes this 
data and makes sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-novd2d