Re: stable vs. experimental archive

[John W. Eaton]

On 22-Apr-2009, Robert T. Short wrote:

| One big difference between the stuff I have done and the 
| octave model is that I always insist on a lot of regression testing to 
| make sure bugs get eliminated and stay eliminated, but that is really a 
| matter of discipline on the part of the programmer.

I admit that I tend to be lazy about adding tests when for bug fixes.
We should all probably try to do a better job with that.  Likewise
with adding documentation and NEWS entries for new features.  Also,
patches to add tests are always welcome...

| I actually have several local repositories that I consider 
| experimental.  I play around, add debug code, whatever, but don't feel 
| at all guilty about trashing the whole thing.  I see no reason to even 
| make these publicly available at all.  However if someone wants to play 
| around with something serious lik (e.g. making octave multithreaded or 
| some such), then a public repository would be valuable so more people 
| can see what is going on, contribute, etc.

Yes.  To me, this is the most important thing about distributed
version control systems.

jwe