Release Management and being sober

[Caleb Tennis]

On Wednesday 26 February 2003 10:29, Eray Ozkural wrote:

> I think we're making a few mistakes regarding release management.

> Nothing too grave, but if we don't make a plan then we are going to get
> recognition as the source of bug-ridden development software like Microsoft
> Visual Studio.

> I think we should, independent of KDE release cycle, must define a few
> phases in development so that we can smoothly progress into a beta release.

I agree completely.

> My suggestion is:
>
> -- a feature freeze date
> -- a bug-fixing phase, all bugs considered important for a beta milestone
> should be fixed.
> -- a TODO decapitulation phase in which we attack the items in TODO
> emphasizing those features already present in KDevelop 2.x but lacking in
> HEAD.

I also agree, but I think that may be harder than it sounds.  Our active 
development base right now isn't very high.  Our new feature generation is 
quite a bit active, but the bug hunting isn't being hit that hard. :)

My goal is to get in sync once again with KDE releases.  I would like to do 
that by 3.2 final.  Since that's at least 6 ( and more likely 9 ) months 
away, I think we've got plenty of time to realistically meet that goal.  

However, I want to push out having multiple development branches as long as 
possible.

I have planned time this summer to finish a lot of the Gideon odds and ends, 
like website and documentation.  So, feel free to leave the bulk of that work 
to me.  Everyone else: focus on the code, the bug reports, and responding to 
technical questions on the mailing list (particularly if the problem in 
question is related to your work).  I'm trying to close code bugs as I can, 
but as everyone here knows, it's hard to know the details of all of the parts 
of gideon.

I've been waiting for Coolo to release a 3.2 timeline, but since I haven't 
seen it yet I'll just go ahead and get a rough one prepared for Gideon.  If 
we assume a 6 month timeframe, does everyone find 3 more months of new 
features, and 3 months of bug fixes / string changes realistic?

Perhaps putting together a list of "active maintainers" would be helpful too.

Caleb (comments/praises to this list, flames to /dev/null please)   :)

PS:

The main reason I wanted to call the next release "Beta" isn't necessarily 
because we are a Beta, but because it would attract more people to start 
using Gideon, generating more feedback and potentially more help with the 
effort.  It's a double edged sword, but I figure the good most likely would 
outweigh the bad.