[Project_owners] A MozDev Mission Statement?

John Dobbins john at brandxcomputers.com
Wed Sep 17 20:01:17 EDT 2003

Pete Collins wrote:

>> Lets face it. I have to live up to 3 versions of the code that is 
>> shipped by mozilla.org. I won't maintain another.
>> Note: I'll give this a day or two, but right now, I see just one 
>> argument and no real reason to keep this thread (and my offer of 
>> time) alive. Thanx to Fabio for bothering to be on-topic, though.
> Confucious say:
>   "When you build house on weak foundation, house fall down".
>   "When you build app on moving platform, app fall down"
> This is the conundrum we've been facing since before the existence of 
> Mozdev.
> --pete
Before Mozilla 1.0 was released there was some excuse for things being 
unsettled. When you are developing for a platfom that is under 
development itself a moving target can be expected. I Understood that. 
What I didn't understsnd was the lousy communications. It wasn't a case 
of  "In Two weeks we will be making some checkins that will cause the 
following problems.....". It was SURPRISE! your application dosen't work 
with today's build! or SURPRISE! That bug fix for Mozilla that you spent 
days working on that is allmost ready is busted! Any work you did over 
the last two weeks was a waste of time.

Changes prior to 1.0 aren't what made me mad, it was the damn surprises. 
Changes without warning, ones that not only surprised people working on 
outside projects, but also community members who were trying to fix bugs 
in Mozilla itself.

After Mozilla 1.0 was released point releases shouldn't break 
applications. An application that worked with 1.0 should work with 1.1, 
1.2, 1.3 etc

Things that break aplications, because they rely on features that are in 
the 1 series should have been introduced in a new development tree that 
would become Mozilla 2.0. Things like the birds belong in a 2.0 series 
tree that would target a release date about 2 years after 1.0 was released.

Right now there isn't a stable series of releases that a developer can 
count on being viable for a couple of years. Instead there are 3 or 4 
releases a year anyone of which may or may not break applications and 
addons for Mozilla. If you are lucky your project or bug fix you can't 
get someone to check in might work for 6 months. After that you are 
rewritting it to get it working again instead of improving it or 
starting on a new application or bug fix.

John Dobbins

More information about the Project_owners mailing list