Re: [Phpgroupware-developers] Communication ... again (was PHPDoc s for head)

[Alan Langford]

At 2005 04 27 23:19, Alfonso Guerra wrote:
> Greetings,
> <snip>

Maybe Alfonso won't be reading this. I bet he will. Whatever.

Rant-like attitude aside, he makes some points worthy of discussion. I 
migrated from eGW as well, partially because of the license, but largely 
because I asked about contributing to an eGW module, got a response that 
was worthy of a politician (many words, zero content), pressed my point, 
was accused of having a bad attitude (since when is asking "is there a 
specification and a project plan?" a bad attitude?), and bailed on the 
project despite the fact that it's got some nice features that pgpGW 
doesn't have (yet). I'll freely admit that I too have considered just going 
in and making changes and creating "my" branch. This is partially because I 
might be duplicating effort (there is little way to tell) and partially 
knowing that the odds of having those changes embraced are slim, since they 
might step on some undocumented plan discussed in IRC, or they might 
violate some unwritten philosophical principle.

But Alfonso is dead wrong about open source. Competing forks are good 
things in the end; eventually a few will come to dominate, and then the 
best features will migrate into the dominant products. It's a messy 
process, but competition is evolution, and evolution is messy. It also 
takes time, and developers aren't known for their patience. If you can't / 
won't contribute or wait, or if you don't have the conviction to argue that 
your changes are worthy, then maybe an extensible commercial product is a 
better option.

One thing is clear to me though. Really successful OSS projects have strong 
supporting infrastructure. At this point in OSS evolution, this usually 
means the project has a single corporate sponsor with a strategic motive or 
a profit motive, and that motive focuses the project (for example 
eclipse.org). But that's not always the case; some times there's a strong 
individual who can put all of his or her efforts into it (the Linux kernel 
would be a good example here).

This is what projects that suffer from forking, and from 
single-implementation dialects seem to be missing: Some kind of clear set 
of goals, objectives, time lines (even if they are subject to ongoing 
revision), an easy way to look at features that have been suggested and 
their status (new, approved (with a target release and estimation of effort 
required), rejected (with detailed rationale so people can learn)), an easy 
way to make a small contribution, documentation that makes it easier to do 
a little bit of work without becoming an API guru, frequent commits, clear 
targets for releases... in other words a lot of stuff developers would 
rather not do when there's code to be written!

Every project I've worked on from time to time has a couple of guys who 
just can't resist taking shots at each other. Any imperfection by one gets 
a nasty remark in response. Every possible solution requires some 
interpretation and evaluation, and of course the other guy always makes an 
inferior choice. It's a great ego-fest to them, but merely distracting and 
tedious to the rest. In the end it's just noise and if it's discouraged, it 
goes away sooner or later. We just need to keep stepping in and saying 
"have your pissing match in private, please", or worst case, moderate the 
messages out and remove their audience. We have a bit of that now and man 
is it dull reading.

Let's not forget that phpGW is a pretty cool product, and something the 
contributors can be extremely proud of. Sure there's lots of work to do, 
lots of features to add, and bugs to fix. Maybe nobody as the time, tools, 
or inclination to step up and manage the development more aggressively 
(before anyone jumps on this, please note I'm talking about managing the 
project, not managing the people -- these are vastly different things). But 
the project has come a hell of a long way without that kind of effort, and 
there's nothing to say that it won't continue to do so. I can understand 
Alfonso's frustration, but posting a diatribe and running away is hardly 
useful. It's also not useful to just discount what he was trying to say, 
just because he said it in an inappropriate way. Better to adapt, evolve, 
and just stay focused on producing better software.