Summary and few tips

Hi everyone,

I just read the entire thread from December to January
and I have found out few things about Y.

Mark seems like a nice guy, he created Y, but unfortunately like most of us,
he got a job, working hard and can't find real time to code much.

The same seems to be true for the other 'core' coders.

It seems that some of the 'core' coders have an attitude which is relatively 
normal
in the community, but here that attitude has conducted into the expected and 
typical
'open source management failure'.

Another problem is that 0.x project leaders seems to *see* too big, too 
large, too fast,
thinking god will save them and a huge number of programmers will join their 
toy project,
it's not like that unfortunately. Most of the time, it's mitocracy by *core* 
programmers
with few patches here and there from anyone in the crowd with relatively low 
quality.
So, think low, think small, make it work and people will join your project,
features will come in time. Don't start too big, you'll just crash and burn 
out.

This means that discussing about supporting pixel shadders, unicode or 
whatever francy stuff,
when low level things don't work well, is quite inappropriate, for now.
Use existing stuff and think about the boring, small and concrete stuff.

That being said if it takes 100 dependancies or a weird thing to make Y 1.0,
don't ask, just please do so.

Don't reinvent the wheel, the broken wheel and the zillion dependancies can 
be fix later.
If you gonna put 40h/week on this project and just want to rewrite it,
get rid of arch, sigc++, autoconf, C, C++ whatever, don't ask just do so,
you probably gonna write more code than anybody will write in the comming 
weeks
and your fork will probably be more important than the current statu quo.

If you don't like the GPL, but like the protocol spec, just do clean white 
room fork from
the thesis specs and rewrite the server in whatever way you may wish with 
whatever
license you may wish.

Most people that are bright and capable enough and who start off their first 
open source project
normally fall into the same trap. First, you have plenty of time and then 
projects deadline and
overtime kill you off pretty quickly.

The other problem is that with FOSS, nobody's get a living out of it,
especially when it start of, until it gets really popular.
So, it's hard to recruit people, it's hard to find time ourself to work on 
it, etc.

The only time where people get paid is that it scratch a company itch
and the company decides to invest into the project as a side-effect.
For instance, Dan Sugalski for the Parrot engine [perl.perl6.internals]
or really big project getting sponsered by IBM, OSDL, etc.
Let's be realistic, probably nobody's gonna sponsor an immature toy project 
in 0.x stages,
just a tip, if you get to revision 2.x or 3.x maybe you have more chances to 
get sponsered.

Therefore, if some sort of capable newbie join your project, you should 
embrace that person,
not reject him/her. Especially, if that person can contribute time to your 
project.
At first sight, doesn't matter if there's bugs or security flaws, those can 
be corrected by you
and over time you can train that person to avoid such glitches.
Even if that persons is kinda annoying with what sounds like stupid 
questions to you.
If that person cannot use CVS, Subversion, arch, diff or whatever, it 
doesn't matter.
Ask him/her to send you plain files by email, ftp or sftp or mailing list or 
whatever.

I admit that its very frustrating for competent developers to welcome 
newbies
but that's life, it's about the same in the industry anyways, so deal with 
it.

Patience is a value. If you can write a FAQ, that's good,
that will save you time over repetitive questions.
Simply repost the FAQ every 2 weeks to the mailing list.

As far as I can see, there's no more any *real project leader* for Y or Y+.

There's two *honorable leader positions* mainly Mark and Andrew,
but in terms of mitocracy there are no more, unfortunately.

So, if you *really* have free time, be very conservative please...
and wish to make Y or Y+ or any other fork based somehow
on the Y protocol or something similar, please say so.

Of course, you should know basic programming,
but it's okay if you don't know any repository system
whatsoever or fancy stuff, that can be deal with in time.

I see that the wiki don't work anymore... oh well. =(

So, if they are any volunteers please say so.

Have a nice day!

Sincerely yours,
Fred.