18 Jul
2009
18 Jul
'09
12:36 a.m.
I think I would be a good example of a potential new developer, one that
would be valuable and interested in the project. So perhaps my opinion may
count to an average of the way others like me would see ReactOS.
Maybe my opinion might be also valuable because I have been managing large
scale projects for 12 years (embedded and distributed OSs, RT, mass
transport automation, electric distribution networks automation, robotics,
scientific satellites, etc), but never with open source projects, so take it
accordingly, I don't intend that my opinion is necessarily valuable for
ReactOS.
The goal is very, extremely attractive, and once wine is working, we think
with reason that a very meaningful part is done. I'm sure feel very
attracted.
But after studying the code a bit and following the mailing lists for a
while, the perception of organization and progress is poor, what introduces
the doubt: there is a lot of work to be done, would my effort be wasted?
I assume that every developer in the team is motivated by the goal, while
each one may have specific areas of interest, preference, of experience.
Everyone would understand that their own efforts would be a waste if the
whole project does not success. Everyone needs a minimum infrastructure and
organization to progress efficiently. So, everyone should understand the
need to address high priority areas that are slowing down everyone.
The highest priorities IMHO are:
1) Documentation. I know most devs like me hate having to write it.
But complex projects can't progress well without a minimum well structured
documentation. Someone has to define a basic hierarchy and rules. Whenever
someone knew or learned anything that is undocumented, update the docs, so
that others can work coherently. Maybe some of you think there is some, but
to me (please remember my outsider point of view) it seems very poor.
2) Drivers installation. The drivers needed to run ReactOS in any
hardware or VM are already available. It is prioritary, at least, to have
.inf parsing and adv/setup api working well. If developers can't start the
OS in the platform they need to test hardly will anyone be efficient, if
does not leave the effort at all. Of course drivers can fail after
succesfully installed, as so many things, but bugs in .inf installation are
first time stoppers and need to be addressed asap.
3) Bug reporting and tracking database. I'm sure you all agree
regarding the necessity of a well designed one. Someone has to do it, a
project like this can not progress far without it.
Again, take this with a grain of salt, I don't know ReackOS like most of
you, I intended more to show first time perceptions that may atract or not,
than being accurate in the criticism.
After explaining my perception, I tell you that I sincerely want to
contribute as soon as I can, I just don't have enough time currently.
Jose Catena
DIGIWAVES S.L.