Date: Mon, 01 Mar 1999 14:05:47 +0100 From: Gianmarco Giovannelli <gmarco@scotty.masternet.it> To: ports@freebsd.org Cc: freebsd@scotty.masternet.it Subject: Too many open pr ? An idea... Message-ID: <4.1.19990301131907.009936f0@194.184.65.4>
next in thread | raw e-mail | index | archive | help
Hello to everyone. I want only to describe an idea I have to deal with the
too many open pr (ports).
I think that there are a lot of hackers/users/programmers that are trying
to improve the availablity of software on FreeBSD. The ports are a good
idea, but in my opinion suffer of some drawbacks.
1) If the port is not committed is unusefull; infact too many freebsd users
don't read the gnats or the port mailing list, but rather try to find what
they need in our ports tree.
2) If the port is old or older than the current version available on the
net is unusefull too, and this is a nonsense if we have a port or un
upgrade lying here and there waiting to be committed. Sometimes our
distfiles are not fetchable anymore.
It want NOT to be a critic towards any of the committers. I think and know
they do all the best they can to improve our OS, but the time flows
quickly for everyone, and perhaps some of them have also a real life to
live. :-)
So if the problem is the time I think a good thing is to try to lesser the
load of the committer.
And where is the main time waste a committer spend when trying to commit a
port ?
To correct the port file, to check the PLIST, to see if the {$PREFIX} are
right.
Well after reading a message dealing about all this process I have the idea
from which this "madness" is born.
Infact I thought that all this operations are not so difficult and a
"trusted" users can do it easily.
So this is my idea:
What about to create a (some) groups of trusted users who have the task to
filter the ports submitted and test them before the committer? If the port
is wrong they contact the author and then until the problem is not fixed
they don't give the ok to the committer. So the committer has not to deal
about wrong things, because when he received the "checked port file" he
knows it was analyzed by others persons. Surely he can try a "make install"
to see if everything is ok, but if the "user-teams" have done a good job
there will be nothing more to do than committing.
I think the trusted user groups can also be divided in some areas of
interests/knowledge like the port tree is : games, x11, sysutils and so on...
Ok, this is a crazy idea, but I think it can have some benefits.
1) More people get involved in FreeBSD project, even if we don't do any
"bazaar". If a person feels to partecipate in the project itself and knows
he is make an usefull jobs is surely happier and is motivated to produce :-)
2) We can speed up the committing time and lesser the load on each committer.
3) We enlarge the test base on different hardware/software configuration.
3) We could have more ports :-). If I know my ports will be committed soon,
I am motivated to send them and not to leave them in my HD. The succes of
an OS don't depend only on how much reliable is (look the win family :-)
but on how many applications can (run|be found) on it.
Obviusly it has also some drawbacks .
1) We need to find a bunch of "voluntaries" trusted users, perhaps a
coordinator.
2) We have to organize the thing better I have proposed :-)
I think I have annoyed too much for this time :-)
Thanks for attention.
Any comment is welcome.
Best Regards,
Gianmarco Giovannelli , "Unix expert since yesterday"
http://www.giovannelli.it/~gmarco
http://www2.masternet.it
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-ports" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4.1.19990301131907.009936f0>