I disagree. I have personally fixed up a number of nearly complete
patches posted here and finalised them. Probably without the initial
legwork being done I wouldn't have done so.
Heh. And, a lot of those patches that you fixed up were ones I rejected for
various reasons and the original author didn't feel like fixing.
Yep. Agreed. That was my point though. They provided a good starting
point, and a clue to the right location in the source.
I think because you are so familiar with the source that you
overestimate the issues for us causal bug fixers in dipping in and
fixing a few things because it needs a bit of digging around to find the
right location. The current programing style uses inheritance, threads
and lots of callbacks /synchronisation (which is very sensible and I
have absolutely no problems with it), but it can make it a little
difficult to follow the code short of literally stepping it through the
debugger (which is sometimes hard to setup if it's time critical and
works differently when you step the code)
A real example might be that I wrote a small bit of code to record
whether a show had been watched or not. It's been in my tree for a year
or two because I just haven't had the time to figure out how to use to
do something useful like colour in the programs in some different
colour. I guess however, if I had posted it (and I would accept that
it's technically unfinished from one point of view), then perhaps
someone else would have finished it off and we would have this feature
in now...? (Which actually I must do because I would quite like this!!)
Why bother even attaching a patch if it's completely untested?
As above. It sometimes helps the rest of us get started fixing the
problem "properly". Agreed it's of little interest to you, except as
perhaps a barometer of where people are spending their time
For _me_ to
test if it's fixed, I'll have to go find a bttv card, open up my dev box,
stick it in, configure it, etc, and test it. Strangely enough, that's
exactly what I'd have to do if there wasn't a patch attached. Figure all
that would take an hour or so. The problem sounds relatively simple, so
debugging it from scratch would doubtfully add much time at all, either.
If the patch were tested, all I'd have to do would be to look it over to make
sure it didn't break anything else. That'd take about a minute, maybe 5 at
the most.
Which would you rather do?
Agreed - If I were you I would definitely want to do the latter. I
totally agree with that, and as such I think you did the right thing to
reject the patch. In fact I would accept a further generalisation that
all patches should occupy a max of a few mins attention in order to be
deemed good enough to check in.
All I was doing was *very* gently trying to suggest that the exact way
that one rejects a patch may influence the chance of a new patch
appearing... I know I have submitted some complete bollocks to other
projects, but they are usually rejected with some gentle suggestions on
how to meet the required standard. Written english is also read very
differently by different cultures (tell me about it, I am British and we
are regularly misunderstood by even Americans!)
Anyway, I was broadly agreeing with you. And my points were vague and
not targetted at you either - there are some others on the list who I
feel are a little harsh to newbies getting started.
If I may rephrase your original email:
"Álvaro, Thanks for your patch. Well done for tracking down this issue. *I* (Ed)
certainly don't have a bttv card, and Isaac doesn't have one handy. Before checking this
into the main code base could you please test it and *verify* that it does in fact fix
the issue. If anyone else has a BTTV card who could test this and repro the original
problem then please let us know your results"
I think it's fair enough that the main developers don't want to spend more than
1-2 mins maximum on any patch submitted, so the submitter needs to do all the
main legwork.
All the best
Ed W
_______________________________________________
mythtv-dev mailing list
[EMAIL PROTECTED]
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
