[mythtv] Request: MythHelp

Stuart Auchterlonie stuarta at squashedfrog.net
Wed Feb 7 11:59:23 UTC 2007

usleepless at gmail.com wrote:
> Stuart, List,
> On 2/7/07, Stuart Auchterlonie <stuarta at squashedfrog.net> wrote:
>>> it wouldn't be so bad if you guys actually knew proper database design and
>> ipc.
>>> instead time is wasted on "hickups" because a lousy database is used,
>>> with a lousy schema. threads are created to work around the problems
>>> that arise. ipc is done with ordinary variables and busywait loops (
>>> while(!done) usleep(50); is just a context-switch banger ). calling
>>> query.isActive() && query.size()>0 && query.next() is just plain
>>> overkill.
>> Your patch to fix this issue is missing....
> so now there is at least acknowledgement this is actual an issue?

No, it's the first i've heard of it...

> well, let's call that progress.
> they are not missing in my 0.18-build. i would be happy to supply you
> with a couple of hundred patches against 0.18-fixes.

Generally accepted procedure for bug handling.
Step 1. Report bug with information on how to reproduce.
Step 2. Provide a patch for *review* that fixes the problem.
Step 3. Dev's review issue & proposed fix and work out if work
is required or the problem should be addressed in a different way etc.

These patches would have been welcome when 0.18 was current....
Now they clearly need a rewrite...

I would even be happy to sit down and review them with a view to
submitting them provided they were done in the correct way.
(1 bug, 1 patch)

> i has been made clear to me that for the "developers" making mythtv
> multi-db-compatible is out of the question.

Some don't want it, some don't mind. The boss doesn't want it.
There are plenty of other things that need doing first.

> then why on earth would i contribute such patches? please answer that.

Because instead of fixing bugs in the appropriate way, you have gone
straight to an enhancement that was generally accepted as not required
at that point in time.

Ever heard of making it work first, then adding features?
Or do you prefer the Microsoft way of features before bug fixes?


More information about the mythtv-dev mailing list