[mythtv] Status Question: Is current SVN useable?

Kevin Fox Kevin.Fox at pnl.gov
Tue Jan 22 00:50:57 UTC 2008


But, as previously mentioned, if its more stable, has less bugs, then
why not release it. Its not worse then what people are already running.
Getting more features out quicker may get you more developers to help
fix the remaining bugs.

I think that's part of the problem though. Going to SVN is not always an
option. I've considered submitting patches, but since I am on the stable
version, I'm not very encouraged to submit to something that's already
horribly out of date. I'm probably not the only one either. Release
Early, Release Often(1) has a lot of benefits I think Myth misses out
from. As far as documentation getting a little out of date, I can live
with that. I'm use to that from tons of other open source projects. In a
lot of projects, I consider any documentation at all a bonus. Having
reasonably current binaries in distro's that track fairly closely to
development is a much better thing then up to date documentation. It
means I have a reasonably good chance of submitting a bug fix or feature
myself based on what code I already have, or be able to talk to a
developer that says something other then "Oh, we know about that, we
fixed that in SVN 4 months ago, please wait a few months till we release
and it gets into your distro" kinds of things. (Not heard specifically
from the Myth project. Just a wild example)

Most other open source projects went away from long development cycles
because of this.

Kevin

No ill will intended. Just intended as constructive criticism

1.
http://catb.org/~esr/writings/cathedral-bazaar/cathedral-bazaar/ar01s04.html

On Mon, 2008-01-21 at 15:09 -0500, David Backeberg wrote:
> Aside from the reasons already given, (deferring to Isaac, more bugs
> to fix, etc.) I also think the documentation isn't quite ready. When
> Myth makes a release, I believe there is usually a guide of the sort
> that came with .20.2 that mentioned the changes to expect and how to
> use them. People who aren't code hackers and bug smooshers can help
> out in that area. I know when I went from 20.2 to SVN I noticed tables
> added to database and some other things.
> 
> But honestly, I would personally love to see a release myself. I'm
> another long time user who wanted feature X and thus required jumping
> to SVN. One thing that I've wanted (aside from mythweather) is a more
> integrated Mythflix setup. I've decided that I should put up or shut
> up, and thus joined the dev list to see how hard it would be to just
> do this myself and submit a patch.
> 
> On Jan 21, 2008 2:50 PM, Kevin Fox <Kevin.Fox at pnl.gov> wrote:
> > So, if SVN is the most stable, and has a ton of bug fixes/new features,
> > why not just release it as is? Release Early, Release Often, has its
> > benefits. Why not make the 2 months from now release 0.22? What would it
> > hurt? It would allow people to just "apt-get update" (or whatever) all
> > the hard work that was added to Myth, that is out of reach of a lot of
> > people.
> >
> > Kevin
> >
> >
> > On Sun, 2008-01-20 at 21:28 +0000, Stuart Morgan wrote:
> > > On Sunday 20 January 2008 20:32:57 Reza Naima wrote:
> > > > But I am at a personal crossroadds and wanted advise in helping make a
> > > > decision as I've not followed the lists closely for ages and wanted to
> > > > see why the normal stream of newer versions stopped, and what expect
> > > > down the pipeline.
> > >
> > > There are multiple reasons why 0.21 has been so long in the making and I would
> > > probably struggle to find a single one on which everyone could agree.
> > >
> > > One of the major ones is that Isaac, the lead dev, has been too busy for the
> > > last few months to be actively involved in development. Since Isaac has
> > > always been the driving force behind setting target dates for new releases
> > > we've been a little slow to set a deadline. Many developers run trunk on
> > > their production machines (yes, it's that stable). Most of us have forgotten
> > > what 0.20 was like and we've no personal motivation to release a new version.
> > >
> > > Another reason is that whilst the project grows in size the number of active
> > > developers hasn't kept pace, if anything there might be less people making
> > > commits now than there was a year ago. That doesn't mean the project is
> > > stagnating, far from it, but there is a massive backlog of tickets to work
> > > through.
> > >
> > > Although 0.20 couldn't be described as feature complete, it was the first
> > > release where many of the must-have features had finally been implemented,
> > > it's quite natural once you reach that point for subsequent releases to be
> > > fewer and further between.
> > >
> > > There are simply too many changes, big and small since 0.20 to remember them
> > > all, but many are significant. SVN is probably the most stable version of
> > > mythtv we've ever had, even accounting for some of the recent instability
> > > caused by some major changes (backend auto-discovery and multirec). We're
> > > probably less than 2 months from a release now, we've set an arbitrary target
> > > in mid to late Feb but I should make clear that the date is NOT FINAL.
> > >
> >
> >
> > _______________________________________________
> > mythtv-dev mailing list
> > mythtv-dev at mythtv.org
> > http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
> >
> _______________________________________________
> mythtv-dev mailing list
> mythtv-dev at mythtv.org
> http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev



More information about the mythtv-dev mailing list