libcmyth and XBMC

[Daniel Cleaver]

Hi all,
I'm not sure whether you guys are aware or not that XBMC
(http://xbmc.org) currently use the libcmyth library?

I'm looking at expanding on XBMC's use of MythTV (and therefore the
library) to provide full PVR functionality (currently only basic
watching of live and recorded shows is possible, but no EPG).

My question is -- what's the best way to go about adding functionality
to libcmyth for XBMC's needs? Branch at the current trunk revision?
Would you guys be willing to allow merges back? Or only if they suited
MediaMVP? In which case XBMC would have to maintain its own fork with
merges being fed into it from the original libcmyth?

Excuse my ignorance on such matters -- I'm new to this open source
development lark. Figured I've used it for so long, might as well get my
hands dirty :)

Regards,
Daniel Cleaver


------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[gettler]

On Wed, 14 Oct 2009 23:35:17 BST, Daniel Cleaver wrote:

> Hi all,
> I'm not sure whether you guys are aware or not that XBMC
> (http://xbmc.org) currently use the libcmyth library?
>
> I'm looking at expanding on XBMC's use of MythTV (and therefore the
> library) to provide full PVR functionality (currently only basic
> watching of live and recorded shows is possible, but no EPG).
>
> My question is -- what's the best way to go about adding functionality
> to libcmyth for XBMC's needs? Branch at the current trunk revision?
> Would you guys be willing to allow merges back? Or only if they suited
> MediaMVP? In which case XBMC would have to maintain its own fork with
> merges being fed into it from the original libcmyth?
>
> Excuse my ignorance on such matters -- I'm new to this open source
> development lark. Figured I've used it for so long, might as well get my
> hands dirty :)
>
> Regards,
> Daniel Cleaver

This has come up before.

The correct solution, really, is for us to create a standalone git repo
that contains libcmyth only.  Then development could proceed there by
everyone who is interested in maintaining it (from mvpmc, xbmc, ...).

I tried splitting the code apart into a seperate repository a while back,
but never quite got git to retain the development history correctly.

Would this work for everyone?  I can make sure that those interested have
write access to the repo.

Jon


------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[jm_maclaren]

From: gettler@... [mailto:gettler@...]
Sent: Wednesday, October 14, 2009 6:35 PM
To: Daniel Cleaver
Cc: mvpmc-devel@...
Subject: Re: [Mvpmc-devel] libcmyth and XBMC


On Wed, 14 Oct 2009 23:35:17 BST, Daniel Cleaver wrote:

> Hi all,
> I'm not sure whether you guys are aware or not that XBMC
> (http://xbmc.org) currently use the libcmyth library?
>
> I'm looking at expanding on XBMC's use of MythTV (and therefore the
> library) to provide full PVR functionality (currently only basic
> watching of live and recorded shows is possible, but no EPG).
>
> My question is -- what's the best way to go about adding functionality
> to libcmyth for XBMC's needs? Branch at the current trunk revision?
> Would you guys be willing to allow merges back? Or only if they suited
> MediaMVP? In which case XBMC would have to maintain its own fork with
> merges being fed into it from the original libcmyth?
>
> Excuse my ignorance on such matters -- I'm new to this open source
> development lark. Figured I've used it for so long, might as well get

my
> hands dirty :)
>
> Regards,
> Daniel Cleaver

This has come up before.

The correct solution, really, is for us to create a standalone git repo
that contains libcmyth only.  Then development could proceed there by
everyone who is interested in maintaining it (from mvpmc, xbmc, ...).

I tried splitting the code apart into a seperate repository a while
back,
but never quite got git to retain the development history correctly.

Would this work for everyone?  I can make sure that those interested
have
write access to the repo.

Jon


------------------------------------------------------------------------
------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and
stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel



Jon
This would be great.  I use libcmyth, probably an old version too, on
the roku hd1000 in my mythroku application that is still used by a few
remaining luddites without one of the fancy new media streamers such as
the popcorn hour!
James

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[MVallevand]

On Thu, Oct 15, 2009 at 2:01 AM, Maclaren, James M <maclaren@...> wrote:
> This would be great.  I use libcmyth, probably an old version too, on
> the roku hd1000 in my mythroku application that is still used by a few
> remaining luddites without one of the fancy new media streamers such as
> the popcorn hour!
> James

James what you did on the roku hd1000 is excellent and far fancier
than anything done for MythTV on the Popcorn Hour.

Martin

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[Eric Lund]

Jon, James, Martin,

Splitting libcmyth out would be fine with me.  I think it is a great
idea.  It is nice to know that that code lives on in other projects.

Speaking of the popcorn hour, I finally upgraded my A110 and it stopped
talking to myth altogether.  I had it running as a UPNP client of Myth,
but it malforms (or Myth misinterprets) URLs in the latest firmare.  I
am not sure this is a bad thing, though.  I upgraded myth and swtiched
over to using NFS and mythrename (the nifty little PERL script that
makes sensible filenames out of Myth data) and I don't think I will go
back.

That said, I would like another player, and I am not sure the PH is
where I want to go at this point.  Is the Roku a good direction?  What
Roku products are supported my MythRoku? My MediaMVPs are all out of
commission at this point, since they can't handle the bandwidth.

                                        Eric

MVallevand wrote:
 > On Thu, Oct 15, 2009 at 2:01 AM, Maclaren, James M <maclaren@...> wrote:
 >> This would be great.  I use libcmyth, probably an old version too, on
 >> the roku hd1000 in my mythroku application that is still used by a few
 >> remaining luddites without one of the fancy new media streamers such as
 >> the popcorn hour!
 >> James
 >
 > James what you did on the roku hd1000 is excellent and far fancier
 > than anything done for MythTV on the Popcorn Hour.
 >
 > Martin
 >
 > ------------------------------------------------------------------------------
 > Come build with us! The BlackBerry(R) Developer Conference in SF, CA
 > is the only developer event you need to attend this year. Jumpstart your
 > developing skills, take BlackBerry mobile applications to market and stay
 > ahead of the curve. Join us from November 9 - 12, 2009. Register now!
 > http://p.sf.net/sfu/devconference
 > _______________________________________________
 > Mvpmc-devel mailing list
 > Mvpmc-devel@...
 > https://lists.sourceforge.net/lists/listinfo/mvpmc-devel
 >

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[MVallevand]

On Thu, Oct 15, 2009 at 10:34 AM, Eric Lund <eric@...> wrote:
> That said, I would like another player, and I am not sure the PH is
> where I want to go at this point.  Is the Roku a good direction?  What

Personally I think we just need to devote a month to get the PCH's and
MythTV working and everyone will be pleasantly surprised.  You can't
compare what you see with the basic Syabas player, with what GBPVR
users are seeing.  Playing mpeg-ts and mpeg-ps files  of reasonable
quality (ie not too many severe internal errors) they get livetv with
2 second channel ranges, comskip, navigation forwards and backward,
controllable OSD over video, on all broadcast tv captures that I have
found including HD-PVR format.  Music and video playback starts with
MVP like speed and you can have simultaneous audio  playback with
control of the UI.

The only real limitations now are  AC3 pass through is not working,
the firmware lacks support for AAC HE-LATM used some countries, and
there is no DVB teletext or graphic subtitles.   Also closed
captioning from analog and digital sources in N.A. is only working on
composite and svideo but the HD does play downconverted.

I don't think the Roku supports h264 so it is limited to North
America.    For non-mpeg files the PCH player does a reasonable job
and it  is much better than the Roku.

Martin

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[Daniel Cleaver]

Jon,
That sounds great. I guess you'll want to enforce some rules though
about what people can commit? Particularly to prevent breaking mvpmc's
myth functionality.

Or were you thinking of branching it and only merging back when mvpmc is
happy with the changes?

Daniel

On Wed, 2009-10-14 at 18:35 -0500, gettler@... wrote:

> On Wed, 14 Oct 2009 23:35:17 BST, Daniel Cleaver wrote:
> > Hi all,
> > I'm not sure whether you guys are aware or not that XBMC
> > (http://xbmc.org) currently use the libcmyth library?
> >
> > I'm looking at expanding on XBMC's use of MythTV (and therefore the
> > library) to provide full PVR functionality (currently only basic
> > watching of live and recorded shows is possible, but no EPG).
> >
> > My question is -- what's the best way to go about adding functionality
> > to libcmyth for XBMC's needs? Branch at the current trunk revision?
> > Would you guys be willing to allow merges back? Or only if they suited
> > MediaMVP? In which case XBMC would have to maintain its own fork with
> > merges being fed into it from the original libcmyth?
> >
> > Excuse my ignorance on such matters -- I'm new to this open source
> > development lark. Figured I've used it for so long, might as well get my
> > hands dirty :)
> >
> > Regards,
> > Daniel Cleaver
>
> This has come up before.
>
> The correct solution, really, is for us to create a standalone git repo
> that contains libcmyth only.  Then development could proceed there by
> everyone who is interested in maintaining it (from mvpmc, xbmc, ...).
>
> I tried splitting the code apart into a seperate repository a while back,
> but never quite got git to retain the development history correctly.
>
> Would this work for everyone?  I can make sure that those interested have
> write access to the repo.
>
> Jon
>

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[gettler]

There are several options.

We could have a single development stream and have the mvpmc.org server
generate tarballs on days when the code changes.  Each project could then
track the library as they see fit.

Or each project could have their own development stream within the libcmyth
repository, and they would build from their own tip.  This would require
the developers to merge their own changes back and forth between a master
tip, but each project would be in control of what they are releasing.

Also, we'll want do official releases of the library every so often.

Jon

On Thu, 15 Oct 2009 19:45:08 BST, Daniel Cleaver wrote:

> Jon,
> That sounds great. I guess you'll want to enforce some rules though
> about what people can commit? Particularly to prevent breaking mvpmc's
> myth functionality.
>
> Or were you thinking of branching it and only merging back when mvpmc is
> happy with the changes?
>
> Daniel
>
> On Wed, 2009-10-14 at 18:35 -0500, gettler@... wrote:
> > On Wed, 14 Oct 2009 23:35:17 BST, Daniel Cleaver wrote:
> > > Hi all,
> > > I'm not sure whether you guys are aware or not that XBMC
> > > (http://xbmc.org) currently use the libcmyth library?
> > >
> > > I'm looking at expanding on XBMC's use of MythTV (and therefore the
> > > library) to provide full PVR functionality (currently only basic
> > > watching of live and recorded shows is possible, but no EPG).
> > >
> > > My question is -- what's the best way to go about adding functionality
> > > to libcmyth for XBMC's needs? Branch at the current trunk revision?
> > > Would you guys be willing to allow merges back? Or only if they suited
> > > MediaMVP? In which case XBMC would have to maintain its own fork with
> > > merges being fed into it from the original libcmyth?
> > >
> > > Excuse my ignorance on such matters -- I'm new to this open source
> > > development lark. Figured I've used it for so long, might as well get my
> > > hands dirty :)
> > >
> > > Regards,
> > > Daniel Cleaver
> >
> > This has come up before.
> >
> > The correct solution, really, is for us to create a standalone git repo
> > that contains libcmyth only.  Then development could proceed there by
> > everyone who is interested in maintaining it (from mvpmc, xbmc, ...).
> >
> > I tried splitting the code apart into a seperate repository a while back,
> > but never quite got git to retain the development history correctly.
> >
> > Would this work for everyone?  I can make sure that those interested have
> > write access to the repo.
> >
> > Jon
> >

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[gettler]

On Thu, 15 Oct 2009 09:34:33 CDT, Eric Lund wrote:
>
> Splitting libcmyth out would be fine with me.  I think it is a great
> idea.  It is nice to know that that code lives on in other projects.
>

I've taken the first stab at this.  The problems I had rewriting history
seem to have gone away with git 1.6.5.

So have a look at http://git.mvpmc.org/ and the cmyth.git repository.  I
rewrote the history from HEAD of mvpmc.git to remove everything that was not
needed by an app that wants to use libcmyth.  Then I made it compile on my
PC, and added my mythfuse app as an example (it creates a filesystem for
your myth server).

For everyone interested in libcmyth, please take a look at this.  If I've
missed anything, I would prefer to fix it now by recreating the repo, rather
than after people start using it for real.

FYI, it is read-only at the moment, so you won't be able to push into it.

Thanks.

Jon


------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[Daniel Cleaver]

Wow! That was fast!

I've managed to build the libraries fine. The mythfuse example doesn't
compile though. Scratch that -- just went to check the exact error and I
don't have the fuse headers installed :S

I'll let you know how I get on with using the library. Still in the
early stages though.

Thanks,
Daniel

On Thu, 2009-10-15 at 20:29 -0500, gettler@... wrote:

> On Thu, 15 Oct 2009 09:34:33 CDT, Eric Lund wrote:
> >
> > Splitting libcmyth out would be fine with me.  I think it is a great
> > idea.  It is nice to know that that code lives on in other projects.
> >
>
> I've taken the first stab at this.  The problems I had rewriting history
> seem to have gone away with git 1.6.5.
>
> So have a look at http://git.mvpmc.org/ and the cmyth.git repository.  I
> rewrote the history from HEAD of mvpmc.git to remove everything that was not
> needed by an app that wants to use libcmyth.  Then I made it compile on my
> PC, and added my mythfuse app as an example (it creates a filesystem for
> your myth server).
>
> For everyone interested in libcmyth, please take a look at this.  If I've
> missed anything, I would prefer to fix it now by recreating the repo, rather
> than after people start using it for real.
>
> FYI, it is read-only at the moment, so you won't be able to push into it.
>
> Thanks.
>
> Jon
>
>
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Mvpmc-devel mailing list
> Mvpmc-devel@...
> https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[dteirney ()]

Would be great to have a separate libcmyth project that is contributed to by all those using it.

I've also done some development of libcmyth for use with XBMC Media Center. For example, finished the work already started to get commercial breaks from the backend using the Myth Protocol rather than the MySQL query. Also a patch to support 0.22 over and above what was added in the mvpmc repository.

There are a number of other changes within XBMC to support building with the Microsoft C compiler, which is very pedantic, e.g. all variables must be declared at the top of the method. Lots of #ifdef _MSC_VER around the place for other reasons.

I don't know the reasons for all of the differences between the XBMC version and the current libcmyth trunk, but I might be able to get some time off elupus to provide some information about the changes that aren't obvious.

What is the next step to try and push back some of the changes that have been made externally to the mvpmc repository? I've got the git version on my machine and can start sending patches with rationale.

Thanks!

David

[gettler]

I think the next step is for those involved to agree that the current
cmyth.git repo is a good starting point.  Then we can start taking patches
from mvpmc and XBMC so that it is buildable by both projects (as well as
standalone).

Jon

On Mon, 19 Oct 2009 00:00:56 PDT, dteirney wrote:

>
> Would be great to have a separate libcmyth project that is contributed to by
> all those using it.
>
> I've also done some development of libcmyth for use with XBMC Media Center.
> For example, finished the work already started to get commercial breaks from
> the backend using the Myth Protocol rather than the MySQL query. Also a
> patch to support 0.22 over and above what was added in the mvpmc repository.
>
> There are a number of other changes within XBMC to support building with the
> Microsoft C compiler, which is very pedantic, e.g. all variables must be
> declared at the top of the method. Lots of #ifdef _MSC_VER around the place
> for other reasons.
>
> I don't know the reasons for all of the differences between the XBMC version
> and the current libcmyth trunk, but I might be able to get some time off
> elupus to provide some information about the changes that aren't obvious.
>
> What is the next step to try and push back some of the changes that have
> been made externally to the mvpmc repository? I've got the git version on my
> machine and can start sending patches with rationale.
>
> Thanks!
>
> David
> --
> View this message in context: http://www.nabble.com/libcmyth-and-XBMC-tp25900
>674s24862p25944258.html
> Sent from the MediaMVP Media Center - Development mailing list archive at Nab
>ble.com.

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[Eric Lund]

David,

The pedantic compiler issues don't seem to me like they should need to
be ifdef'ed.  If there is a code generation issue with the Microsoft
compiler, then, by all means, protect it, but I, for one, hate excess
ifdef's, so if you are just making the code work well for a stricter
compiler, by all make the changes in the common code.

That said, is it that the variables have to be declared at the top of the
function or as the first thing in a delineated block?  Most C compilers are
just fine with:

int foo()
{
    ...
    if (a == b) {
       int i;
       ...
    }
    ...
}

In fact, I believe this has been a documented legal syntax in C since the
original K+R book, and it places the scope of 'i' in the inner block. Many
true C compilers (as opposed to C++ compilers doing double duty as C
compilers) will have a problem with:

int foo()
{
    ...
    if (a == b) {
       a++;
       int i;
       ...
    }
    ...
}

So, if you have a choice of moving an inlined definition to the top of the
function versus moving it to the top of the block in which it is defined,
and the compiler will let you, please move it to the top of the block to
avoid any possible changes in scoping that might bite you.

Jon, I didn't get around to trying out the myth fuse over the weekend.
maybe tonight.  I now have an 0.21 myth server (I was running 0.20 up
until last week) so we'll see if it still works with 0.21.

                                        Eric

gettler@... wrote:
 > I think the next step is for those involved to agree that the current
 > cmyth.git repo is a good starting point.  Then we can start taking patches
 > from mvpmc and XBMC so that it is buildable by both projects (as well as
 > standalone).
 >
 > Jon
 >
 > On Mon, 19 Oct 2009 00:00:56 PDT, dteirney wrote:
 >> Would be great to have a separate libcmyth project that is contributed to by
 >> all those using it.
 >>
 >> I've also done some development of libcmyth for use with XBMC Media Center.
 >> For example, finished the work already started to get commercial breaks from
 >> the backend using the Myth Protocol rather than the MySQL query. Also a
 >> patch to support 0.22 over and above what was added in the mvpmc repository.
 >>
 >> There are a number of other changes within XBMC to support building with the
 >> Microsoft C compiler, which is very pedantic, e.g. all variables must be
 >> declared at the top of the method. Lots of #ifdef _MSC_VER around the place
 >> for other reasons.
 >>
 >> I don't know the reasons for all of the differences between the XBMC version
 >> and the current libcmyth trunk, but I might be able to get some time off
 >> elupus to provide some information about the changes that aren't obvious.
 >>
 >> What is the next step to try and push back some of the changes that have
 >> been made externally to the mvpmc repository? I've got the git version on my
 >> machine and can start sending patches with rationale.
 >>
 >> Thanks!
 >>
 >> David
 >> --
 >> View this message in context: http://www.nabble.com/libcmyth-and-XBMC-tp25900
 >> 674s24862p25944258.html
 >> Sent from the MediaMVP Media Center - Development mailing list archive at Nab
 >> ble.com.
 >
 > ------------------------------------------------------------------------------
 > Come build with us! The BlackBerry(R) Developer Conference in SF, CA
 > is the only developer event you need to attend this year. Jumpstart your
 > developing skills, take BlackBerry mobile applications to market and stay
 > ahead of the curve. Join us from November 9 - 12, 2009. Register now!
 > http://p.sf.net/sfu/devconference
 > _______________________________________________
 > Mvpmc-devel mailing list
 > Mvpmc-devel@...
 > https://lists.sourceforge.net/lists/listinfo/mvpmc-devel
 >

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[Eric Lund]

As one of those involved, I think it is probably ready for prime time,
at least from the perspective of appearing to work.  Anyone else?

                                        Eric

On 10/19/2009 07:37 AM, gettler@... wrote:

> I think the next step is for those involved to agree that the current
> cmyth.git repo is a good starting point.  Then we can start taking patches
> from mvpmc and XBMC so that it is buildable by both projects (as well as
> standalone).
>
> Jon
>
> On Mon, 19 Oct 2009 00:00:56 PDT, dteirney wrote:
>> Would be great to have a separate libcmyth project that is contributed to by
>> all those using it.
>>
>> I've also done some development of libcmyth for use with XBMC Media Center.
>> For example, finished the work already started to get commercial breaks from
>> the backend using the Myth Protocol rather than the MySQL query. Also a
>> patch to support 0.22 over and above what was added in the mvpmc repository.
>>
>> There are a number of other changes within XBMC to support building with the
>> Microsoft C compiler, which is very pedantic, e.g. all variables must be
>> declared at the top of the method. Lots of #ifdef _MSC_VER around the place
>> for other reasons.
>>
>> I don't know the reasons for all of the differences between the XBMC version
>> and the current libcmyth trunk, but I might be able to get some time off
>> elupus to provide some information about the changes that aren't obvious.
>>
>> What is the next step to try and push back some of the changes that have
>> been made externally to the mvpmc repository? I've got the git version on my
>> machine and can start sending patches with rationale.
>>
>> Thanks!
>>
>> David
>> --
>> View this message in context: http://www.nabble.com/libcmyth-and-XBMC-tp25900
>> 674s24862p25944258.html
>> Sent from the MediaMVP Media Center - Development mailing list archive at Nab
>> ble.com.
>
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Mvpmc-devel mailing list
> Mvpmc-devel@...
> https://lists.sourceforge.net/lists/listinfo/mvpmc-devel
>

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[dteirney]

Eric, you are right. The variables need to be defined at the top of the block they are scoped within. This is what has been done. The #ifdef statements are not used to change the location of the variable declaration. That would be madness.

The #ifdef/#ifndef statements are generally used to include platform specific headers. For example,

#ifndef _MSC_VER
#include <unistd.h>
#endif
#include <stdio.h>
#ifdef _MSC_VER
#include <winsock2.h>
#else
#include <sys/socket.h>
#include <netinet/in.h>
#include <netdb.h>
#include <arpa/inet.h>
#endif

They are also used to remove blocks of code, for example.

#ifndef _MSC_VER
        void (*old_sighandler)(int);
        int old_alarm;
#endif

Eric Lund wrote:

David,

The pedantic compiler issues don't seem to me like they should need to
be ifdef'ed.  If there is a code generation issue with the Microsoft
compiler, then, by all means, protect it, but I, for one, hate excess
ifdef's, so if you are just making the code work well for a stricter
compiler, by all make the changes in the common code.

That said, is it that the variables have to be declared at the top of the
function or as the first thing in a delineated block?  Most C compilers are
just fine with:

int foo()
{
    ...
    if (a == b) {
       int i;
       ...
    }
    ...
}

In fact, I believe this has been a documented legal syntax in C since the
original K+R book, and it places the scope of 'i' in the inner block. Many
true C compilers (as opposed to C++ compilers doing double duty as C
compilers) will have a problem with:

int foo()
{
    ...
    if (a == b) {
       a++;
       int i;
       ...
    }
    ...
}

So, if you have a choice of moving an inlined definition to the top of the
function versus moving it to the top of the block in which it is defined,
and the compiler will let you, please move it to the top of the block to
avoid any possible changes in scoping that might bite you.

                                        Eric

[gettler]

On Wed, 21 Oct 2009 02:59:40 PDT, dteirney wrote:

>
> Eric, you are right. The variables need to be defined at the top of the block
> they are scoped within. This is what has been done. The #ifdef statements
> are not used to change the location of the variable declaration. That would
> be madness.
>
> The #ifdef/#ifndef statements are generally used to include platform
> specific headers. For example,
>
> #ifndef _MSC_VER
> #include <unistd.h>
> #endif
> #include <stdio.h>
> #ifdef _MSC_VER
> #include <winsock2.h>
> #else
> #include <sys/socket.h>
> #include <netinet/in.h>
> #include <netdb.h>
> #include <arpa/inet.h>
> #endif

I think the above is a good use of #ifdef _MSC_VER

> They are also used to remove blocks of code, for example.
>
> #ifndef _MSC_VER
> void (*old_sighandler)(int);
> int old_alarm;
> #endif

I think we should change this type of thing, however, to some sort of
feature ifdef.  That way users of the library can easily turn off features
they may not want or cannot support.

Jon


------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[Daniel Cleaver]

Hi Jon,
I notice you've made quite a few changes to the source recently. I've
just updated my local copy, but I'm now getting the following error when
trying to compile:

ar rc libcmyth/libcmyth.a libcmyth/bookmark.o libcmyth/connection.o
libcmyth/debug.o libcmyth/event.o libcmyth/file.o libcmyth/freespace.o
libcmyth/keyframe.o libcmyth/posmap.o libcmyth/proginfo.o
libcmyth/proglist.o libcmyth/rec_num.o libcmyth/recorder.o
libcmyth/ringbuf.o libcmyth/socket.o libcmyth/timestamp.o
libcmyth/livetv.o libcmyth/commbreak.o libcmyth/mythtv_mysql.o
libcmyth/mysql_query.o
ranlib libcmyth/libcmyth.a
scons: *** [libcmyth/libcmyth.so.0.1] AttributeError : 'NoneType' object
has no attribute 'genstring'
scons: building terminated because of errors.

This is from a clean clone of the repo. It looks as though it's an scons
problem, but I'm not too hot with scons yet. Any ideas what's going on?

Cheers,
Daniel



------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[gettler]

On Wed, 28 Oct 2009 16:56:04 -0000, Daniel Cleaver wrote:

> Hi Jon,
> I notice you've made quite a few changes to the source recently. I've
> just updated my local copy, but I'm now getting the following error when
> trying to compile:
>
> ar rc libcmyth/libcmyth.a libcmyth/bookmark.o libcmyth/connection.o
> libcmyth/debug.o libcmyth/event.o libcmyth/file.o libcmyth/freespace.o
> libcmyth/keyframe.o libcmyth/posmap.o libcmyth/proginfo.o
> libcmyth/proglist.o libcmyth/rec_num.o libcmyth/recorder.o
> libcmyth/ringbuf.o libcmyth/socket.o libcmyth/timestamp.o
> libcmyth/livetv.o libcmyth/commbreak.o libcmyth/mythtv_mysql.o
> libcmyth/mysql_query.o
> ranlib libcmyth/libcmyth.a
> scons: *** [libcmyth/libcmyth.so.0.1] AttributeError : 'NoneType' object
> has no attribute 'genstring'
> scons: building terminated because of errors.
>
> This is from a clean clone of the repo. It looks as though it's an scons
> problem, but I'm not too hot with scons yet. Any ideas what's going on?
>
> Cheers,
> Daniel
>

That does seem odd.  With scons v0.97.0d20071203.r2509, it all builds
just fine.  But with scons v1.2.0.r3842 I get the same error you do.

I'll look into it...

Jon


------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[gettler]

On Wed, 28 Oct 2009 12:10:38 CDT, gettler@... wrote:

>
> On Wed, 28 Oct 2009 16:56:04 -0000, Daniel Cleaver wrote:
> > Hi Jon,
> > I notice you've made quite a few changes to the source recently. I've
> > just updated my local copy, but I'm now getting the following error when
> > trying to compile:
> >
> > ar rc libcmyth/libcmyth.a libcmyth/bookmark.o libcmyth/connection.o
> > libcmyth/debug.o libcmyth/event.o libcmyth/file.o libcmyth/freespace.o
> > libcmyth/keyframe.o libcmyth/posmap.o libcmyth/proginfo.o
> > libcmyth/proglist.o libcmyth/rec_num.o libcmyth/recorder.o
> > libcmyth/ringbuf.o libcmyth/socket.o libcmyth/timestamp.o
> > libcmyth/livetv.o libcmyth/commbreak.o libcmyth/mythtv_mysql.o
> > libcmyth/mysql_query.o
> > ranlib libcmyth/libcmyth.a
> > scons: *** [libcmyth/libcmyth.so.0.1] AttributeError : 'NoneType' object
> > has no attribute 'genstring'
> > scons: building terminated because of errors.
> >
> > This is from a clean clone of the repo. It looks as though it's an scons
> > problem, but I'm not too hot with scons yet. Any ideas what's going on?
> >
> > Cheers,
> > Daniel
> >
>
> That does seem odd.  With scons v0.97.0d20071203.r2509, it all builds
> just fine.  But with scons v1.2.0.r3842 I get the same error you do.
>
> I'll look into it...
>
> Jon

I pushed up a fix which seems to work for both versions of scons that
I tried (v0.97.0d20071203.r2509 and v1.2.0.r3842).

Jon


------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel

[Daniel Cleaver]

On Wed, 2009-10-28 at 13:02 -0500, gettler@... wrote:

> On Wed, 28 Oct 2009 12:10:38 CDT, gettler@... wrote:
> >
> > On Wed, 28 Oct 2009 16:56:04 -0000, Daniel Cleaver wrote:
> > > Hi Jon,
> > > I notice you've made quite a few changes to the source recently. I've
> > > just updated my local copy, but I'm now getting the following error when
> > > trying to compile:
> > >
> > > ar rc libcmyth/libcmyth.a libcmyth/bookmark.o libcmyth/connection.o
> > > libcmyth/debug.o libcmyth/event.o libcmyth/file.o libcmyth/freespace.o
> > > libcmyth/keyframe.o libcmyth/posmap.o libcmyth/proginfo.o
> > > libcmyth/proglist.o libcmyth/rec_num.o libcmyth/recorder.o
> > > libcmyth/ringbuf.o libcmyth/socket.o libcmyth/timestamp.o
> > > libcmyth/livetv.o libcmyth/commbreak.o libcmyth/mythtv_mysql.o
> > > libcmyth/mysql_query.o
> > > ranlib libcmyth/libcmyth.a
> > > scons: *** [libcmyth/libcmyth.so.0.1] AttributeError : 'NoneType' object
> > > has no attribute 'genstring'
> > > scons: building terminated because of errors.
> > >
> > > This is from a clean clone of the repo. It looks as though it's an scons
> > > problem, but I'm not too hot with scons yet. Any ideas what's going on?
> > >
> > > Cheers,
> > > Daniel
> > >
> >
> > That does seem odd.  With scons v0.97.0d20071203.r2509, it all builds
> > just fine.  But with scons v1.2.0.r3842 I get the same error you do.
> >
> > I'll look into it...
> >
> > Jon
>
> I pushed up a fix which seems to work for both versions of scons that
> I tried (v0.97.0d20071203.r2509 and v1.2.0.r3842).
>
> Jon
>

Thanks for that. Works fine now with my v1.2.0.r3842.

Daniel



------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Mvpmc-devel mailing list
Mvpmc-devel@...
https://lists.sourceforge.net/lists/listinfo/mvpmc-devel