All Developers, please read.

[tschak909]

I wish I could. However, I would not be able to dedicate as much time to this project as would be needed in these first critical stages. I am currently on the verge of releasing my first research project as Free Software (Niki, a Wiki-web engine based on the results of my research on alternative user interfaces.), and I wouldn't want to let you guys down.

With that said, as soon as that project stabilises (should be a couple of months from now), I would be glad to lend a hand in coding, documentation, helping set up servers, whatever is needed of me.

-Thom

[Zaerc]

I'm sorry if I wasn't clear enough, but I think we're on the same page here.

You (Paul) have done a tremendous job so far.  And the last thing I want to do is to saddle you up with more work.  If it wasn't for your hard work and dedication we wouldn't be here today.  I didn't mean "we are on our own" in the sense that you have abandonned the project/community, but more like that we shouldn't wait for you to hand us everything on a silver platter (like you have been doing so far, which is pretty amazing in itself).

So the point I was trying to make is that it's time for the rest of us to get of our asses and see how can we help out, preferably in such a manner that will take as little effort from you as possible.

One of those things we can already do (without your help) is simply documenting the things we figure out.  The rest will sooner or later require at least some of your assistence so I guess we will need to find a way to reduce your workload one way or another to make some room for that in the mean time.

[bulek]

Hi,

I'll try to add few thoughts, since I've been working a lot also on Pluto system. Since I'm not experienced Linux guy, I could easily be missing something but anyway :
1. for a good start, I think working SVN and sqlcvs would be just fine
2. enhancement should be entered in coordinated way - For example under Pluto, where I helped a bit with Motion wrapper. I've contacted corresponding developers and explained what I'd like to add. Then he approved my work, I took down SVN source, compiled just new binary for Motion Wrapper and add it to my formal release. Then I tested it in my environment and then applied those changes into SVN - "official" testers did try it and it was in next release.
3. Same mechanism could be used for sqlcvs. I know this would narrow contributions, but again maybe better than nothing at all...

Currently we're not getting any enhancements into sql database or in source code, so maybe this would be improvement. And later, if community grows, we can take another steps.

What I'm currently also missing very badly is developers discussion... If you notice, I have several question regarding (how to best implement this, how to efficiently add this and to do some brainstorming on some future additions and how to implement them, or to learn important details about LMCE & Pluto internals etc..)...

HTH,

regards,

Bulek.

[Zaerc]

About the SVN part, as we all see the need for one or more people to approve and apply modifications requested by the community.  I can see a couple of options on that:

• We could use a branch in the current LMCE SVN, but that would likely mean extra work for Paul, at least to give access but probably a lot more.
• Somebody could set up a (private) SVN using the tarball sources and accept patches on that.
• We could also try sourceforge, but I have no idea what that requires.

But before a decision is made on this, I think we should figure out and document how to build the sources, at least up to the point of .deb packages which can be used to up/down-grade an existing installation.  I'm confident that we're able to do that with the pieces we already have.

Once we have that out of the way that we can set up a seperate community develpment repository and feed it so the changes can be tested by more people.  By this time a decision should be made on how to deal with SVN.  That should give us plenty of time to discuss all our options I think, and SVN is only really usefull if people know how to build from it anyway.

So I guess that's the direction I would like to go in...

[nite_man]

Sourceforge can be a good to keep the project there. Many projects are hosted there (I have a few ones). I'm not sure only about license. Maybe it may cause some problems.

[webpaul1]

I first tried source forge, but the problem there is that they only host small projects.  Each project only can use 100mb, and the forums/bug tracking, etc., are really geared towards small projects.  LinuxMCE's sources were too big.

[Zaerc]

Ok, I guess we can scratch sourceforge off the list then, thanks Paul.

[tschak909]

I could buy a xen slice and set things up there?

-Thom

[dopey]

That would be wonderful. I wouldn't mind contributing to it either.

The question is what size do we need? Slicehost seems to have very reasonable prices for this, it's just a matter of just how much storage it would need. I don't think we would need that much bandwidth, as long as we don't allow everyone to download the the dvd every day...

Another problem is who will be responsible for setting up and maintaining the slice. As I said before, I'm not experienced enough in that area to do it myself and I really don't think Paul should have to do it.

[webpaul1]

Sorry if I'm not totally familiar with that 'slice' hosting and what the advantages are over a normal dedicated server.  But it seemed quite a bit more expensive than a normal dedicated server.  I'm using a server from go-daddy.  It's about $100/month for a p4 with 1gb ram, and then you pay $40/month for each extra 1TB of bandwidth; linuxmce uses about 2tb/month.

The problem, though, with using one of those go-daddy servers as a build box is that it only comes with Fedora.  When I do the building I always do it with the LinuxMCE/Kubuntu distro so all the headers and stuff match.  I guess it could be built on a fedora box, but it's probably more work.

[dopey]

With slicehost you pay $70 for 1GB Ram 40GB storage and 400GB of bandwidth plus you can install several different distributions and they don't oversell, so you really are getting the bandwidth and resources you purchased. You can also buy extra bandwidth, but it seems that price is negotiable. Really, they are similar concepts that's just implemented differently.

[tschak909]

and if the thing explodes, it can simply be rebooted from the management console, or even re-imaged. I find less and less of a need to use dedicated boxes anymore.

[tschak909]

Paul? What if we got a slice from gplhost.com or somewhere similar?

also, what ARE the immediate priorities currently in the development branch?

-Thom

[seanyseansean]

Am I being stupid here, or wouldn't it be a good idea to use Ubuntu's Launchpad for the project hosting?

https://launchpad.net/

[webpaul1]

The immediate priorities, imho, are (1) building a community, and (2) any stability or installation issues that will allow a majority of users to get going with LinuxMCE.  Most of the emails I get are for feature requests.  But LinuxMCE already has a gazillion features, and while there certain new features that would be really nice to have, I think the community will be better off if the features that are already there are accessible by everyone.

So, if I was to put the priorities in order, I would say:

1) Find someone who has knows the FOSS world well, has some clout in the FOSS community, and has some free time, and who can act as a community liason to start building a team.
2) Find someone who's good at writing docs, ORG charts and other tools to document LinuxMCE, both enhancing the existing quick start guides, but also laying the foundation for developer docs so that as developers join they can jump in right away.  There's actually a ton of info already in the wiki, I think it just needs a clear roadmap.
3) Working with the team to identify the high priority tasks.  My personal votes are: A) improving the integration with MythTV.  I still get complains that Myth is less stable in LinuxMCE than in stock Ubuntu, which baffles me given that LinuxMCE is stock Ubuntu + stock MythTV.  But since the PVR is such a key component, this needs to be resolved.  B) Trying to determine what are the installation issues that prevent a lot of users from getting their systems going.  C) Implementing the PIN code to protect private media (ie keep the kids from watching Dad's porn ).
4) The next big, new task will be re-doing the UI to make it easily re-skinnable and themable and improving the designer stuff.  This way the creative design guys can work on making it look really beautiful, since the aesthetics seems to be the first thing people pay attention to.