Open-source front end

[SirPoonga]

Well, if you want flame wars, just goto mame.net....

[rampy]

Anyway, I can see there is no way a project like this can come about.  At least from this group of people.  Too many opinions about wht is the best way.

FWIW, I think the project can and still be done... I do think open source is preferable way to go (greatest benefit to the community and greater potential pool of contributors) and it opens it up beyond this immediate group.

The project does need some structure in order to choose a direction.  A board of directors so to speak... after a group of core/lead devs/maintainers are chosen/volunteer they could then democratically work out the goals, design, language, etc...  people that are put off by the decisions can choose not to contribute or fork/start their own open source FE in their language of choice...

It would be neat if at the minimum, even if NO code is actually created some sort of standards/specification is hammered out.  FE's could even have a little badge that says "UberFrontEnd 2002 compatible" or something (i'm blue skying a bit here... work with me)

I dunno... maybe we can make it platform/language agnostic modular framework (kinda like what .NET hypes it self to be  -- now I curse M$ daily, but hopefully you know what I'm implying... the whole common IL runtime interpreter thingiemabob...)

*shrug*

Rampy

[Dave Dribin]

FWIW, I think the project can and still be done... I do think open source is preferable way to go (greatest benefit to the community and greater potential pool of contributors) and it opens it up beyond this immediate group.

In theory I agree with you, but reality has proven otherwise.  Game Launcher has been an Open Source FE for almost 3 years now.  I have only once received a patch (which I am grateful of!).  But in general, I have gotten almost zero interest in people wanting to contribute.  It looks like most other Open Source FEs have had similar lack luster followings.

Perhaps I am an ass and people just don't want to work with me.  However the real reason I think is that most people using GL (and FEs in general) do not have programming skills.  Open Source tends to work best when the users are technical and can contribute back to the project.  I think that's why most successful Open Source projects are more infrastructure type stuff: Apache, Linux, Perl, Python, Sendmail, Bind.  The users of these projects are programmers or sysadmins.  They have the ability to fix mistakes, add features, and contribute back.  I don't think a FE has that luxury.

I am open to suggestions (even personal ones) as how to make Game Launcher more approachable to someone who wants to contribute.  I am reluctant to completely scrap GL and start over, though, unless there is significant benefit.

There's a lot of cool stuff going on in the CVS version of GL.  There will be themes in the next version.  And it will be very flexible themes.  I've embedded the Lua language into GL and all themes will be written in Lua.  This should give an enormous amout of power to the theme writer.  I can't wait to get it fully working.

BTW, Lua is an embedable scripting language meant for extending applications.  It's small, effecient, and very portable.  And there's no semicolons or curly braces, either.  I've already tested it under DOS, Windows and Linux.  For more info check out:

http://www.lua.org/about.html

I also like having a common back end, or at least a common file format for game meta-information.  This could get rid of my map files.  It would especially be cool to get information for emulators other than MAME.  MAME is nice in that you can get most of the information you need from -listinfo.  Most other emus do not have that capability.  It would be nice to have a consolidated file with all the NES info, for example.

-Dave

[Lilwolf]

I think it could happen... but it would work best if we had different people excited about different pieces or modules and then combining them.

It would be great if we could also pull in parts from other FE's especially initially (so if soemone already has -listinfo parsers, we dont' have to do it right away and we aren't testing everythign at once).

btw, JFront is opensource also, but I don't think anyone has ever downloaded the source.

[Dave Dribin]

I think it could happen... but it would work best if we had different people excited about different pieces or modules and then combining them.

True, but these could be different modules within the same program/project.

btw, JFront is opensource also, but I don't think anyone has ever downloaded the source.

Other Open Source FEs are EmuWizard (and Cocktail) and AdvanceMenu.

There have actually been 500 downloads of Game Launcher's source since the last version, which really baffles me.  It's not that easy to compile and no one has asked for help.  I don't know what people are doing with the source.

-Dave

[)p(]

I also like having a common back end, or at least a common file format for game meta-information.  This could get rid of my map files.  It would especially be cool to get information for emulators other than MAME.  MAME is nice in that you can get most of the information you need from -listinfo.  Most other emus do not have that capability.  It would be nice to have a consolidated file with all the NES info, for example.

-Dave

And that is precisely what the dat and carver.ini combo can do for us. The clrmamepro dat files are in mame's listinfo format. Allready there are basic dat files available for a lot of systems. It is not perfect yet but with a bit of standardization of the fields to use I think they can serve as the common file format for out fe's to retrieve the meta information...especially because we all have a way to parse them allready...

Peter

[Lilwolf]

I have to admit.  I never thought of using t.he .dat files... and I love the idea.

but this will also assume that you have the rom files right?  Can you make a dat file of all your own roms?  I've only downloaded them

[)p(]

I have to admit.  I never thought of using t.he .dat files... and I love the idea.

but this will also assume that you have the rom files right?  Can you make a dat file of all your own roms?  I've only downloaded them

No the datfile contains all the known roms/games. The fe can generate the gamelist in any custom format it wishes from that and check for the availabilty of roms/games and do whatever. This way we have a common base for game information for all fe's to use. Personally I think that is enough...and easily implemented in most of the current fe's because we all alrready parselistinfo fomrat and catver.ini files...

Peter

Peter