aMule Forum
English => Feature requests => Topic started by: gbordiga on July 22, 2008, 12:29:51 AM
-
I think would be cool if amule would be splitted into a library and ui/core, just like pidgin and libpulpe (i think). Amule would benefit of a cleaner and more maintainable code and having an ed2k library would probably create new ed2k apps using the library and probably contributing to it.
Comments are welcome.
-
A lot of work.
-
If an independent library should be created, I see EC more as candidate for this.
-
If an independent library should be created, I see EC more as candidate for this.
The idea is to have library that is actually useful for other people.
A lot of work.
Not an amount of work, but global lack of direction in this project prevents creation of such library.
-
A "global lack of direction" as you call it is what allows this project to keep a handful of people working on what they like most, as opposed to what someone wants them to do.
Feel free to commission any work from any of the other developers for a fair fee if you feel something specific must be done.
EDIT: And let me point out that this project is not aimed to create "A ED2K network library for p2p client coding", but an actual p2p client. It's about coding our client, not helping other people code their clients, tho it's certainly welcome if someone can convince me that they want to work on split the ed2k part out of this project into a separate library while keeping a sane interface and extensibility, quality coding, complete porting of the application to use that library and of course all of that for free.
-
If an independent library should be created, I see EC more as candidate for this.
The idea is to have library that is actually useful for other people.
That's what I mean. We already have EC-bindings for some langs (at least I think so, as they are named like this). I think it would be useful to have a lib that can be used to link apps that are aimed to control aMule. So they don't need to be recompiled for a new svn tarball, when the only ec-change is the date. They could link against libec-v20080722, and the version is only updated when something really is changed.
-
Feel free to commission any work from any of the other developers for a fair fee if you feel something specific must be done.
EDIT: And let me point out that this project is not aimed to create "A ED2K network library for p2p client coding", but an actual p2p client. It's about coding our client, not helping other people code their clients, tho it's certainly welcome if someone can convince me that they want to work on split the ed2k part out of this project into a separate library while keeping a sane interface and extensibility, quality coding, complete porting of the application to use that library and of course all of that for free.
Thought i posted my message in the "Feature requests" forum. :o
I know amule is not "A ED2K network library for p2p client coding" but a lot of amule's code could just be reused by other projects. Just wanted to know the developers view.
-
Just wanted to know the developers view.
Since aMule is an Open Source project, everyone is free to take the sources and build an ed2k library. It is just that we won't do it.
-
Feel free to commission any work from any of the other developers for a fair fee if you feel something specific must be done.
EDIT: And let me point out that this project is not aimed to create "A ED2K network library for p2p client coding", but an actual p2p client. It's about coding our client, not helping other people code their clients, tho it's certainly welcome if someone can convince me that they want to work on split the ed2k part out of this project into a separate library while keeping a sane interface and extensibility, quality coding, complete porting of the application to use that library and of course all of that for free.
Thought i posted my message in the "Feature requests" forum. :o
I know amule is not "A ED2K network library for p2p client coding" but a lot of amule's code could just be reused by other projects. Just wanted to know the developers view.
I actually wasn't talking to you, but lfroen.
-
I think it would be useful to have a lib that can be used to link apps that are aimed to control aMule. So they don't need to be recompiled for a new svn tarball, when the only ec-change is the date. They could link against libec-v20080722, and the version is only updated when something really is changed.
Wrong all over. Generally speaking, casual users should not compile anything. That's why they're called "users". All other folk (developers, testers, etc) are supposed to recompile. That's an idea. Adding dll complexity only worth it in case of significant amount of shared code between different applications. And here, you have neither significant amount of code to share nor third-party applications. Now, do you propose to create such library in every language we support? Java? C#?
A "global lack of direction" as you call it is what allows this project to keep a handful of people working on what they like most, as opposed to what someone wants them to do.
No. Handful of people (me including) are working on this project because it is quite useful application which gratefully accept contributions.
This allowed project to get to this point. Although, it think that direction in sense of feature priorities, for example, would benefit this project.
And let me point out that this project is not aimed to create "A ED2K network library for p2p client coding",
Yes, you right, literally. But I do think that it would be good idea to split codebase to ed2k core in form of static lib (as we do with EC) and write GUI around it. IMHO it worth an effort