Wrong code

[anjohnson]

Two problems:

1. The code in this (epics-rip/caj) repository actually contains Matej's JCAS pure-Java CA Server implementation, but the README file talks about it being a CA client. If the JCAS is intentionally being abandoned (I don't remember discussing that at all in the core developer meetings, I would prefer that we explicitly state we are doing so) this archive repository should be named epics-rip/jcas and the README should properly reflect what the code actually does.
2. The original CAJ and JCA libraries are currently described on this SourceForge page; JCA provided the client API and the JNI implementation, while CAJ provided the pure Java implementation. It is reasonable that we should keep using the JCA name for the client API, but we should also have an archive of the JNI C code, and it should appear under a project named epics-rip/jca or epics-rip/jca-jni — it was never called CAJ so should not use that name.

[ralphlange]

This is the original caj repository, and it contains the code that always was in the caj repository.
Kunal has imported the caj parts into the jca repository, keeping their full commit history.
If Matej's JCAS is contained in this repository, it was always that way, and if the README doesn't mention it, it never did.
So - if the issue is really "wrong code", it was always wrong.
@shroffk, can you comment on the JNI situation? I remember we discussed that, but I don't recall the details.

[shroffk]

I moved the JNI bits to its own project
https://github.com/shroffk/jca-jni

We should move that project here...under epics-rip

[shroffk]

should I transfer ownership?

[ralphlange]

Sure. The yard is big enuff.

[shroffk]

done

[ralphlange]

So the remaining thing is the Java server?
I honestly never knew there was one until earlier this week.

[ralphlange]

I'd be fine with putting that on the agenda of the next telecon. Unless there's a hurry, which I don't expect.

[anjohnson]

@shroffk thanks for archiving the JNI code.

No rush for the server discussion.

Has anyone in the CS-Studio group looked at PSI's CA libraries?