kaffe 1.1.8 (Was: Re: [kaffe] Future directions for Kaffe)

[Dalibor Topic]

Kiyo Inaba wrote:
> So, from releasing point of view (based on Dalibor's and Riccardo's
> wish list), we still have to
> 
> D] Either fix the serialization bugs leading to the two failing
> regression tests, or mark them as XFAILs, and file bugs for it in GNU
> Classpath's tracker, adding the failing tests to Mauve

The XFAIL part is now done.

> D] Merge in the dropped arm jit patch
> 	Merged by Dalibor, but verified not working by Kiyo.

I'm currently waiting to finish making kaffe build with a classpath 
install first, as building glibj.zip in the emulator takes too long.

> D] Merge in the dropped cygwin patch

Still todo.

> D] Merge in the interix port

Still todo.

> D] Remove the merged in fastjar and zlib and just use the jar tool on the
> system

Done.

> D] Remove the merged in GNU Classpath and just use an existing installation
> 	I did the lower-half separation (separate build) but not yet
> 	change the upper-half to point the location of classpath.
> 	Ideas like 'with-classpath-prefix', 'with-classpath-libdir'
> 	or 'with-classpath-includedir' involved in Cacao should be
> 	needed.

I've got a local version of that building, using a slightly adapted 
classpath.m4 from Cacao and a local install of classpath. I'm working on 
making the regression tests build & work with classpath 0.93 (closest to 
the version in our tree).

> R] "fix it at best" of classpath so that swing/gtk peer runs

That could work out of the box with the locally installed Classpath, 
once the previous part is done. I'll also merge in the stack of build 
system patches that went into Classpath after 0.95 to make it work with 
automake 1.10.

D) Merge in tbfg's m68k patch. Kiyo, could you take a look at it?
http://www.kaffe.org/pipermail/kaffe/2007-June/104949.html

cheers,
dalibor topic