Binary Compatibility ABI work

Ranjit Mathew rmathew@hotmail.com
Fri Mar 5 13:44:00 GMT 2004


Andrew Haley wrote:
[...]
> Although these patches work after a fashion they're not suitable for
> mainline.  They're buggy, messy. and very much a work in progress.
> 
> However, some folk have got to hear about this work and they might
> think I'm being unnecessarily secretive.  To counter this appearance,

That's unfortunate - as a fellow though-not-so-regular GCJ
hacker, I fully appreciate that you do this in your free
time and that you'd won't like to inflict something that
you're not fairly satisfied with yourself onto the
community.

However, I would request you take some time and
document the new ABI as much as possible - I know
it is not fully fleshed out, but even a 100m
overview and implementation strategy would be
really useful. I also know that this information
can be gathered by perusing the mailing list
carefully, but it's hard.

Instead of a branch, which might be an overkill
if no one else steps forward to hack at it or
help test it, is it possible for you to post
diffs against the mainline? (I hope you *are*
writing comments detailing your intentions/algos in
the new code, aren't you? ;-))

My 2p.

Ranjit.

-- 
Ranjit Mathew          Email: rmathew AT hotmail DOT com

Bangalore, INDIA.      Web: http://ranjitmathew.tripod.com/



More information about the Java mailing list