RE: RE[5]: SWT Crash Problem with MingW/GCJ (UPDATED)

["Erik Poupaert" <erik dot poupaert at chello dot be>]

>>>>> Right now, since we're releasing 3.3 soon, the 3.3 branch is a nice
>>>>> balance between stability and up-to-date-ness.  So I'd recommend
>>>>> trying it out, and also submitting patches against it.
>>>>> Sometimes if you're using a less-recent compiler, you might write a
>>>>> patch only to find the problem has already been fixed.  However, you
>>>>> have to weigh this against the pain of building a recent compiler;
>>>>> from time to time it won't work and if you update on the wrong
>>>>> day... bleah.  The risks are higher on less common platforms (i.e.,
>>>>> everybody notices when x86 Linux breaks).

I'm actually waiting for either a new release candidate for minGW to be
available for download or else for the minGW version to build natively.

The procedure that consists of first modifying the patches to match the new
version in CVS, and then cross-compiling from another platform to minGW, is
still beyond me. I haven't spent time with it, because I'd rather spend time
building applications with the compiler and isolate and report problems
while doing so.

As soon as the minGW version builds natively, however, it will be another
ball game altogether. I'll be able to work with something more familiar,
that is, the libjava java sources and in the CNI sources that support it.