On 05/25/2018 03:17 PM, Ulrich Wilkens wrote:
On 05/25/18 19:12, Jon Trulson wrote:
On 05/25/2018 10:36 AM, Ulrich Wilkens wrote:
Hello,
at the moment all BSD and Solaris platforms are broken. Even some Linux
platforms doesn't built anymore. And this is for several reasons. The
error mentioned in the previous mail is only the first one. Beside of
that a lot of new warnings are generated.
Really? I've seen the warning go from the thousands down to 403
(after my fix for the problem Dietmar reported). I will commit that
one to master in a few minutes.
Yes, but I'm talking here about the BSD systems. Most, but not all new
warnings are coming from makedepend which is still in use for BSD.
Right. I'd love to just get rid of makedepend. If you could arrange
for clang to do dependencies like what we do for Linux with gcc/g++...
[...]
Otherwise its impossible for me to fix the broken systems.
If you are talking about the mass removal of old systems patches from
Chase, we can stop putting those in master, leaving that free for
testing. Only actual fixes would go in there, and anything else goes
into cde-next-2. Would that work for you?
True, I'm talking about Chase's patches. Your suggestion sounds good.
But we should agree about a starting point. As far as I've seen the
ultrix removal is pending. Should it go into master or cde-next-2?
IMO, mass removal of old OS's should probably be a low priority, saved
for AFTER the next official release, if Chase is up for that. Such large
patches are hard to review and verify, and have the definite possibility
to destabilize things.
Is it possible to merge master and cde-next-2 afterwards? I'm not very
familiar with git. SVN was able to merge changes automatically whenever
possible. Can git do this too?
Yes - what I usually do is periodically rebase master onto cde-next-2
(or whatever topic branch I'm using). When done, then we just take the
rebased patches and apply them onto master. This avoids a merge commit,
and any potential merge conflicts (if done right of course :)
Also, I was not aware you were even trying to do any testing (or that
you were even on the list anymore :)...
Please feel free to let me know when you are doing this kind of stuff
so I can act appropriately in the future.
Well, I'm still doing things from time to time. The first thing I will
describe in another mail. The other is a port to macOS which is nearly
ready. It has only two annoying problems left (not in CDE, but in the
implementation of X11 on macOS).
Nice!
Linux systems will need (by default) to have tirpc-dev installed now.
If thats not the case on your Linux test systems, you will likely have
build problems, just FYI.
Well, the problem is related to tirpc, but its a little bit different:
glibc has removed support for sunrpc, so /usr/include/rpc is nearly
empty.
For systems using the new glibc tirpc is now mandatory. Unfortunately
not all places in CDE use tirpc even if its enabled. This is fixed by my
patches.
Damn, I thought I had all of those... I didn't notice any issues.
Anyway, I think it's fine if you want to start your testing run now.
We'll keep things in cde-next-3 until that's done and your patches are
merged. Then we can go from there.
[...]
--
Jon Trulson
"But when I'm in command, every mission's a suicide mission."
- Zapp Brannigan
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
cdesktopenv-devel mailing list
cdesktopenv-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/cdesktopenv-devel
