> Now, this leaves the question where to put this compiler switch. If it
> was be needed for all targets, it might be best to put it into some
> ressource string yet to be defined. If it is not needed always, or
> perhaps on Linux only, it might be acceptable to put it into the compiler
> Now I DID clean up the mess from previous failed attempts before trying the
recompile, and it DID work.
Nice that it works now!
Yes, it is because a **very** old fpc bug.
Even using "-B" (build all) fpc does not clean all previous compiled files.
One of the cause is because fpc forget the
Hello fredvs,
following up my previous post about the error compiling the "language"
libraries, I found the cause now. The clue & key was the compiler's
suggestion "recompile with -fPIC"
Well I had tried that before, and it hadn't helped. But this might have
been due to my not cleaning up the
> It seems this version was meant to already use "mselang" compilatiin or at
> least compilation via llvm that it was meant to use, but as the regular
> What are you (& other list members) thinking?
I think there is a confusion, there is no call (yet) to mselang (the
compiler).
There is
Hello Sieghard.
Thanks to give attention to msegui.
> calling "Make" for a project I set up, "msei18n.trp", returns this error:
Are you trying to compile on a Unix machine (Linux or FreeBSD) ?
If yes, you should add "-fPIC" parameter to fpc compiler option:
I use this macro with the gui
Hello fredvs,
by the way, I got the latest msei18n programs to compile and run now, both
the one from release 4.6.2 and from your gitlab repository.
But they don't work correctly. Apart from an error shown on calling:
"Sessionmanager Error: SESSION_MANAGER environment variable not defined"
(that
6 matches
Mail list logo