Re: [lmi] First version of the illustrations pagination patch

[Vadim Zeitlin]

On Sat, 14 Jul 2018 12:51:09 +0000 Greg Chicares <address@hidden> wrote:

[...]
GC> ...and perhaps jump to a later wxWidgets version at that time. But we
GC> don't need to make that decision today,

 Yes, absolutely. I forgot to add my usual refrain about this being not
urgent at all, but it was implied.

GC> and it's possible that other wx regressions will be discovered in the
GC> coming week.

 For once, I'm more optimistic and rather think that I could get some high
DPI fixes into wx until then. But, to be honest, the most likely outcome is
that we're both wrong.

GC> I imagine I should first commit and push PR 86 in full in any case, as
GC> upgrading wx again is just a forty-character change on one makefile
GC> line.

 Yes. If I may be immodest for a moment, it's really great how simple it is
to upgrade wxWidgets used by lmi now.

GC> Thus, I don't see any need to test today's wx HEAD with MinGW-w64 g++-7
GC> immediately.

 Yes, again, I definitely agree with the "immediately" part.

GC> I'm not eager to upgrade MinGW-w64 to a later version than debian
GC> provides.

 Yes, but I hope that Debian will provide it soon. Maybe I'm over
optimistic again however, as I wasn't aware of the problems with MinGW-w64
8 that you mention. But at the very least I'd like the code to compile
cleanly with native Linux g++ 8, which is available in Sid since some time
already -- this can do no harm and g++ 8 enables a few new warnings, some
of which are potentially useful (and some of which, as always, are not so
much...).

 Anyhow, to summarize, I definitely support the plan of merging the current
PR in first and updating wxWidgets later. Maybe it could be combined with
the next (still unfinished, unfortunately) PR adding proper header support.

 Regards,
VZ