marble for mingw-win32? (+ some notes)

[Lubomir I. Ivanov]

On 19 June 2013 14:34, Tomaz Canabrava <tcanabrava at kde.org> wrote:
> I hate this cellphone email client...
>
> Em 19/06/2013 08:34, "Tomaz Canabrava" <tcanabrava at kde.org> escreveu:
>
>>
>> Em 19/06/2013 07:13, "Lubomir I. Ivanov" <neolit123 at gmail.com> escreveu:
>> >
>> > does anyone have an idea where i can find pre-compiled marble devel
>> > packages for windows/mingw/qt5?
>>
>> Doesn't exists, marble doesn't works with qt5 yet, so we are using qt4.
>>

ok, i see.

>> >
>> > i really don't want to go trough these steps for the sake of including
>> > the widget:
>> >
>> > http://techbase.kde.org/Projects/Marble/WindowsCompiling#Compiling_Marble_using_MingW
>>
>> You don't need the full marble, just the libmarblewidget, it should be
>> easier than all that trouble.
>>

no experience with said library, so may have to invest more time into that.

>> > also it doesn't seem possible to disable marble in subsurface ATM and
>> > are there any plans to do that?
>>
>> It was decided against the use of optional libraries, and marble gives us
>> a great looking map ( way better than osmgpsmap)
>>

i'm sure it's better than osmgpsmap...

>> > some other notes:
>> > i found inconsistencies in usage of Q_SLOTS: against slots: in some
>> > qt-ui/*.h files.
>> > so i manually replaced slots: with Q_SLOTS: in all places because i
>> > was getting an error. not sure what's going on there...
>>
>> Me neither.

i've noticed usage of both slots: and Q_SLOTS: in qt-ui/*.h.
what is the final resolution on the syntax? again, not sure what's
going on but qt5 is spitting errors regardless of -DQT_NO_KEYWORDS.

>>
>> > divelistview.cpp: i had to do a:
>> >         #if QT_VERSION < 0x050000
>> >                 header()->setClickable(true);
>> >         #endif
>> > since setClickable() doesn't seem to be a method in QHeaderView from
>> > Qt5?
>>
>> Well, we are using Qt4 at the moment but it's good to make the code
>> correct for qt5 in the future.
>>
>> > also is it me or the current state of the Makefile(s) is
>> > worse/more-complicated than the GTK version...?
>>
>> Define "more complicated", qt didn't introduced complexity, thiago worked
>> on the makefile to make it safer, better. If you think that handwritten
>> makefiles are way too complex to be a sane environment, the I agree. =)

qt's moc and uic tools do introduce complications. that's how i see it..
but as long as the Makefile works for users i'm happy about it and as
long as they don't have to edit anything.

attached is the diff i had to do to make it compile and run with Qt5
on Windows...you can probably see that there are some issues related
to anonymous enums, QToolTip and .moc files inclusion that i have not
much of an idea about. the Marble "globe" is obviously disabled.
also attached is a Makefile copy-paste abomination that was used.

that was a pointless exercise in a way, but more of check if "can it
be done, at this point in time?".

lubomir
--
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Makefile.win32
Type: application/octet-stream
Size: 7643 bytes
Desc: not available
URL: <http://lists.hohndel.org/pipermail/subsurface/attachments/20130619/0635f0fb/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: qt5.0.1_mingw_4.7_win32_no_marble.diff
Type: application/octet-stream
Size: 24832 bytes
Desc: not available
URL: <http://lists.hohndel.org/pipermail/subsurface/attachments/20130619/0635f0fb/attachment-0003.obj>