PCManFM-Qt now supports Qt 5.1.

[PCMan]

Hello,
here're good news and bad news.
The good ones:
1. PCManFM-Qt now fully support Qt 5.1.
2. The X11 related code now uses XCB in Qt5, not Xlib.

The bad ones:
1. Qt5 is much heavier than I thought. The memory usage dramatically
increased after porting to Qt 5.
2. The Qt5 version of pcmanfm-qt uses about 10-20 MB of resident
memory (according to htop). The shared memory increased, too.
3. Some weird crashes never happened in Qt 4 were noted.

Because of the significant increase of memory usage, I'm not so
excited about the port to Qt 5. I'm not sure if there are possible
ways to fine tune this part. :-(

[PCMan]

On Sat, Apr 5, 2014 at 5:53 PM, PCMan <pcma...@gmail.com> wrote:
> Hello,
> here're good news and bad news.
> The good ones:
> 1. PCManFM-Qt now fully support Qt 5.1.
> 2. The X11 related code now uses XCB in Qt5, not Xlib.

The use of X11 related stuff is minimal. This won't affect our future
wayland port.

> The bad ones:
> 1. Qt5 is much heavier than I thought. The memory usage dramatically
> increased after porting to Qt 5.
> 2. The Qt5 version of pcmanfm-qt uses about 10-20 MB of resident
> memory (according to htop). The shared memory increased, too.

Sorry, I mean the Qt5 build uses "additional" 10-20 MB, which is quite a lot.

[PCMan]

OK, I examined this again with a memeory profiler (massif) and got
some more info.
The most memory hungry part is not in Qt 5.
I ran my desktop inside Virtualbox so the result may be influenced by vbox.
In the report generated by massif, the code used most memory is in
VBoxOGLcrutil.so.
So it's possibly caused by the OpenGL support of virtualbox.
Later I'll try to test it again on my laptop, on which archlinux runs natively.
Cheers!

[PCMan]

Some more tests are done. I tried to turn off the 3D acceleration
support of Virtualbox.
Then, I did heap profiling again. This time the memory usage became normal.
With process monitor "htop", the resident memory used is still more
than that of the Qt 4 version.
However, with massif heap profiler, the heap used is decreased.
So, overall Qt 5 is not that bad. :-)
Cheers!

[PCMan]

OK, I did some more tests today.
Now, PCManFM-Qt compiled against Qt5 can run in Wayland!!!
Known problems:
1. Qt5 uses significantly more memory regardless of backend used. Both
xcb and wayland platforms uses much more memory than Qt4.
2. Mouse cursor position and context menu are buggy.
3. Desktop icon manager does not work.

Well, I really think that Wayland and its Qt support are still far
from production use. However, in future development we need to keep
Qt5 and wayland in mind so later the porting can be easier.

Cheers!

[gary sheppard]

@PCMan,

That is nice news to hear! Any idea if some of that overhead might be due to extra debugging turned on inside of QT5? Like you I believe Wayland and QT5 should be kept in the loop so to speak. Perhaps keep it on an actively turned on "back burner" to keep simmering along and developing?

This project just keeps getting better, and better. Thank you all, for all of the hard work!

--
--
You received this message because you are subscribed to the Google
Groups "Razor-qt" group.
For more options, visit this group at
http://groups.google.com/group/razor-qt?hl=en

---
You received this message because you are subscribed to the Google Groups "Razor-qt" group.
To unsubscribe from this group and stop receiving emails from it, send an email to razor-qt+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.