Lars Gullik Bjønnes wrote:
Helge Hafting <[EMAIL PROTECTED]> writes:
| Lars Gullik Bjønnes wrote:
|
| >| >To set optimization level you should use --enable-optimization='-O2'
| >| >
| >| >
| >| Ok, tried this. More exactly:
| >| $ ./configure --prefix=/usr/local --with-frontend=qt --with-gnu-ld
| >| --with-qt-includes=/usr/include/qt3 --disable-debug
| >| --disable-concept-checks --enable-optimization=-O2
| >
| >You should use '--disable-stdlib-debug' as well. that is a real time
| >consumer, nice when developing but not else.
| >
| So --disable-stdlib-debug is not implied by --disable-debug?
No, I do not think so... (hmm ... I cannot remember how I implementet
this bug stdlib debug and debug are realy ortogonal)
Well, this helped -- a little.
With
$ ./configure --prefix=/usr/local --with-frontend=qt --with-gnu-ld
--with-qt-includes=/usr/include/qt3 --disable-debug
--disable-concept-checks --disable-stdlib-debug --enable-optimization=-O2
I got 4.7s starting lyx instead of 5.2s
Starting with LANG=C brings this further down to 4.6s, so
internationalization
is not the big culprit.
The test method is to run
time lyx
while clicking the mouse fast in the place where the closing button is
going to
appear. Repeated clicks are used so my reaction time won't affect the
result.
I barely see the main window, it is removed as soon as it appears.
The same kind of timing for lyx-qt 1.3.6 from debian is 0.7s, with or
without
translations. So it doesn't seem like a debian problem.
Is there some sort of profiling I could try?
Helge Hafting
