On Tuesday 31 October 2006 03:35, Marco van de Voort wrote:
> On Mon, Oct 30, 2006 at 07:02:13PM -0800, johnf wrote:
> > If writting a wrapper was easy why haven't we got a completed QT
> > interface?
>
> The same reason as why GTK exists at all. QT had license problems for a
> long while. There simply was not much interest.
>
> Moreover, Lazarus is way more than a wrapper. It is a full designer system.
>
> > Why does the GTK2 have so many bugs?
>
> All widget sets have bugs.
>
> > With regard to the speed of Python. I truly wonder if today we have to
> > be concerned with performance since most of the time is spent in the GUI.
>
> All the python apps I've seen (and that were not that many, mostly small
> time utils) were also annoyingly slow in the GUI.
>
> Also a lot of buildprocesses have scripted parts, (xorg, tetex, openoffice
> to name but a few), and they actually manage to spend more time in
> generating a few files using a script than compiling the C app.
>
> > Like FPC 'c' can be used to improve any specially slow routines in
> > python. BTW the link suggest that Python won a few of the test. I find
> > this fact to be strange considering FPC is compiled.
>
> Multiple reasons possible. Python delegating to more optimized C code (IOW
> mostly bound not by Python code), or a simplistic benchmark (like some
> recursion stuff) that the parrot engine can optimize away.
>
> > I like FPC and Lazarus. But to suggest that Python is not an equal I
> > believe is miss leading and not taking the facts in consideration. BTW
> > most of what I have said also applies to Ruby.
>
> Now you really spoiled it. The only Ruby app I know is the FreeBSD
> portupgrade package, and it is unbearably slow.
OK guys I like FPC because it's compiled. What I'm suggesting is the OOP in
python works. Is it perfect - no (no private or protected vars)! Does the
GUI run slow - maybe. But no general apps written in python - you must be
kidding. In fact I would suggest there are more python apps in general use
than FPC apps. Does that mean they could not have been written in FPC - no.
Just that there are more.
I still say the QT and GTK2 for the LCL is not ready. A lot may have changed
but two months ago I could not compile 90% of my forms for GTK2 so I
delivered GTK1. OK so it's my fault that I didn't donate more time to the
GTK2 project to improve the GTK2 LCL. But I did help get the postgres
interface running better with a few bug fixes.
I don't like Ruby but not because it is slow but because it changes the way I
think about OOP. The only app I know of is "Ruby on Rails" which seems to
get many excited.
See new thread - I'll offer a bounty for the either QT or GTK2.
Now that I have said the above I have some very serious doubts about the
future of FPC. This thread is proof that my concerns are real.
M$ is moving to .Net - period. The only consideration is when will FPC
support .Net. IMHO not in the next year. But I read statements suggesting
that it will never be required. That's plain nuts.
Although, FPC and Lazarus have a very devoted group. The group is not very
large. In fact most of the work is done by five people (I could be wrong on
the number but you get the point). What happens if they decided all to have
lunch and get hit by a bus crossing the street. I'm afraid FPC and Lazarus
would die. Our community is not as large as the Python community or the
current hot fad Ruby. Of course there are advantages to a small group but I
would still feel much better about the future if there were many more.
John
_________________________________________________________________
To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
archives at http://www.lazarus.freepascal.org/mailarchives
