Hello Rene
> Hi Valeri,
>
> I am definitively lost with your reactions about Qt and in particular
this one.
> If Masa succeeds in generating the dictionary for the Qt classes, his
solution
> is far superior
I have to beg a pardon for my confusing comment.
Oo sure, It is great !!! But Qt is just a GUI library. For example I
personally don't need to see the Qt (foundation) classes CINT
dictionary. What I am interesting in is to be able to create a CINT
dictionary for my classes derived from the Qt. At the moment I have to
hide the Qt methods with #ifndef __CINT__
CPP which is ugly of course.
Certainly it is great if Masa can create a CINT dictionary for the user
class derived from Qt. (I believe he can)
> to the very poor solution based on Javascript proposed by
> TrollTech. Even Tony Johnson rejected the Javascript solution for JAS
!
I think it is a separate story. Tony wants to use CINT dictionary
inside of Java.
TrollTech wants to use there own "dictionary" (well V.Onuchine had
proved it is primitive vs CINT) with its "own" Java-like language. They
have more chance to succeed.
Anyway I am not against of Qt you know :-) just see my web site
http://root.cern.ch )
Valeri
>
> Rene Brun
>
> Valeri Fine wrote:
> >
> > Hello Masa, I understand the simple thing probably will work,
> >
> > On other hand if understand TrollTech philosophy properly, Qt is OO
> > class library. This means the end user is expected to create his/her
own
> > class by sub classing some Qt class.
> > I think it would be more useful if you may have created a "proxy" to
> > communicate
> > the Qt Scripting http://www.trolltech.com/products/qsa/index.html.
> > The same way the ROOT communicates the MySQL, for example. This is
> > rather simple to implement and would have allowed accessing the Qt
from
> > CINT "in full."
> >
> > ----
> > Best regards
> > Valeri
> >
> > > -----Original Message-----
> > > From: Masaharu Goto [mailto:MXJ02154@niftyserve.or.jp]
> > > Sent: Sunday, October 26, 2003 5:29 AM
> > > To: Valeri Fine; 'Melitta Dragaschnig'; cint@pcroot.cern.ch
> > > Cc: roottalk@pcroot.cern.ch
> > > Subject: Re: [ROOT] Re: [CINT] Qt library
> > >
> > > Hello Valeri,
> > >
> > > Thank you for your attention.
> > >
> > > I had exactly the same concern. But, things were not as bad.
> > > First, the event loop thing, it is a problem, of course. But, if
> > > it is a simple program, it doesn't hurt so much. It works like
> > > Visual Basic which you can do significant GUI development
> > > within a single thread.
> > > Second, the parsing thing was solved by using preprocessor.
> > > Nice thing about Qt was it does not have many tricky macros.
> > >
> > > Since I've only tried a couple of simplest example, I expect to
> > > face those problems as I proceed. But, so far, it is not so bad.
> > >
> > > Thank you
> > > Masa Goto
> > >
> > >
> > >
> > > ----- Original Message -----
> > > From: "Valeri Fine" <fine@bnl.gov>
> > > To: "'Masaharu Goto'" <MXJ02154@niftyserve.or.jp>; "'Melitta
> > Dragaschnig'"
> > > <mdragasc@edu.uni-klu.ac.at>; <cint@pcroot.cern.ch>
> > > Cc: <roottalk@pcroot.cern.ch>
> > > Sent: Sunday, October 26, 2003 11:25 AM
> > > Subject: RE: [ROOT] Re: [CINT] Qt library
> > >
> > >
> > > > Hello Masa,
> > > >
> > > > Let me call your attention like all other GUI libraries under
> > Windows
> > > > this implementation should face the notorious "event loop" and
> > "second
> > > > GUI thread" problems.
> > > >
> > > > Another problem with QT -based code is the automatically MOC
> > generated
> > > > C++ files. I suspect it would be hard for CINT to parse it out
> > properly.
> > > >
> > > > For example Qt header files are populated with C++ like
> > constructions
> > > > like:
> > > >
> > > > class myQt : QObject {
> > > > Q_OBJECT
> > > >
> > > > protected slots:
> > > > void myProtectedQtSlot();
> > > > signals:
> > > > void my Signal();
> > > > };
> > > >
> > > > Thank you
> > > > (some details one can find with the source code (see:
> > > > http://root.bnl.gov )
> > > >
> > > > ----
> > > > Best regards
> > > > Valeri
> > > >
> > > > > -----Original Message-----
> > > > > From: owner-roottalk@pcroot.cern.ch
> > > > [mailto:owner-roottalk@pcroot.cern.ch] On
> > > > > Behalf Of Masaharu Goto
> > > > > Sent: Saturday, October 25, 2003 10:11 AM
> > > > > To: Melitta Dragaschnig; cint@pcroot.cern.ch
> > > > > Cc: roottalk@pcroot.cern.ch
> > > > > Subject: [ROOT] Re: [CINT] Qt library
> > > > >
> > > > > Hello Melitta,
> > > > >
> > > > > I made a progress on embedding Qt library in Cint.
> > > > > Basically, it wasn't so difficult. Using VC++6.0, I can
> > > > > access Qt library from the Cint interpreter. So far, a couple
> > > > > of simple Qt tutorial programs run on it. I want to play
> > > > > and refine it a little more. I plan to include preliminarily
> > > > > version of Qt-Cint in next Cint release.
> > > > >
> > > > > Thank you
> > > > > Masa Goto
> > > > >
> > > > >
> > > > > ----- Original Message -----
> > > > > From: "Melitta Dragaschnig" <mdragasc@edu.uni-klu.ac.at>
> > > > > To: <cint@pcroot.cern.ch>
> > > > > Sent: Sunday, October 19, 2003 9:35 PM
> > > > > Subject: [CINT] cint with ffmpeg and imagemagick
> > > > >
> > > > >
> > > > > > Hello -
> > > > > >
> > > > > > >At this moment, Qt library has not been ported onto Cint.
> > > > > > >Some people tried it but there were problems parsing the
> > > > > > >header files. I believe it can be solved if I have the Qt
> > > > > > >environment on my PC. If you can wait for a couple of
> > > > > > >weeks, I'll investigate and let you know what I can do
about
> > > > > > >the Qt. Also, there are other people who want to use Cint
with
> > > > > > >Qt.
> > > > > >
> > > > > > I'll happily wait for further notifications
> > > > > > The QT Homepage http://www.trolltech.com/
> > > > > > provides good information about qt.
> > > > > >
> > > > > > greetings, Melitta
> > > > > >
> > > > > >
> > > >
> > > >
> > > >
This archive was generated by hypermail 2b29 : Thu Jan 01 2004 - 17:50:16 MET