Thanks, Tim. I reported the argument-handling bug on the mailing list, the other thing escapes me for the moment but it's an obscure code path and if it will trigger, you get a SIGSEGV and it's easy to fix then ;)
So far I'm getting the Xt code to use Qt. I will then backport improvements made in the Tk version. I fully understand that you want nothing further to do with Xt-based code, it's a very royal pain. The Xt version doesn't work very by itself -- it has lots of painting glitches around the widgets. The main area works fine, but Xt widgets -- not very much. It's interesting that merely re-implementing parts of Xt using Qt fixes all those problems for "free". I work off the latest 3.6 release, so if your multi-window changes are in 3.7, I haven't gotten to it yet. I don't think I want to support TCL at all. If someone else will contribute code -- fine, but I won't touch it. Python seems a best match, especially since you can get nearly native performance using LLVM, CLR and JVM-based implementations. Only cPython is using bytecode interpretation, other implementations I mention are using an optimizing JIT. Frederic Bonnet posted once on TCLCORE that it'd be nice to have a TCL frontend to LLVM, but I'm not aware of any runnable code anywhere to do that. Besides, TCL lags behing pretty much everything else in terms of availability of libraries, and I really got turned off by the fact that they can't seem to agree on what's the OOP model they want to use. This is just my personal set of pet peeves, I don't intend to discuss it, code contributions welcome yada yada ;). Python integrates very well into a C++ Qt application thanks to the PythonQt project <http://pythonqt.sourceforge.net/> -- it's pretty much as simple as it gets. Cheers, Kuba On Sep 23, 2010, at 9:27 AM, Tim Edwards wrote: > Dear Kuba, > >> fixes: >> - text input work >> new stuff: >> - toolbar highlighting >> - tool cursors >> - loading files via menu - see >> <http://home.wowway.com/~mandasoft/xcircuit-4.png> >> - handling application closing >> bugs: >> - scrollbars and panning aren't functional yet >> So far most bugs are due to my lack of experience with Xlib/Xt semantics, >> and occur in the glue code. There were IIRC two genuine bugs I've found >> in Xcircuit code so far, one was some obscure type error somewhere, the >> other was incorrect handling of arguments with XtOpenApplication. That's >> not bad :) > > Please forward to me any reports of true bugs, because I would like to > correct them in my version. > > Bear in mind that I have not even bothered to compile the Xt version for > several years; all my development work is on the Tcl/Tk version. Most > errors you find are likely to be parts of the Xt code that broke when I > changed something in the Tcl/Tk version. There is not too much difference > between the way the two versions behave. However, I would urge you to get > the original Tcl/Tk version compiled and running, and compare it to the > Xt version. In particular, the Tcl/Tk version supports multiple windows, > but the Xt version does not---I didn't feel like writing all the GUI > routines for Xt. It takes up enough of my time to support one version. > > ---Tim > > +--------------------------------+-------------------------------------+ > | Dr. R. Timothy Edwards (Tim) | email: [email protected] | > | Open Circuit Design, Inc. | web: http://opencircuitdesign.com | > | 22815 Timber Creek Lane | phone: (301) 528-5030 | > | Clarksburg, MD 20871-4001 | cell: (240) 401-0616 | > +--------------------------------+-------------------------------------+ _______________________________________________ Xcircuit-dev mailing list [email protected] http://www.opencircuitdesign.com/mailman/listinfo/xcircuit-dev
