Re: What OO extension for 8.6?



Georgios Petasis wrote:

I had the impression that 8.6's OO was targeting extension developers. Is it good enough for everyday use? I will check it...

Regarding my problems with itcl, its an issue discussed many times. No bug reports exist and none will be made, since there is not a simple script that can reproduce the crash. I suppose if the bug was simple enough to discover, it would have shown up in the itcl test suit.

I have provided a crash report with a stack trace, but this provides little information. I have offered to send the crashing app to itcl developers, but I don't know if there is anyone :-). I have two crashing apps, which work fine in 8.4 & 8.5 with itcl 3.4. They crash in 8.6 with itcl 3.4, 4.0b3 & itcl-ng (which is still in development). [Ok, with itcl-ng there is no crash, but instead I am getting an error on constructor :-)]

Is there a decent extension that will work also in earlier tcl versions?
Or an 8.6 OO "implementation" for at least 8.5?

Itcl 4.0 and Itcl-NG are the same thing.

As far as I know, Arnulf is working on it still. He and Donal
are likely to be the key players in hunting down your problem.

How much infrastructure is needed to build and test the crashing
app? Does it require C extensions, or can it run with stuff from
the Tcl repository? Is it all scripts? (If so, you can attach
them with a bug report.) Does it require complex user interaction
to reproduce the error?

Distilling things down to a simple test case helps, but it's
not an essential prerequisite to a bug report. If you have something
that crashes Tcl and Itcl built from the HEAD, we do have a few
expert bug hunters that can start chasing things.

Forget trying to run 3.4 with 8.6. We *know* that won't work.

--
73 de ke9tv/2, Kevin
.



Relevant Pages

  • Iwidgets install problem.
    ... Well, it's everyone's favorite bug reporter, back to bug you about yet ... another bug when trying to build Tcl and extensions from scratch. ... (hereafter itcl and itk, respectively). ... the config process depends on a shell script, ...
    (comp.lang.tcl)
  • Re: Tcl/Tk 8.5.3 RELEASED
    ... Why the official website so much encourages people to use Tcl if their input ... Apparently Tcl is not very popular these days and "fix the bug yourself" ... It's unfortunate that your bug report languished for so long, ... runs on Windows, OS X, and Linux. ...
    (comp.lang.tcl)
  • Re: Converting from Itcl to TclOO?
    ... It is not only the issue with Itcl, but also problems I detect with code execution that involves recursion. ... If there is such a thing, it is DEFINITELY a Tcl bug - 99% probability some NRE bug, ie, my responsibility. ...
    (comp.lang.tcl)
  • Re: Converting from Itcl to TclOO?
    ... It is not only the issue with Itcl, but also problems I detect with code execution that involves recursion. ... If there is such a thing, it is DEFINITELY a Tcl bug - 99% probability some NRE bug, ie, my responsibility. ... I still see the problem with the fresh ActiveTcl install. ...
    (comp.lang.tcl)
  • Re: Converting from Itcl to TclOO?
    ... It is not only the issue with Itcl, but also problems I detect with code execution that involves recursion. ... If there is such a thing, it is DEFINITELY a Tcl bug - 99% probability some NRE bug, ie, my responsibility. ... Yes I have a test script, and does not involve itcl (its a converter actually that registers itself as Itcl). ...
    (comp.lang.tcl)