[IronPython] IronPython 1.0 Beta 1 / Mono 1.1.12 Compilation

[Zoltan Varga]

                               Hi,

  The upcoming mono 1.1.13 release will run IronPython 1.0 beta 1.

                   Zoltan

On 1/5/06, Jeff Griffiths <jeffg at activestate.com> wrote:
> Hi Jim,
>
> Thanks for taking the time to go over our issues. I'll see about
> building a snapshot of Mono to see if that helps, and have passed along
> all of your comments to the dev team. Mono is important to us because
> Komodo is available for OS X and Linux as well.
>
> I remember vaguely the bit about using Python's libs with the IronPython
> interpreter, thanks for reminding me. We'll grind on this some more and
> let you know how it goes.
>
> cheers, JeffG
>
> Jim Hugunin wrote:
> > Hi Jeff,
> >
> > It would be cool to see IronPython running in Komodo.
> >
> > The makefile doesn't work out of the box on anything except MS's
> > SSCLI implementation.  Your change to gmcs looks like the right one
> > to compile on Mono.  The TAB bug is one we need to fix.
> >
> > The general experience people report with Mono is that IronPython
> > will run successfully on versions released after a given version of
> > IronPython.  I know that the Mono team uses the test suite we ship as
> >  part of their testing process.  Because Mono is still in beta for
> > the CLIv2 features that we use, it can sometimes take a release for
> > them to catch up.  I expect that this lag will go away as Mono gets
> > to a final CLIv2 compatible release and IronPython development slows
> > as we approach 1.0.  Interestingly enough, the binaries that we ship
> > will often run on Mono even if the sources can't be completely
> > recompiled. You might want to try that.
> >
> > We'll take -u as a feature request, but if you have a work-around
> > it'll be pretty low priority.  Our primary hosting work is centered
> > around making the CLS PythonEngine class as solid and easy to use as
> > possible.  It never occurred to me that there'd be a desire to use
> > the command-line as a pseudo-hosting API.
> >
> > IronPython doesn't ship with any of the standard Python libraries
> > that are implemented in Python.  It is designed to be used with the
> > Python-2.4 standard Lib directory - just add this to your site.py or
> > copy into the Lib dir with IronPython.  You can get an initial os
> > module by doing this.  That'll get you os.getcwd.  The os.path you
> > get will be the ntpath (there was a recent message about this).
> > You'll probably want to replace that with the posixpath module on a
> > Unixy platform.  Os.popen is probably some work, but is a priority
> > for us.  This is something we want to get cleaned up.
> >
> > Similarly, you should be able to use the standard logging.py module.
> > I just tried this in beta 1 and it fails because our time module is
> > missing localtime.  We're working on these issues too, but you might
> > find a stub module is the best way to get things going in the very
> > near term.
> >
> > Thanks and let us know if you get IronPython running in Komodo - Jim
> >
> >
> > -----Original Message----- From: users-bounces at lists.ironpython.com
> > [mailto:users-bounces at lists.ironpython.com] On Behalf Of Jeff
> > Griffiths Sent: Thursday, January 05, 2006 11:38 AM To:
> > users-ironpython.com at lists.ironpython.com Subject: [IronPython]
> > IronPython 1.0 Beta 1 / Mono 1.1.12 Compilation
> >
> > I looked into getting IronPython working on Linux with Mono 1.1.12
> > this morning. In my defense, I just hadn't seen anything about
> > whether this would work either way, so these problems are likely
> > known issues?
> >
> > Some problems:
> >
> > 1. makefile: I changed this:
> >
> > CSC=gmcs
> >
> > then got this:
> >
> > makefile:16: *** missing separator (did you mean TAB instead of 8
> > spaces?).  Stop.
> >
> > I changed the 8 spaces to a tab in vim, then got a little farther:
> >
> > jeffg at apple Src> make gmcs -t:library -r:bin/IronMath.dll
> > -out:bin/IronPython.dll -recurse:IronPython/*.cs
> > IronPython/Compiler/NewTypeMaker.cs(1011,17): error CS0019: Operator
> > `==' cannot be applied to operands of type
> > `System.RuntimeMethodHandle' and `System.RuntimeMethodHandle'
> > IronPython/Hosting/PythonEngine.cs(76,20): warning CS0219: The
> > variable `o' is assigned but its value is never used Compilation
> > failed: 1 error(s), 1 warnings make: *** [bin/IronPython.dll] Error 1
> >
> >
> >
> > Is this a mono version incompatibility with 1.0 Beta 1? ( just
> > guessing )
> >
> > Background: we're looking at IronPython to see what kind of effort is
> >  involved in debugging it with Komodo. Using the MS .Net Runtime,
> > currently the issues are:
> >
> > - IronPython has no '-u' option. We can hack around this as we did
> > for Jython in the past.
> >
> > - IronPython does not have an 'os' module, which the PyDBGP client
> > currently depends on. Current os module usage includes: os.environ,
> > os.path.*, os.popen, os.getcwd. We have clever ideas about working
> > around this in a fepy-specific code path.
> >
> > - IronPython does not have the logging module. We could fairly easily
> >  cobble up a stub one that was good enough... or look at porting
> > enough of the logging module.
> >
> > cheers, JeffG
> >
> >
> > _______________________________________________ users mailing list
> > users at lists.ironpython.com
> > http://lists.ironpython.com/listinfo.cgi/users-ironpython.com
> > _______________________________________________ users mailing list
> > users at lists.ironpython.com
> > http://lists.ironpython.com/listinfo.cgi/users-ironpython.com
> _______________________________________________
> users mailing list
> users at lists.ironpython.com
> http://lists.ironpython.com/listinfo.cgi/users-ironpython.com
>