Experience with vb2py?

[asdf sdf]

vb2py sounds terric but i don't want to be the first guineau pig who
gambles his project on it.

i have a large VB project lying around that may have to be upgraded. i
would love to see it go to python.

anybody have experience with vb2py, the vb to python converter?

how solid is it? how complete?

my VB project use custom controls, is a multi-tier ActiveX project, with
a custom (non-ASP) architecture, using sockets and a COM component on IIS.

is vb2py ready for this scale?

i see that vb2py converts to pythoncard. that raises concerns because
i'm not aware that pythoncard is suitable for large, complex, multipage
forms containing hundreds of controls.

can vb2py handle that kind of conversion?

 

[Peter Maas]

vb2py sounds terric but i don't want to be the first guineau pig who
gambles his project on it.
>
> i have a large VB project lying around that may have to be upgraded. i
> would love to see it go to python.

It's reasonable to ask for opinions but if you want to use it for
a large project you should experiment with it anyway. Create a little
hello world dialog and try to convert it. Populate it with more
complicated controls, add database and network connections and try to
convert it again. My bet is that vb2py won't do *all* of the work for
you. I once tried to convert a medium sized project (25 kloc) from
vb6 to vb.net and the migration assistant supplied by MS aborted after
a 3 hours run. Python and VB are much more different so I assume that
vb2py saves the low level typing work like converting loops, ifs and
defs which is very valuable but the high level conversions that require
to understand the code will probably be up to you.

Please don't forget to report your results in c.l.py because
i don't want to be the first guinea pig who gambles his project
on it.

Mit freundlichen Gruessen,

Peter Maas

 

[asdf sdf]

It's reasonable to ask for opinions but if you want to use it for
a large project you should experiment with it anyway. Create a little
hello world dialog and try to convert it. Populate it with more
complicated controls, add database and network connections and try to
convert it again. My bet is that vb2py won't do *all* of the work for
you. I once tried to convert a medium sized project (25 kloc) from
vb6 to vb.net and the migration assistant supplied by MS aborted after
a 3 hours run. Python and VB are much more different so I assume that
vb2py saves the low level typing work like converting loops, ifs and
defs which is very valuable but the high level conversions that require
to understand the code will probably be up to you.

certainly a detailed investigation would be required before making any
project committments.

realistically, if a bunch of people are not already having success
converting minor projects then it isn't ready for the big ones yet.
vb2py is attempting something difficult. in my experience with
conversion programs, everytime you scale it up, you shake out some test
cases you never thought of. i'm sure that applies here too.

 

[MetalOne]

I tried to use vb2py about a year ago for a project of about 15,000
lines. vb2py was not ready for that kind of a task. I encountered
problem after problem. The author of the package was extremely
helpful and glad to here about the problems. He seemed willing to
almost immediately fix any problems that I had. However, I was on a
very tight deadline and after a few days of experimenting with vb2py
it was clear that it was not going to work for me.

On another note, the VB code that I was working with was unbelievably
bad. I found that I could reduce the amount of code from between 5
lines to 1 to about 30 lines to 1 depending upon the code sections.
Frequently this entailed completely different solutions to the
problems. The author of the VB code had absolutely no idea how to
factor out any kind of redundancy in the code.

The vb2py translations did not produce particularly optimized Python
code for the VB code. It also became clear that if I wanted the
project code in Python that rewriting the code by hand would be vastly
better than automated translations. So much so that I could not
recommend the automatic translation even if it worked.

In the end, the project was cancelled. Bad VB code was deemed better
than spending money to fix the code.