Re: Why Linux?

I will stay out of the flame war, but this is an interesting topic.
Someone just re-posted my original note describing my AcuRite clone. I
have been thinking about this for some time. Here is my delima:

I have a working AcuRite clone under Win95/VB. Its only real drawback is
the lack of real time control- once I send the motors off on a move my
VB gode is completely locked up in a .ddl. I cant abort the cut, and I
cant update the display to show whats happening. The program does work,
and I have made real parts with it. Ive also extended it to read g-code
generated from Bobcad, and I can output real AcuRite files, which my
friend can run on his real AcuRite.

I see a couple solutions to this:

1) I have considered fixing this by buying I-LPT. I bought an old
version (my very first expirement with the Sherline lathe) but I would
need the new version to get the real-time control.

2) I bought a Tiny Tiger microprocessor and implemented the movement
function in this. I can communicate to this via a serial line. This
works, but it requires an external hardware box, and Ive knocked
FlashCut for that, Im not to eager to follow it. If I ever consider
selling my software, I dont want to also get into the hardware business.

3) I could try to write a real-time .dll, but that is re-inventing
I-LPT, and would still have the limitations of WIN95, and be caugh up in
the endless upgrade to support Microsoft.

I have considered moving my project to use EMC/Linux. (I am a senior
Unix consultant in real life, for which people may my company outrageous
ammounts of money for me to administer their unix systems) I can cope
with Linux just fine. But, the GUI tools on Linux are weak. I have a
rough start on a VB like program using Perl/Tk- It puts up a toolbox
window, you click controls, place them on a form, and the perl code
writes the perl/tk code to run that form. You would 'simply' have to
write the callback code. To finish this I need the tk to emc binding,
which Fred recently said he was doing.

Is there anyone else here that would consider joining some kind of
collabrative project to develop this under EMC/Linux?

Is it more valuale to continue with the WIN95/VB track?

Or should I chuck this whole thing and return to making models, which is
why I got into all this in the first place!

ron
http://www.ultranet.com/~ginger



ron