[phoenix-project] Re: libcomedi
- From: Georges Khaznadar <georges.khaznadar@xxxxxxx>
- To: phoenix-project@xxxxxxxxxxxxx
- Date: Sun, 4 Sep 2005 13:32:28 +0200
Hello,
I browsed http://www.livejournal.com/users/pramode_ce/5161.html, I see
that you know about Comedi more than I do :)
Pramode C.E. a écrit :
> Don't know whether there are enough COMEDI based
> applications to justify a port of the Phoenix driver
> to the COMEDI framework.
Presently there are quite few applications written on top of libcomedi I
am aware of.
- Qtscope
- Qt-wave
Those are packages hosted by USBDUX' authors, I turned into .deb bundles
- Ktimetrace
This package is a rough frontend to drive data acquisition, I wrote
a little python script to have it output the data directly to a new
XmGrace session.
- python-comedilib
Python wrapper for Comedilib
Those applications just know libcomedi.
However there are other noticeable environments connected to libcomedi
in one or another way :
- you may find useful the Toolkit for Scilab, see the document
http://www.google.fr/url?sa=t&ct=res&cd=5&url=http%3A//www.linuxdevices.com/files/rtlws-2004/RobertoBucher.pdf&ei=FskaQ5-mEruWRra33I8N
written by Roberto Bucher,
- there are ongoing works about connecting Comedi and Labview, see
http://www.openg.org/tiki/tiki-index.php?page=Comedi+Library+for+Linux
=====
So I come back to your question: are there enough COMEDI based applications ?
From a developer's point of view, it may be easier to modify the source
code of an open "oscilloscope software" to comply with already written
drivers, than to rewrite and proof-test the drivers to comply with
Comedi.
Let's now consider a customer's point of view: for example I must
regularly advocate to convince colleagues that we can use free software
to teach physics. A recurrent question question which arises is about
the possible « isolation » of a solution. They are conscious of the
isolationist behaviours of our retailers: they are two of them, Company
Jeulin and Company Pierron, selling hardware and software badly
interoperable in a variety of subtle ways ... Why should we undergo
another submission to a newer isolationist company, whose future
behavior is still unknown?
When I proposed buying USBDUX interfaces
(http://www.linux-usb-daq.co.uk/), the fact that they were bound to
Comedi made a kind of guaranty : these particular interfaces would be
swappable with a lot of other devices
(http://www.comedi.org/hardware.html), so the solution we choose can
still work even if we cannot get newer supplies from the same retailer
in the future. The fact that there are still few end user software was a
drawback, but it sufficed that some basic acquisition frontends be able
to communicate with other data-managing programs to make us happy.
That's why I advocate to use common standards whenever they exist and
are not bound to a hostile standards-maker.
Best regards, Georges.
Other related posts:
