[volt-nuts] do you like Labview in your labs?

John Lofgren jlofgren at lsr.com
Mon Dec 6 15:41:22 UTC 2010


In answer to the original question of this thread: No, I don't like to use LabView.

LabView is a very polarizing application.  You either love it or hate it.  I've always felt betrayed by NI portraying it as an easy to use graphical programming system that should be easy for people who aren't necessarily programmers.  In reality it's just a very clumsy and obtuse graphical overlay for a hierarchical programming structure.  You still need be proficient in C, or a C-like language to understand it.  In addition to that it's a huge resource hog.  The installed size is absolutely enormous.  Add to that their deviations from the IEEE 488 standard that stop it from working with generic interfaces and you really have something to loathe.

For a lot of simple data collection tasks I long for the days of HP Instrument BASIC or GW BASIC.  What I don't long for is all of the hardware setup issues with manual card setup and drivers.

Both LabView and Matlab are the 800 pound gorillas in the market.  Everybody knows them, or has at least heard of them, partly due to their penetration in the educational institutions.  It's a vicious cycle that works to their advantage.  You bring up a generation of students on your product and they will want to use it in their working life.  When it comes time to hire more people the former student, now experienced engineer, looks for more people who know the system they are familiar with.  Once that cycle is established it's hard to break.

I completely agree about the comfort factor.  Many companies, especially larger ones, seem to feel better about having something that they can pay for and get "support" rather than relying on a group of users.  I think that there is a fear of Opensource from the standpoint that it's less controlled than a purchased product and it might disappear or change in a negative way without warning.  Buying a package or product is bu no means a guarantee of stability, either, but companies just seem to be comfortable with the old way of doing things.  I also wonder if there is a perception that Opensource projects are inferior to pay solutions.  Whether it's right or wrong I could see how people further up the management chain could be suspicious of the quality of something that they don't pay for.

I, for one, am very glad to see  Opensource flourish.  I'm grateful for people who have the time and the passion to support something they believe in or see a need for


-John

My views and comments are not to be associated with my employer


-----Original Message-----
From: volt-nuts-bounces at febo.com [mailto:volt-nuts-bounces at febo.com] On Behalf Of shalimr9 at gmail.com
Sent: Monday, December 06, 2010 8:34 AM
To: Discussion of precise voltage measurement
Subject: Re: [volt-nuts] do you like Labview in your labs?

"I can't for the life of me understand why anyone would set up their tent
in NI labview land."

Some people like the comfort it gives to pay for something. The perception that if anything goes wrong, someone will be there to help you. What they don't realize is that most commercial software vendors are resorting to user forums to help users with problem, just like to free software folks have done for ages.

I had a need to learn Matlab quickly to help my son with an assignment (he is an EE student). We were able to share his login to the free on-line tool Matlab provides to universities (a Citrix client, a little cumbersome and slow, but it did work, mostly), but only one of us could be logged at a time. Since he is several hundred miles from me, that was not convenient. So I checked the Matlab web site for a license or a free trial. They have severe restrictions to their free trial, and they have to approve you first. For about $2,000, you can buy a single user license, what a deal! Keep in mind this is for software that is not even aware that Windows has been supporting long file names since Windows 95, only 15 years ago.

So I downloaded and installed Octave, the free Matlab clone on my machine. It ran all the Matlab scripts we threw at it without a flinch. All the error messages (when present) were the same between Octave and Matlab. Octave wants to be a faithful Matlab clone, so they have the same restrictions on long file names by the way, which I found amusing. I can tolerate that from free software, it is inexcusable for expensive commercial software.

I don't understand why the university would force the students to use Matlab (even though they provide a free login, this login expires when the class is over) when quality free tools like Octave are available.

By the way, I don't think Octave will replace Spice any time soon, but it looks like a very useful tool. That was an interesting exercise for me :)

Didier KO4BB

Sent from my Verizon Wireless BlackBerry

-----Original Message-----
From: Chuck Harris <cfharris at erols.com>
Sender: volt-nuts-bounces at febo.com
Date: Sat, 04 Dec 2010 10:34:54 
To: Discussion of precise voltage measurement<volt-nuts at febo.com>
Reply-To: Discussion of precise voltage measurement <volt-nuts at febo.com>
Subject: Re: [volt-nuts] do you like Labview in your labs?

You can also use the same openVISA layer with scilab, python, octave, c++, ...
essentially all other open source languages and packages.  So things are
pretty nice for GPIB users in open source land... much nicer than in 'doze
land.  If only because you don't have to ask anyone's permission to make
changes that make things work better for your needs.

For those that think python is impossible to use, here is a snippet that
will read from your keithley voltmeter:

import visa
keithley = visa.instrument("GPIB::12")
print keithley.ask("*IDN?")

There are many dozens of instruments that have support already in the library,
and adding additional instruments is pretty trivial.

Plotting is as simple as using the python graph function.

Here is another package called pyVlab, which is a python based clone of
Labview:

http://pyvlab.sourceforge.net/

I can't for the life of me understand why anyone would set up their tent
in NI labview land.

-Chuck Harris

Ivan Cousins wrote:
> Some may not be aware that there is an alternative to the NI VISA layer
> that is open source
> (GNU General Public License).
> I use it with GNU octave as an alternative to Labwindows-Labview.
> I like open source tools on linux so they can be changed if needed.
> On the web page it is mentioned:
> "You might be interested that your vxi11 package can be compiled on
> Cygwin/Windows."
> I have not tried this.
>
> VXI11 Ethernet Protocol for Linux at:
>
> http://optics.eee.nottingham.ac.uk/vxi11/
> <http://optics.eee.nottingham.ac.uk/vxi11/>
>
> John C.
>



More information about the volt-nuts mailing list