Should We Use Free Software 4

[emil]

Should we use free software for our consulting projects or study. Is it ethical and safe to use unsupported “black box” software tools from unknown sources? If we are after free software, following the same logic, should we provide free professional services?

 

[dik]

There is a lot of good 'free' software, including OS's like Linux, etc... Not a problem, but with *all* software, you have a professional responsibility to make sure that it is fit of purpose.

 

[dubourg]

I would never do it. It's like geting in a plane that was not tested. It is both an ethical and a safety/waranty problem.

 

[emil]

dubourg I completely agree with you. But, I am surpassed that so many engineers are looking for free software, and I suspect many are using free software in theit projects.

 

[Luca]

I' m not so sure that a free software must be as a non tested plane. Only one example. In Suse Linux professional, you find the z88 Finite Element program. Till nowaday I never used it, but I installed this pakage last week-end.
This software, if I good remember, was developped by a german university. You find the name of the professor and of other people involved in the program.
As it is a free software, they don't give support....
Anyway they have credentials.
Often the manuals of free software are very exaustive...
I found interessting things in the net...

Anyways, in the office I use only expensive software...

Bye
Luca

 

[rlnorton]

Keep in mind that even with expensive commercial software no company stands totally behind the results. Every agreement that I've seen says that the user is responsible for validating the results. The software developers naturally don't want to be sued when your structure fails, whether the problem is the user's fault or a program bug.

With any software, expensive or free, the user should confirm the results by comparison with theory or other programs for test cases that are similar to the analysis being done.

 

[austim]

Fist, I would like to support those saying that ALL software is suspect, whether 'supported' or not.

I clearly recall the trouble that I had during the early development of Stardyne persuading the local reps that the quadrilateral element had a fundamental flaw - after all, what would a mere structural engineer know of such things?

Eventually they reluctantly accepted my proof that there was a bug that produced impossible stress values if any of the nodal coordinates was not an exact binary fraction - so 1.015625 would be OK, but 1.015624 would not!

Now, to the question - is it ethical or safe to use free software for study ? My answer is an unequivocal 'yes : both ethical and entirely safe'.

If the software developers who provide free downloads (eg SAP2000, CadreLite, Risa3D, WinAvansse student edition - generally limited in some way) of 'student' or 'lite' versions do not intend them to be used for study, what other possible purpose do they have? And if the software is being used in the way intended by the developer, how can there be an ethical problem?

As for safety - since no physical product can result from such study, safety cannot be a problem.

Should we provide free professional advice? On odd occasions I do just that, and I'm sure I am not alone. Mostly for marketing purposes (set a sprat to catch a mackerel, as they say).

 

[feafan]

Why not, if your intention is (only) to learn something new (should we kill someone who stole a book from a bookstore?), ‘but don’t tell to no one’. Also, be prepare to experience a new dimension of ‘pain’, when such kind of software shows his true face.
Of course, I also think that usage of this software in ‘serious’ purpose (when you have to stay behind your results and conclusions) could be very dangerous.
But, after you become an experienced user, there is a plenty of reasons to spend a thousands of $ to purchase an authenticated software package.

 

[emil]

I wish my dentist can offer some free samples or service…

 

[ishvaaag]

True, emil..., but some things better be taken in somewhat relaxed ways. I myself make my own sheets in Mathcad for design things; they are I intend rational, but "tested" not all I would call them. If I have something external to calculate, such RISA 3D, I use it; but also mine's...and it is me who has to decide what to put in the plans. Having no cost has nothing to do with the convenience of what programmed; its contents has. And furthermore what told above respect the warrant of the expensive software shows that legal warrant is nil. They may give you support and maybe will...till challenged to enter in a placed lawsuit. So it is all our wits, as always.

 

[DaveVikingPE]

Should we purchase used/old engineering books? Should we even borrow a book from the library and use it in an analysis?

Should we use software at all? Free or purchased?

The ethics are not so obvious, but the way I see it, professionals owe absolutely nothing to the software industry. We do, however, owe it to ourselves to appreciate and purchase the old textbooks and if we cannot afford to purchase them we should take advantage of the local library or rummage sales.

On the issue of software: software producers make money selling software, not using it. And once you're purchased their product, unless you upgrade, they'll make no money off you. Personally, I consider it unethical for software companies to sell a product that "may" have a few bugs in it, but they'll fix 'em on the next upgrade...

Concerning free software, if the tools are available, why not use them? Besides, if you make a mistake, the lawyers aren't going to sue the software company, they're going to sue you, and if you say, in court, "...the computer said it was OK..." you're going to lose your license and become a laughingstock for a while.

There are a lot of very nice spreadsheets, MathCAD, HyperCard and other free engineering analyses software that have been around for years and year. The authors ask for nothing in return, thus, what is the issue? Is it ethical to cut corners using software at all? Is it ethical for software companies to extort upgrades/license fees out of companies?

Is it ethical for software companies to sell the proverbial "bill of goods" with poor documentation, no customer service, etc.?

Is it ethical to use the free "works" programs that came with the computer or should we trash them and purchase a more (sometimes very) expensive program?

My opinion is that it is ethical to use the best tools you can afford to do the job. If one is in "learning" mode or happens to be a student, money is hard to come by, so you use what you can. If you're a professional being paid, you use what the company has. If you're a consultant, you use what's best for you and your customer. For example, if a job requires that you provide electronic drawings in Microstation format, and your office is an AutoCAD shop, you DO NOT give the customer .dxf files...

 

[DaveVikingPE]

I should have mentioned, MatchCAD isn't free, but there are a lot of MatchCAD sheets folks have put together that are.

 

[dozer]

This is off the subject of the original question but since Dave brought it up, here goes. I share his frustration that software vendors release products with bugs in it but don't seem to care since they maintain that it will be fixed in the next release, which you are of course going to buy aren't you?

I sometimes wonder if this isn't just a ploy to keep you updating. I'm experiencing something right now along these lines. I won't use the name of the software to protect the guilty. We have a certain FEA package that we quit updating because a) we're cheap and b) they didn't have any improvements that we really needed.

Recently I've discovered some, what I consider to be, serious bugs. The only way to get 'em fixed? Buy the upgrade of course. Now wait a minute why should I have to pay for somebody elses mistake?

Maybe we should start another thread, but I'm curious of other peoples thoughts and experiences on this subject.

 

[brad]

rkillian,
In dealing with some of the software vendors, I can certainly appreciate your cynicism. I personally know several people who do work or have worked for various CAE companies and I do not believe that there is any "Department of Bug Development" in any of the CAE companies. Bugs are an unfortunate fact of life in complex software development.
However, the way some companies handle their bugs often creates tremendous frustration for the customers. I also won't name companies, but it has been my impression that some companies do definitely have the attitude you have described ("just buy the next release&quot, while some of the others often encourage their support staff to help figure out workarounds for known bugs.
One piece of advice--definitely check the company's on-line (meaning world-wide-web) help section, to see whether their known bugs are listed. Oftentimes they'll list bugs, with possible workarounds listed. This is often useful in avoiding a new purchase of software.

 

[DAVE1]

daveviking
Your comments " the computer said it was ok" could not be more true. During my degree programme the message was drummed in.
cheers
dave1

 

[ishvaaag]

Whether we want it or not, propped or not and consider it proper or not the use of software is to stay. How a designer otherwise would individually check and provide the silly authority with 30000 checks?

So if we still have the common sense of separately checking the main elements in other way is more our wisdom than code enforcement what causes that. Good then have free tools for that.

On the other hand, the checklists required by agencies and codes are doing nothing but to growth without sense, and to one irrational requirement, an irrational answer follows. Nor to say the admonitions of required precision in the calculations in the nonlinear realm for structures; for what I have read, the nonlinear models simply don't predict the behaviour (I have a number of pearls from articles on that saying just so), none of them (er, may be after statistical adjustement when trying-testing something for the aerospatial industry) and so the charge of the codes exacting nonlinear analyses maybe or many times of the unknown class maybe looked with interest as a way of developing our extant knowledge on how the structures function but never as an accurate tool as of today...then resulting that the enforcement of such clauses is just full-of-fear cover-all of potential liabilities more than anything else...to such extent that for so much items in the nuclear industry where the safest analyses are paramount and available we read them deciding the buildings stand for the biggest earthquakes in the elastic realm! So much for non-linear design, today, free or not.