Re: -- Wolfram Research QA process defect
- From: Dave <someplace@xxxxxxxxxxxxxxxx>
- Date: Tue, 04 Mar 2008 10:56:35 +0000
Martin Rubey wrote:
Dear Dave,
Dave <foo@xxxxxxx> writes:
Have you considered open-sourcing your code? Parallel processing could be
used by getting a large number of people to hunt for bugs, in the same way is
projects like Seti
I agree with some things you say, but on the other hand, I think that Wolfram
and MapleSoft could just as well be grateful to Vladimir.
I suspect some individuals in Wolfram Research and Maplesoft are grateful. I suspect there are others who just wish he would not exist.
Or did he so far
discover only bugs/shortcomings that Wolfram knew about anyway?
Only someone from Wolfram Research could answer that, but it seems unlikely they know about all the bugs he finds. I suspect he finds some they do not know of.
I believe one of Vladimir's gripes with Maplesoft was that the same bug would exist for a long time, even after he published it. I get the feeling there are a number of employees in Wolfram Research who do read sci.math.symbolic and do act on the bug reports. Vladimir did for a number of years repeatidly report bugs in Maple and left Mathematica alone. He seems to have switched tactics now.
Finally, why doesn't Wolfram open source Mathematica? Yes, they want to earn
money. Even though Vladimir has bad chances, it seems to me that he would like
to earn money, and so far he chose not to do it via the open source
route. (Although I believe he would be better off doing so.)
As I am sure you know, open-source and making money are not mutually exclusive. But acting in a very unprofessional manner and making money are likely to be mutually exclusive. That I suspect is Vladimir's biggest problem.
I've no idea, but perhaps Wolfram Research and Maplesoft feel the software they use for detecting bugs is better than Vladimir's. The fact he finds bugs they have not found does not prove his software is any better than theirs. Testing the software must be very computer intensive, so no matter how much testing you do, someone else may have tested some other combination of parameters.
I suspect there would be some who would like to help Vladimir in his quest to find bugs. He could do that by open-sourcing, or perhaps producing binaries for different platforms. If someone uses Mathematica a lot for integration, they might be willing to run tests on the integration functions when a machine is otherwise idle. SETI does not release the source code for the clients - only binaries.
Another possible option is to release the code but under a licence which prohibits companies such as Wolfram Research and Maplesoft using it for bug hunting without paying a licence fee, but allow open-source programs such as Sage to use it for no cost. This could result in greater improvements to open-source software and might result in extra income for Vladimir, but I still suspect his greatest problem is the way he reports the bugs.
Finally, I must say that I find it rather surprising that people choose to work
for Maple / Mathematica without getting paid. In fact, they even pay
themselves! Yes, I know, the free and open source alternatives are not as good
with respect to some important features (in other cases, they are better,
though). But why should we promote expensive, closed source software
unnecessarily?
I'm not sure "promote" is the correct word here. One does not normally promote software by proving it is buggy. It might well have the opposite effect and make people look at open-source programs like Sage. A number of people use Linux as Windows is so buggy.
.
Martin
- References:
- Re: -- Wolfram Research QA process defect
- From: Martin Rubey
- Re: -- Wolfram Research QA process defect
- Prev by Date: Re: equation
- Next by Date: Re: equation
- Previous by thread: Re: -- Wolfram Research QA process defect
- Next by thread: equation
- Index(es):
Relevant Pages
|
