I am currently in my third year (with funding going to be over by this month). Based on my assignment, I believe that I had been selected for my PhD funding (in the year 2009) for some software verification. In the first year, I was working with a postdoc 'X' and it has worked fine. Again, in the second year I was working with a new postdoc 'Y' (this guy was working on some software release which he did not release until he resigned just 5-6 months back). This new postdoc guy was working on this software (some enhancement over some previous guy's 'Z' work over 15 years) from last two years during his postdoc. The software designed by 'Z' was never validated against any experimental result, nor any commercial software (there are already existing software in the market but are very costly and hence univ. has never purchased these, these commercial software are never available in free manner). Hence in the last 3-4 months back, the software designed by 'Y' (which is based on some previous guy's 'Z' work) was never validated and released for my PhD. I am having my tough time, validating the work done by 'Y' (which never works) which is again based on the work done by 'Z' (which also did not work, somebody in my deptt. had provided the feedback as rubbish software, ever released in the univ.). The result generated by the developed software provides an altogether different result than the experimental result (this included some hardware design and experimental setup), commercial available software (having worked with some of these companies, I have still license left with me and hence I am able to validate).
Now my problem is, I am not getting anywhere with the deptt. designed software (which has already taken about 17years of postdoc experience). I do not understand how some persons' failure can be passed to PhDs. Compounding the issue, the guy 'Z' (who was working over 15 years) have stopped coming to the deptt. just after the resignation of 'Y'.
Can anybody help me what should I do in this case.

Well, from what you've stated, I think you have a wealth of information. You could compare those software and highlight their weaknesses?

Dear Dan,
Thank you very much for your mail. My prof. has also told me in the similar line in the past (It has verified the software - outcome). But the problem was, they wanted to release this software to other universities (since commercial software is very costly). So, they are asking to redo this software (accomplished by guy 'Y') and there is still some flaw in the basis software of the guy 'Z'. 2nd problem is my funding is ending this month and I do not know how to juggle with finance and my task.

Alright, first of all, who are 'they'? Who are the people who told you to do this? Could you clarify? Is it part of your PhD research objective to modify the software? What did your Prof. say about the matter?

Thank you very much for your reply. I am being guided by a group of professors. Releasing the software for univ. academic purpose and across partner univ. was the plan initially and I believe that's why I was appointed with a small living funding. The professors have the confusion (and still there is some belief) that the developed software can not go terribly wrong. Developing the software was never a part of my PhD (this was said to me time and again), but the belief that software can not be so wrong is causing all these issues. This is aggravated by small living allowance (provided to me) and that too is going to get stopped.

Hi,
After some period, I am writing this mail again as I am still facing my hard time. As mentioned in the previous mail, the software designed by Y and Z is not working and it does not agree with the experimental result and commercial software's result (while experimental result and commercial software's result agrees). The problem is, my professors believe that software Y and Z works and can produce better result than this commercial software which is not ((even after lots of convincing, supported by many journals, books, manuals etc.). It is not like that software Y and Z does not work at all but it does not produce better result than the available commercial software (which is bigger cause of the problem) for the agreement with experimental result. These software can take some considerable time (it requires a complete refurbishment). Does anybody think that I can still proceed with my result.

hey stephensen,

the dissension between academics and the industry is never ending. still, there might be a way in shaa Allah to resolve the issue.

is there any part of the software developed by Y and Z which is better than the one commercially produced?
u may have to look hard to find it but it's ur best shot of making this work.

for instance, the softwareCommercial might be better in term of speed, but it may lack the accuracy of softwareYZ. alternatively, it could be lean but it doesn't cover as much functionality as softwareYZ.

to do this, u gotta ask urself ..

why is the software required?
how does the software help the user? how can it become effective?
what does the software need, really need to be effective?
how does softwareYZ respond to this need as opposed to softwareCommercial?
which part of softwareYZ is better than softwareCommercial in a particular aspect, say aspect(A)?
why is it better?
how can it be proven that (softwareYZ > softwareCommercial) in aspect(A)?

a phd is all about contribution. if u can prove that softwareYZ is eligible in making a contribution from a specific aspect, then it could work.

hope it helps. all the best!

Hi Pikirkool,
Thanks for your reply. I am not sure, how to justify that particular portion of the software works better than available commercial software. Like, For example somebody has been said to develop an operating system and the person who developed this, is not sure whether this will stand against any available operating system (Windows, Linux) and he disappears from the univ. (not able to get a contact). Now I know the software is required for any system to run but it can not compete against these operating system (first reason, it has been developed by a good team of people, not by a single person, second reason these are always updated in terms of effectiveness, resource requirement, speed and throughput), except being free. This was also acknowledged by the developer and he left the uni without any contact. And my problem is to show this software works better, which can be only possible either by altering the theory behind this software (as against the standard theory).

hey stephen,

a superior softwareCommercial can't win against all the benchmark of softwareYZ
not in every single case of comparison.
u should consider the paradigm shift in 37signals' getting real philosophy.

there must exist a dimension or aspect where ur softwareYZ reigns supreme.
:) u just gotta zoom in a lot more and find ur focus.
else, all those years of research would be in vain.

despite the progressiveness of the commercial software, it wasn't built out of in depth research,
like what ur faculty did. therefore, there must be a niche not covered by this commercial software.
u gotta find that niche dude.