Path/File Access Error - Comport.txt

Hello,

I have been unable to run my experiment in MediaLab since Sunday. I keep getting this error message: Path/File Access Error. This message pops up before I can even enter the participant ID and condition. The message also pops up repeatedly, opens up numerous Mozilla firefox/Internet Explorer windows, and makes it impossible for me to exit MediaLab with hitting control alt delete.

My experiment won’t run even after : 1) I re-created the experiment with all 135 of my files TWICE, 2) I transferred my files from my flash drive to the computer’s C Drive and re-uploaded so that the path says C Drive :……., 3) the Psych IT person tried to run the experiment on his computer, 4) after the PsychIT person re-installed MediaLab version 2008 1.13 (I believe) on my computer, and 5) the PsychIT person gave me full rights to the program.

Please advise. I need to resume my data collection as soon as possible. I cannot figure what’s wrong and neither can the PsychIT person who has been assisting me since Tuesday. Apparently, another researcher at my institution complained to the same IT person that MediaLab versions 2006-2008 were problematic for him. Is there a glitch in the program? Please advise so that I can run my experiment. Is there a limit to how many files I can include in an experiment?

[I][edited by moderator][/I]

Hello - Just wanted to 'jump in here' - I am Tim, the 'PsychIT' person Lori is referring to. Within the past two weeks, we are getting an influx of problems associated with MediaLab v2008.1.13. We have v2006.2.40 that I have been installing from...

[I][Tim--sorry I edited here because I got confused by the version info. Is it that you had 2006 on these machines and all was ok, but now that you are installing 2008, you are having this path/file access error?][/I]

The other thing I keep coming across, is a 'generic' license code (99999).
I am not sure if it has something to do with this or not however.

[I][Note from moderator: the 99999 code is used for IP-based Licensing][/I]

I have done much searching throughout the forums - and for most cases, changing permissions to [I]FULL CONTROL[/I] - seems to have helped (e.g., when somone else logs on, and they get prompted for a License Code, even after a proper one has been added to the machine).

The issue that Lori is referencing though has not been corrected by adjusting the security settings, i.e., she is still getting the [I]Runtime Error 75[/I], along with [I]path/file not found[/I]. Now, as an addendum - I have learned from one of my co-workers that, problem could be resolved by re-creating the experiment to the C:\ location...

Thanx!!

Thanks Tim and Lori for the detail. Lori, you mentioned in one of your emails that you are getting the same path error message if you try running one of the basic samples that come with MediaLab, is that right? If so, is that the case with all the machines that you have been upgrading to v2008?

Hi Blair,

The sample experiment is working at the moment although it was not working last night. I just tried to run the sample experiment on
another lab computer with 2008 and it worked there as well. My experiment, however, is still not running.

[I][Tim--sorry I edited here because I got confused by the version info. Is it that you had 2006 on these machines and all was ok, but now that you are installing 2008, you are having this path/file access error?][/I]

To answer this: When I started installing MediaLab on computers here - my predisessor had two versions on our server: 2006v2.39 and 2006v2.40
When I installed, I chose the 2006v2.40. Approximately a couple weeks later, I start getting questions about MediaLab not working. That is when I discovered that MediaLab would initiate an 'M' Code. So following the instructions, I sent an email, and got a response back from John with a License Code. The 'M' code I provided was from an installation I did on 'MY' desktop as a test. So I plugged in the License Code - all worked well on my machine. I then remote accessed into 6 other machines that I had recently installed MediaLab onto, ONLY to realize, that the License Code did not work for any of the others. Another quick email back, and I was informed that each machine needs to get its OWN License Code. Again, no problems there. Then I started getting a few more requests about MediaLab not working in a couple of our Labs, or individual machines - and each case, a License Code was needed - so I would send in the 'M' code; get a License Code - everything was 'OK'. Then a second lab began having access issues. While it worked for either myself, or an 'admin' on the machine, it would not work for anyone else who logged in 'normally'. Before I found the answer in the forums (about changing the Security), I had two more reports from two different Labs on this.
In addition to these reports, there came Lori's problem (path/file not found error), and a second which gave the 'Run Time Error 75, path/file not found' error. Again, it 'seemed' that the problem would be resolved by changing the security.
I have not been able to follow up just yet with the other individual, to see 'if' they are still getting the 'Run Time Error 75' still - but last week Thursday, March 5th, Lori came down to my desk about the problem.
Jumping back a little first - after the report of the Run Time Error 75; that individual, as well as Lori, BOTH reported, that while it didn't work on 'one' computer, it DID work on more than one computer in our Labs (they both used different Labs I believe). This got me looking at the different versions of MediaLab that were installed - and it was 'then' discovered, that on the machines that were having the 'issues', the Actual version of MediaLab was showing as 2008v1.13 (even though, I had installed 2006v2.40 on these machines). The machines in the Lab that were 'working' - were either 2006v2.39 or 2006v2.40 (AND, that is when I learned of the 'IP Based Licensing' code that was being used (that code seems to be used in ALL of our Labs - most of which were installed prior to my starting)).

Back to Lori coming to my Office on March 5th - we spent several hours troubleshooting this. From my desk, I remote accessed into her computer. The version of MediaLab on her computer is 2008v1.13. Her Experiment was Originally put together on 'that' computer - with the path to her USB Jump drive. She was able to run the experiment on the Lab computers. I spoke with her a couple days prior, asking her to get version numbers from the Lab computers; we found them to be the 2006v2.39/2006v2.40. Interestingly enough, on her second 'test'....of two Lab computers, one worked, the other did not - giving the 'path/file not found' error. Coming back to her machine (which 'was' working) - she needed some of her Research Assistants to be able to work on the experiment - so she copied her entire data folder from her USB Jump drive to the local C:\ drive. Thinking that this might be a 'permissions' issue, I made the appropriate Security changes; the only thing this fixed, was the research assistant could not login to MediaLab without it asking for an 'M' code - but Now, the experiment would not work for anyone, even Lori, when she was logged in.
So when she came to my office, I too could not get the experiment to run with MY login (which has Full Admin priveledges on all machines) on her machine. We kept gettin the path/file not found error. So we used the Experiment Editor, and changed to 'Use Relative Path' from 'Use Full Path Names' (Lori had changed the path earlier, after copying over her data files) - but we still could not get her experiment to work. On Thursday, when I tested - the Sample1 experiment worked just fine.
She still had the files on her USB Jump Drive, so we plugged that into MY desktop, and the experiment ran just fine (the version that shoes on my desktop is 2008v1.13). We did NOT try moving the data files to my HD though, as we ran out of time for the day.
I learned yesterday, from my co-worker, that Lori had manually recreated the experiment - and everything was working - however, I received notification this morning, that it is no longer working, nor is the Sample1 Experiment working (which I belive you received some of that info incorrespondence with Lori).

I hope all this info helps, and I look forward to any more suggestions you might migh have.

OH - almost forgot....in one of the other Labs, the research assistant had actually gone out to the website, and downloaded 2006v2.40, and installed it on two machines. Only one of the machines shows as verion 2008v1.13 however. Is there some type of 'auto update' that might be happening?

[[I]Please ignore the above "OH - almost forgot..." I discovered the problem on this one. It appears that the info in our database pointing to the machine in question, was in fact, pointing to a Different machine - so when I was using RDP to access the machine, and making the changes; they were in fact, NOT on the machine in question. Once I finally sat down in front of the machine, to do some checking and 'see for myself' the problem - I then determined that they were different. A quick change on that machine to the Security settings - and all is 'OK' on that one. I may just 're-verify' some things on Lori's PC 'directly' as well.][/I]

Thanx!!

Is there anymore info on this? I know that someone remote accessed into Lori's computer - and had her email a copy of her experiment - but I have not heard of any solution.

Thanx!!

Yes, we did remotely access the computer and found a few issues. First, it was a 2006 installation with a 2008 version of medialab.exe. That was really odd (as if the newest exe was simply dropped into a v2006 ML folder). We uninstalled and reinstalled and had the same issue with the path access error. Turned out that the comport.txt file was causing the error. If we removed that one file, then everything worked fine. I noticed that the com1 port was set to 9600 baud but that the hardware she was using communicating at 19200 baud. We played around with this but I eventually ran out of ideas--aside from trying it on a different computer which I think she said may have worked. In any case, it seems related to the special hardware being accessed via the comport.txt file.

So does this mean, that we have a corrupted download?

The downloaded file I have been installing MediaLab from (which was downloaded before I started in the Department, which was Nov 2008), is:

MediaLab2006.2.40.exe 23,861 bytes 8/30/2007 modified date

How would installing from 'this' file, put a 2008v1.13 version on the computer? I have done some 20 or so installs, ALL from the same file. Am I expected to have the same problem on ALL the computers?

Or do I need to replace the current file with a 2008 version, and use that going forward?

Thanx!!

-Tim

What it looked like to me was a valid installation of v2006 but where someone had dropped/pasted a v2008 copy of the single file "medialab.exe" into the MediaLab folder--overwriting the v2006 equivalent of that file. Are you guys licensed for v2008 or v2006? If you are not sure let me know and I can have our sales manager pull up your license records.

Well...maybe this will answer. I just had a reqeust for a re-install of MediaLab. So this time, I went out to the SITE, and specifically downloaded the v2008, and installed that.

I then sent in the 'M8' code (instead of the 'M6' codes I had been generating); and was given a proper License Code. So far, I have not received any indicating problems.

(Keeping my fingers crossed)

-Tim