www.cmm.bristol.ac.uk/forum

Hi,

I know this has been discussed before, but I could not find any help in the previous discussions. Perhaps someone has a suggestion what could fix the problem in my case. When trying to run runmlwin using Stata 14 on a MacBook, I get:

The model did not run properly in MLwiN. You most likely clicked the 'Abort Macro' button in MLwiN, rather than the 'Resume Macro' button.

I clicked no button at all. The path has been set to: Any idea what the problem could be?

Note that the datafile I am using is fairly large (~115 MB).

Thank you in advance for any help.

Unfortunately Mac OS X does not have a built in mechanism for automatically invoking Wine when you run a Windows executable, so you will need to write a wrapper script as described in http://www.bristol.ac.uk/cmm/software/m ... .html#unix and set the MLwiN_path to that instead. Alternatively, if you don't need the GUI interface, we now provide a native Mac OS X version of mlnscript (available from http://www.cmm.bris.ac.uk/clients/softwaredownload/) which you could set it to instead.

Thank you for your kind response.

I have tried to set up the scripts as suggested on the website, but I cannot find any x64 subdirectory? Installing the other two scripts does not fix the problem. I have updated the path to:
When executing the script from the terminal, it returns:

wine: cannot find L"C:\\windows\\system32\\mlwin.exe"

The website you referred to also mentions "mlnscript.exe", but that is not provided, at least in the trial version, or?

I would be fine testing the script only version, I do not need the interface, but as far as I see it is not possible to test it before buying it (I am not based in the UK). Of course, I want my software to be legal, but it would be great to test it before buying it - so that I know that it actually works. I know MLwiN quite well, so it is not that I need to be convinced that the programme itself is a good programme, I do know that.

MLwiN startes fine under Wine, I can also read data - but it either quickly crashes or becomes so slow as if crashed?

Best

Sorry, I didn't realise that you were running the trial version. We do not distribute the scripting components with this version, which would explain why you couldn't find the x64 subdirectory or the mlnscript files. For some reason the current directory isn't being set correctly in your mlwin.sh script, however if you change it to the following then it should work: where XXX corresponds to your user name.

I'm not sure why you are getting crashes under Wine, if you could provide any error messages that are displayed when this occurs it would be useful

Thank you for your reply. Writing this made the script work:
The error message is still the same, however.

"The model did not run properly in MLwiN. You most likely clicked the 'Abort Macro' button in MLwiN, rather than the 'Resume Macro' button."

Could a problem be the trial message at the beginning? The path is correct, it works.

How can I retrieve any error message from MLwiN. It is not crashing, it just hangs forever.

Best

Yes, that is possible - does it display the trial message when you start it? We use a third-party wrapper to supply the trial functionality and it may be that something within that isn't fully compatible with Wine. One thing that you could test if you are just interested in how MLwiN runs under Wine would be the teaching version (available from http://www.bristol.ac.uk/cmm/software/mlwin/download/). This only works with specially tagged example worksheets, and therefore isn't usable from runmlwin, however otherwise it works identically to the full version of MLwiN.

Yes, the trial message is shown when I start MLwiN myself in Wine--but that has nothing to do with why it crashes, I presume, it is no problem to click the message away.

What I want to know is whether and how MLwiN runs from within Stata. Do I understand you correctly that this is not possible to test with the trial version, i.e. I would have to buy MLwiN without knowing whether and how it will work?

Yes, this is unfortunately currently the case. If you send me an email (you can find my details on http://www.bristol.ac.uk/cmm/team/) then I should be able to provide you with something better for you to test.