Hi Jerzy,

Have you tried preparing the project in 2015 to work on it in 2017?

If you've successfully worked with Epic xlf before in 2015 that might be the easy workaround?

Or, if these are custom xlf files, have you tried exporting their filetype settings file from 2015 and importing into 2017?

Just a few quick random ideas 'off the top of my head'...

Ali :)

What astonishes me really... I tried the same files on my main PC - and they work there. So why does my older PC with freshly installed Studio 2017 and file types resetted not want to process the files, while my new one does?

There is only one empty TM in the project. All files are XLF, so I use the default XLF file type in Studio. Now the analysis runs on my main PC and I will see, what results we get.

Hi Jerzy,

Yes, I deleted my post once I realized you already said

This error appears also, if I switch off the validation in the XML file type

I was still curious why it would work on one PC and not the other though.
What version of Trados are you using on the main PC? Also, are both PCs fully connected to the internet?

Hi Jerzy,

Yes, I deleted my post once I realized you already said

This error appears also, if I switch off the validation in the XML file type

I was still curious why it would work on one PC and not the other though.
What version of Trados are you using on the main PC? Also, are both PCs fully connected to the internet?

I tried again, after having updated the old PC (let's name it "Corsair", this is the network alias it uses) with the latest bells & whistles from Norton and MS Windows 10. This a i7 3500 machine. TM and the whole project are on a NAS. Studio is the current build 14.0.5746.0.

And I got again the same error as above - some files are not recognized.

The other PC (called "Spitfire") is a i7 6700k machine. It accesses the same NAS in my local network.

Both Corsair and Spitfire have a full Internet connection. Both run Windows 10, both are protected with the same version of Norton AV. Studio was installed on both after having uninstalled the previous beta. The only difference is, that Spitfire has had the very latest beta, while on Corsair I had one of the first betas of 2017.

Now Spitfire has finished the analysis and reported 7 errors - all were "Out of memory exceptions". So I think I will run the analysis again in the night, after a PC restart and with only Studio running there. So far we would have a basis for working.

Now the next question:

In customer files we have not only source and target, but also named groups elements. Here's an example:

    <trans-unit id='CAT.ALT.230.1.A' maxwidth='12' resname='ALT.230.1' restype='Category.Abbr' size-unit='x-characters'>

     <source>Overridden</source>

     <target>Overridden</target>

     <context-group name='Item Information'>

      <context context-type='x-info'>Category abbreviation in item [ALT-230-ALERT STATUS] for category title
       "Overridden". Item description: "alert status"</context>

      <context context-type='x-length'>12 characters</context>

     </context-group>

     <note>• INFO: Category abbreviation in item [ALT-230-ALERT STATUS] for category title "Overridden". Item description: "alert status" LENGTH:
      12 characters STRING ID: CAT.ALT.230.1.A</note>

    </trans-unit>

We have to make sure, the translator can see this additional elements, while this elements do not need to be translated. Studio ignores these completely. When I would create a xml file type with only "target" as translatable item, could I declare this tags above as comments? Even if I know how to create a file type I do not yet know how to declare elements to be just comments in Stuido.