David McDermit

The most likely reason for the files not uploading to Articulate is that you are including segmentation information in the XLF you return and they have no mechanism for handling this.  Make sure this option is switched off when you prepare your projects:

On the freelancer issue... could be related to merging/splitting segments. Try preparing the project again, don't split/merge and pre-translate from the TM.  See if that helps.

Hello Paul

I heard back from the freelancer.

They did not merge/split segments and the box for add segmentation in the XIFF settings was not checked.

David McDermit

Can I see the XLIFF files you’re working with? The source XLF and the translated target? You can send them to pfilkin@sdl.com if this is possible.

Paul, thanks.

Sent you the files and hope you can see something.  Articulate xlf files can be picky.

David McDermit

Thanks for sending me the files.  I have no problem using these files. Using the default XLIFF filetype or using your custom settings open the file.

The only thing I would note is that these source files are not fully segmented and this means you are expecting the CAT tool to complete the segmentation for you.  This opens up the possibility for problems:

• Studio may segment the file in a different way to that expected
• whether or not segmentation information should be added in the target file needs to be clarified as this could also lead to problems
• even simple things like target language codes could be a potential problem

Normally I think it's much safer to provide fully segmented XLIFF files for translation and then you avoid some of these problems.

David McDermit

Thanks for sending me the files.  I have no problem using these files. Using the default XLIFF filetype or using your custom settings open the file.

The only thing I would note is that these source files are not fully segmented and this means you are expecting the CAT tool to complete the segmentation for you.  This opens up the possibility for problems:

• Studio may segment the file in a different way to that expected
• whether or not segmentation information should be added in the target file needs to be clarified as this could also lead to problems
• even simple things like target language codes could be a potential problem

Normally I think it's much safer to provide fully segmented XLIFF files for translation and then you avoid some of these problems.

Paul

We had tried using these files with out that XLIFF filetype custom setting.  It leaves a lot of coding in Trados and that confused the in-house and freelancers just as much.  We ended up with files that had material translated that should not have been and that obviously created upload issues.  Thus the custom settings.  However, with these settings we have had actually fewer upload issues.

Also, these does not seem to have a good pattern.  Same translator could be working on multiple xlf files for a project, like this one.  And some would upload fine and some did not.  In this case two out of three uploaded file.

Articulate has two platforms.  One does put in target language codes and the other does not.  The files for this project does not.  The other puts coding throughout the xlf file.

David McDermit

I think the only way you're going to get to the bottom of this is to work with your client and try to find out exactly what's wrong with the files that fail. They could probably split the files up if it's not obvious and this way you may be able to narrow down where the problem is coming from.  What you need is a couple of segments in the file at most.  Then when it fails it'll be easier to diagnose.

Hi,

Just my 2 cents. This actually is a common problem with Articulate cloud platform. What we would do to fix is rather than trying to fix the current file, we'd just export another XLIFF, prep and translate it using TM, and import it right back. With the Cloud platform, any changes made to Articulate application itself might prevent an old export from getting imported. I hope it helps.

Thanks,

Naoko

thanks Naoko Jones, that's helpful.  David McDermit worth a try.

Paul

It might be possible to break up the file, but almost as challenging as going through this single file to find errors.  This is an online learning development program that is more of a glorified PowerPoint presentation.  We did try  this once and breaking down to each version of what it considers a slide did not bring any real help.

I think I mentioned before, but the error message it gives is also of no use.  It does not specify what the error is or where it could be in the file.

Naoko is correct.  Is a common problem and Articulate itself is not that helpful.

Thank you for taking the time to review this one.   As always, appreciate your expertise.

Naoko

Thanks.  This is in our list of fix options. 

Had tried this option and unfortunately the TM left a large number of segments empty after pre-translation.  Scrolling through the file did help fill in a few more segments, but because of some of the tags articulate puts within does seem to hamper the TM and filling in all segments.

Hi David McDermit,

Doesn't that sort of confirm that we do need to use the new export and go through each empty segments and finish "translation"? It might mean a lot of copy/paste and adding missing inline tags, but I don't see any other way of fixing it unfortunately.

Naoko

Yes, unfortunately it does mean a lot of line by line fixing.

At least articulate is finally realizing that English is not the only language these module we be put in and that people are using more than online translation programs for this.