Two Japanese users have reported that some files they upload to Workspace have had their fonts changed.  One was about a month back and offered few details (except, Excel seemed fine for them).  The latest mentions the fonts coming out in a totally illegible form.  For example:

½�¬‚Q‚O”N“xŽ–‹ÆŠT—v

Has anyone encountered a similar problem?  Or, perhaps issues with diacritics or other non-standard marks not appearing normally, etc.?  Any guesses about what might cause this? 
 

This appears to be rendering entire Japanese text files unreadable once they arrive on Workspace. 

 
Any suggestions are much appreciated.

Do they see this problem only with HTML documents? Currently you have to encode the characters using UTF-8 for HTML documents to render correctly.

Thank you for the reply. 

This has only been mentioned as occuring after upload to OLW.　 Would it be useful to ask that the user downloads the same file and opens it again to check.  Should the conversion problem go away again upon downloading if it's only about the encoding?

I have an English system and so am entering Japanese using IME, myself.  My own Japanese language documents seem to upload to OLW fine (in Word 2000 from Win Home XP 2002 SP 2).  I see the encoding choices in the browser View menu, but only for reading pages.  I did flip through those views looking at what was sent and it didn't translate. 

Could you tell me how most Japanese users can control for the encoding when they either create or send the file, as needed?

Assuming this is the known issue about HTML files I was talking about the file should be readable once downloaded.

If you click on File -> Save As... option in Internet Explorer you can see a dropdown with "Encoding:" caption on the left side. If you then choose UTF-8 the saved file should be readable in OLW.

I appreciate the suggestion and am trying to work through it.  What type of file do you save it as, exactly?  Or, do you need to go to another section of OLW to open it?

Also, is there a menu someplace to prepare Japanese files during upload, rather than converting them inside OLW?  I am guessing the Japanese version  applications would have this, if someone can check one out.

The workaround is only for HTML files. Please let us know if you find out that Japanese users experience problems with reading other documents, e.g. Office Word docs.

The conversion basically happens during the saving operation (for html documents) in the browser - when you see a website you want to save to disk in IE you can choose (via the save dialog) the encoding to use for the file. If you choose UTF-8 you should be able to upload the created file to OLW without any additional conversions and it should render correctly.

Hmm, both of the Japanese users refer to "text files." 

At first I thought it was Word since the first post made a point of contrasting it with a successful upload using Excel.   However, if they're using the term per the jargon, I guess it's simply .txt files or something similar.  I've been asking for application details, but so far they have not provided those.

Thanks for the feedback.  Posted the HTML fix to the Japanese forums.

I'll chirp if they give some more information.

Another user added to this topic on the Japanese side with clearer data.  In this case, it was confirmed that a file saved from Notepad in the default form turned illegible once uploaded.  The user says the default encoding for Notepad in Japanese Windows is ANSI, so in order to make these files legible, they must change the encoding each time they save.

I wonder if it might be possible for someone to confirm which commonly-used applications require that step upon saving.  For instance, does this regularly apply to some or all of the Office applications on Japanese Windows?

Also, are there plans to correct so that the workaround will not be necessary in the future?

We also have another issue involving Japanese fonts, but this one only affects a subset. 

In Japanese, there are times when users wish to write in notations alongside Chinese characters.  These are called furigana.  They usually appear in a little line above the regular script, in a smaller size.  A user has reported that where furigana were used in a document, both the furigana and the regular script are simply gone.  The rest of the Japanese script is intact on Workspace and there is no blank space where the missing characters should be.  It's just not there, and the surrounding characters are adjacent to each other as if it hadn't existed.

This was in Windows XP, IE7 with MS Office Live Update.  The file is from Word 2003. 

Once again, any suggestions are much appreciated.  Cheers.

 
This is based on a pretty aged web forum (written in 2004 dealing with Japanese Word 2000), but it looks like there may be at least a partial fix for disappearing Japanese furigana (and disappearing associated characters too).  Here is the link, but I'll sum up the pertinent portion below.