UI strings not updating - sign of larger problem? Full uninstall

Question

I have a Paratext 9.0.103.41 user who uses the UI in French, but there are some UI strings that are old, sometimes even in English, I think, but for all of the other participants in our recent Paratext course, the UI strings came up fine with version 103.41. Now I guess one would suspect the fr.xml file in C:\Program Files (x86)\Paratext 9\LocData, which I haven’t had a chance to look at yet. But there are other weird things going on on this computer. Like he installed the ShuLatn resource from our course folder, and even though it appears fine on other computers, for him what is supposed to appear like this:

image801×117 27 KB

appears without the hyphens. It appeared fine on the other course computers.

So I’m thinking I might like to do a full uninstall of Paratext 9, and reinstall from scratch. Once I uninstall, I think I might want to make sure to delete some of the folders that Paratext uses, like:

C:\Program Files (x86)\Paratext 9
C:\Users\jeff_\AppData\Local\Paratext90

I’m planning to leave the My Paratext 8/9 Projects folder. Does this sound like a good idea? Can you think of any other places that I should clear out?

Answer 1

I was able to look on the computer, and the C:\Program Files (x86)\Paratext 9\LocData\fr.xml file was older, so I pulled it out of that folder and restarted Paratext - that seemed to fix the old UI element problem. Thanks, anon848905, for the reminder.

And thanks, AWR, for the pointer to check the view in that text collection. The view was set to Preview (Aperçu), where the hyphens don’t appear in the text, but they do appear in the single verse pane to the left (see al-masjuun / almasjuun):

image849×363 102 KB

If I change the view to Standard, then the hyphens appear (along with the SFMs, which is not real helpful):

image852×365 102 KB

So I’m leaving the view as Standard for now, so we can have the hyphens. This happens on my computer, as well as on the translator’s computer. But ShuLatn is a standard resource now, so the developers should be able to replicate the problem. (I’ll report a problem…) These are not special hyphens, just the standard U+002D, so I don’t know why it is doing that.

Thanks for your help.

Comments

That is very strange. I do notice that in Formatted view the hyphens are still visible so that could be another option.

---

Except Formatted view is not an option in the text collection, only Preview, Unformatted and Standard. (And we would generally only want to open up this resource in a text collection.)

Answer 2

jeffh:

I have a Paratext 9.0.103.41 user who uses the UI in French, but there are some UI strings that are old, sometimes even in English, I think, but for all of the other participants in our recent Paratext course, the UI strings came up fine with version 103.41.

jeffh, where you need to look is the C:\Users{USERNAME}\AppData\Local\Paratext90\LocData folder. When Paratext starts imports the data from the C:\Program Files (x86)\Paratext 9\LocData folder to this location. What can happen is that if a user ever opens the localization tool the localization can get frozen here. You can safely delete the contents of the C:\Users{USERNAME}\AppData\Local\Paratext90\LocData folder and it will get regenerated the next time Paratext is opened.

As to the issue of the hyphen, one thing I would check is to verify that his fonts are correctly installed.

Comments

This issue of the multiple copies of localisation files used to be an issue for the localisation team. I assume that the localiser feature can now be removed from PT (if it hasn’t been already), as all the major teams are now using CrowdIn. Unless of course some smaller groups want to localise PT.

As you have seen from other installations, jeffh, PT French should be basically 100% done.

The issue of hyphens in CABRS reminds me of how they used to always be replaced by non-breaking spaces in Preview view and in the right pane of the text collection window. This was fixed in late 2019, I believe (I had some correspondence with John Wickberg at the time).

Answer 3

This problem can happen if the project has hyphen set as a word-break character. Is that the case with this project?

Answer 4

That does seem to be the case. Here is the word-breaking field in the (obviously not modifiable) ShuLatn resource:

But here is that field in the CABRS project, the base project that was used as a basis for that resource:

That change must have been made in CABRS after the language settings were copied to the other project. It’s a bit odd that in both cases, the language listed at the top of the Language Properties dialog is identical:

But I guess that’s my old Paratext thinking kicking in, from when the language settings were shared in the My Paratext X Projects folder. It appears that that language info is actually stored in the (protected) shu-Latn.ldml file that is in the ShuLatn.p8z file in the _Resources folder. And in order to change that, the owner of the resource would need to make that change, and then submit it to the DBL - that would be the mechanism to update the Paratext resource? I think we might have some text corrections for that resource coming up as well, so we might be able to do that at the same time.

Comments

jeffh:

And in order to change that, the owner of the resource would need to make that change, and then submit it to the DBL - that would be the mechanism to update the Paratext resource?

Yes, that is how it would need to be fixed.