Parse glosses missing in Pt 9.2

Question

Since I upgraded to Pt 9.2, I can’t get the data I had previously entered (over several years) to show up in the interlinearizer.
The parse glosses are not showing up, especially the grammatical morpheme glosses. Some of the roots are correct, but not everything.
I use this tool all the time, so I’m really handicapped now.
I saw a previous discussion about setting up the Interlinearizer, but I don’t think it refers to this particular issue.

I have uninstalled 9.2 and went back to 9.1 in the hopes that the interlinearizer would work as it did previously, but alas, I still don’t have access to the parse glosses.

Answer 1

anon570831, and anyone who stops by with the same issue:

Your interlinear data is stored in the database and tagged with the language of the glosses. You can look in your Paratext project directory to see what language code(s) your existing interlinear has been tagging the glosses with. Paratext creates a subdirectory to store the tagging information that is called Interlinear_XYZ where XYZ is the language code. You may have more than one:

It is common to have more than one folder. You may have glosses in English in one project and in Spanish in another project. But when you have one or more variants of English (en-US, en-UK, en), these are also stored separately, even if they are all English. Having two english databases, as indicated in the image above can indicate a problem.

If you have two interlinear_XYZ directories and don’t know which is the language code you have been using, just open each directory and see which one has more files, and with past dates. In the screenshot at top, the Interlinear_en folder had only one file in it that was only 500bytes in size. The other one had multiple books with larger file sizes. Clearly, en-US is the language I need to select when setting up the interlinear if I want to see my glosses back.

Let’s say that all your glosses went missing after upgrading to 9.2 and reopening the interlinear. If you haven’t worked yet in the new interlinear, but noticed your glosses were missing, you will probably only have one directory showing here, lets say it is en-US. This is the one with the language tag you have been using. To see your glosses again, you will need to create a new interlinear in one of two ways:

1. Create an interlinear with no model text, and select en-US as the language.
   OR
2. Create an interlinear with a model text that is en-US. You can see the language of the model text in the Open Resource dialog:

Either way, you will see your missing glosses when you re-open the interlinear.

Comments

Hi dhigby,
Thank you so much for explaining how to do this safely.
I have managed to upgrade to Pt 9.2 and reload all my data for the interlinearizer!
I’m very grateful for your help, thank you!

Answer 2

Please use Help > Give feedback so we can take a closer look.

Comments

Hi, yesterday I was able to communicate with my team exegete and together we tried to resolve the issue. He had the correct parse glosses. I reverted to Pt 9.1, then undid recent changes that were related to the interlinear, then did a send/receive and thankfully, everything returned to how it had been before. I took a screenshot of the wrong parse glosses, as well as the correct ones. I’ll try and upload them using the help>give feedback so you can see what the issue was.

Answer 3

I’m not sure how best to get the 2 screenshots back to you?

Answer 4

Unfortunately, the screenshots probably won’t help us track down the problem. The reason we have people use Help > Give feedback is because it includes log files and other information that help determine what causes the problems the users see. Since you have already worked around the problem, sending us anything probably won’t help, sorry. If something like that happens again, please use Help > Give feedback before fixing the problem so we can try figure out what is causing it.

Answer 5

Okay, thanks for clarifying.
I did originally go through the help>give feedback route (PTXS-30124) and Steve White answered me, but didn’t seem to understand what the problem was. I responded further to him twice but he didn’t answer again, so that’s why I posted something on the support forum in the hope that somebody else may have run into the same problem and have a solution. Maybe you can find the original screenshot that was attached to the first request for help?

Answer 6

Can someone please clarify… Is the expectation that someone updating from PTX 9.1 to 9.2 will retain all of the work that they have done in the interlinear tool? Is there some specific setting that they need to choose in the new interlinear interface to make that happen? It would be helpful if someone familiar with the design could comment on how former users of interlinear can maintain that work after the upgrade.

Comments

Let me try to explain.
Paratext 9.2 takes more control over the interlinear. There were some outside cases that are causing problems when interlinearizing. One I know of is when the source and target are both set to the same language. So if there are bad settings in Paratext 9.1, you may have issues going to 9.2.

Also, because you have to reopen the interlinear after upgrading to 9.2 (even if it was open in 9.1), there is always a possibility someone selects a different model text or language than was previously used. If the language code is different (en-US vs en) then you will get an empty database. This isn’t new… Happened in 9 all the time. But since people have to walk through the process again, after their interlinear is closed wrong configurations happen.

Their previous configuration should be the first thing that they see when opening the interlinear, but there are other options that can lead to an empty dataset with a different language ID. It is usually an easy fix for my team. However, the shock of missing data, combined with failure to understand how Paratext grabs the language code from the model text makes it difficult for users to resolve on their own.

---

Hi dhigby, thanks for explaining this further.
I would still like to upgrade to Pt 9.2, but obviously didn’t understand how to set up the interlinearizer correctly last time I tried to after upgrading. I thought I had done it in the same way as previously in 9.1, but obviously not. Could you give me further instructions so I can set up the interlinearizer correctly and have access to my data. Thanks.

---

I would be happy to help you, if this is still an issue for you. Contact me at [Email Removed]

---

Hi anon758749, thanks for being willing to help. As you see below, I followed dhigby’s instructions and have been able to set up Pt 9.2 with the interlinearizer fully functioning as in 9.1. Thanks anyway!

---

Excellent, I had not seen dhigby’s reply.

Answer 7

My project (CKO) has two Interlinear directories, and I can identify the one that I need to select in order to follow dhigby’s reply with the two options for restoring glosses. However, option 1 resulted in empty glosses. Option 2 had only guesses to the Parse glosses and Word glosses. I did not see the my previous glosses for Parse, Word or Phrase. So I am in the state of “shock of missing data” and unable to resolve the issue on my own. What do you advise? -anon927678

Comments

@anon927678 Did you get this figured out? If not, please let us know and we can try to follow up with you.

---

Thanks for reaching out. No help recd yet. I m out of town this week. Would you be available on M 6/20?

---

Good news! I found a file that had somehow been deleted and it looks like most (if not all) of your parses are there. I need to consult with our team to know how to properly restore it. I can connect with you next week to work on it with you.

We could also probably merge the 2 directories, so all of your glosses will be in 1 version of English instead of 2. Do you have a preference for which translation you use as a model text? The current options are en and en-US. You can see which translations are which English version by doing this:

1. Go to the Paratext menu > Open
2. Select Resources only (#1 below)
3. Filter for English (#2)
4. Sort by the Language column (#3)
5. Look at the versions below (#4)

Here’s an example of what I have on my machine (you may have different translations, and can get more with the Paratext menu > Download/Install Resources):

image853×488 42.1 KB

Answer 8

anon927678,

I just want to cover the basics. There is a toggle for showing parses. Since you had to setup your interlinearizer again after the upgrade to PT9.2, verify the settings in the view menu. Here is a screenshot:

image1347×473 120 KB