Potential misaligned text and image?

#2
by wjbmattingly - opened

First of all, thanks for this dataset. It is truly remarkable!!

I am working within the GUI on HuggingFace as I prepare a video tutorial on this dataset. I may have found a potential mistake here:

image.png

It appears that the text does not align with the image. Notice that the SECUNDUM line corresponds to the wrong image. This could be an issue with the GUI. I am downloading the complete dataset right now to dive in a bit further.

wjbmattingly changed discussion title from Potential mistake? to Potential misaligned text and image?
CATMuS: Consistent Approach to Transcribing ManuScripts org

Hi @wjbmattingly :)

Thanks for reaching out. Yes, we spotted the issue you are talking about. The Caroline Minuscule dataset is the only dataset we realigned (they used cropped picture, we had to put it back within the full manuscript page). Unfortunately, doing so, we messed up some realignment, and we also forgot to CATMuSify some practices.

The good news is, Matena Vlachou-Esthatiou has spotted this, and corrected the transcription. I just haven't had time to recompile (I took some long overdue vacations).

This issue comes with several small others:

  • A few manuscript have the wrong script type (Paris, BnF, lat. 6395; Paris, BnF, esp. 44; Philadelphia, Rosenbach Foundation, ms. 482/2 are all marked as textualis instead of semi-textualis)
  • The Verse/Prose column wasn't properly exported.

This is on my todo list of next week :)

Thanks for the quick reply! That makes perfect sense. I'm coming off a long vacation myself, so I am more than familiar with the feeling =). Looking forward to seeing how this project develops. IMO this is one of the best medieval datasets I have seen in a while!

CATMuS: Consistent Approach to Transcribing ManuScripts org

:)
Just so you know, we have some more manuscripts coming, we are just trying to get things slowly since our preprint is up in review:

  • French Paris, BnF, fr. 20050
  • French Paris, BnF, fr. 2168
  • French Paris, BnF, fr. 12554
  • French Paris, BnF, fr. 5024
  • French Paris, BnF, fr. 574
  • French Paris, BnF, Arsenal ms. 3346
  • French Paris, BnF, fr. 1357
  • French Paris, BnF, fr. 1881
  • Latin Paris, BnF, lat. 13388
  • Latin Paris, BnF, lat. 15176
  • Latin Paris, BnF, lat. 17901
  • Latin Paris, BnF, lat. 12270
  • Latin Paris, BnF, lat. 12449
  • Latin Paris, BnF, lat. 130
  • Latin Paris, BnF, lat. 16085
  • Latin Paris, BnF, lat. 17903
  • Latin Paris, BnF, lat. 8001
  • Latin Paris, BnF, lat. 14137
  • Latin Paris, BnF, NAL 730
  • Latin Paris, BnF, Arsenal ms. 1046
  • Latin Paris, BnF, NAL 632
  • Latin Paris, BnF, Smith-Lesouëf 11
  • Latin Paris, BnF, Smith-Lesouëf 12
  • Latin Paris, BnF, NAL 1909
  • Latin Paris, BnF, Smith-Lesouëf 16

Wow! Very exciting!!

CATMuS: Consistent Approach to Transcribing ManuScripts org

After a long hiatus, I finally got back to the dataset @wjbmattingly .
I will publish a 1.0.1 version of the dataset with "bugfix" level modifications.

CATMuS: Consistent Approach to Transcribing ManuScripts org

The bugfix version is up :) That took way more time than I expected.

ponteineptique changed discussion status to closed

Sign up or log in to comment