Disputatio Libri:Cornelia Mima Maxey.pdf

Hold off correcting from page 15-49

@Persimmon and Hazelnut / others Heads up that I will "macronise" pages 15-49 in one pass, via search and replace or a macroniser tool, once I have corrected the non-macronised text. So please leave those pages as is until this is done :)

You are all welcome to check any page up to and including page 14 of course :)JimKillock (disputatio) 15:43, 15 Augusti 2022 (UTC)

OK. I actually have the vocabulary for Cornelia proofread and on my computer, so hold off on pp. 59-78. Persimmon and Hazelnut (disputatio) 16:05, 15 Augusti 2022 (UTC)

Transferred. Go ahead and proofread the glossary (pp. 59-78) if you are so inclined. Persimmon and Hazelnut (disputatio) 16:49, 15 Augusti 2022 (UTC)

What are you using for your OCR step? Does it really not see macrons at all? Persimmon and Hazelnut (disputatio) 18:42, 15 Augusti 2022 (UTC)

I am - as I mentioned above - ignoring macrons as there are multiple easy ways to add them in bulk in a single automated step across all pages when all is done. The default scan is very accurate for the text (it jumbles it a bit) but does not contain macrons or other diacritics. If you leave them out you can work more quickly and we can complete this job in less time, but if you put them in, that's up to you, it won't matter.

BTW doesn't it warn when someone is editing a page? I just had an edit conflict as we worked on the same scan page. --JimKillock (disputatio) 21:17, 15 Augusti 2022 (UTC)

In any case you seem to have made short work of that! Thank you. I'll make a check for accuracy and run it through a macroniser. What are you using for OCR? --JimKillock (disputatio) 21:29, 15 Augusti 2022 (UTC)

I'm using the iPad app "Text Scanner (OCR)". It seems to miss fewer characters than Wikisource's default scans and it's pretty good at noticing when a letter has a diacritic. (Other than “i”—distinguishing i and ī is apparently very hard.) (It is not so good at noticing which diacritic—I have a postprocessing step where I convert all the â/ã/å/äs to ās.) I noticed that you missed a few macrons on pp. 1–14 and I thought it would be easier to just add them in at the start rather than the end.

Is there a way to do a find/replace across multiple Wikisource pages at once without having to edit all the pages one at a time? Persimmon and Hazelnut (disputatio) 21:39, 15 Augusti 2022 (UTC)

I don't think so; I would simply create a text document with pages 15-49 and run it through a macroniser, and then place it back in. This seemed the quicker way although with your version it's not as helpful as they are mostly already there. --JimKillock (disputatio) 21:54, 15 Augusti 2022 (UTC)

OK, I ran all the pages through this Macroniser; it is also a useful spell checker. The pages do still need a careful read; I have spotted things that are missed. In particular the Macroniser can't tell a lot of ablatives from nominatives in 'a' declensions. However we've made short work of this, I'm quite impressed by your scanner! --JimKillock (disputatio) 16:27, 16 Augusti 2022 (UTC)

Part of the reason I did my part so fast is that I'd been planning on uploading this in 2029 (because then it will be in the public domain even if I missed the copyright renewal), so I had the first third or so already proofread and had done a lot of find-and-replace searches to macronize the later parts.

I've also got fair-sized chunks of Carolus et Maria and A New Latin Primer partly proofread/macronized, so let me know if you decide to start those ones. Persimmon and Hazelnut (disputatio) 17:47, 16 Augusti 2022 (UTC)

OK, will do. The main obstacle is finding / recording the copyright records search, which is a bit of work but not too much. I'll try to do this in the coming days so you can add in your work. --JimKillock (disputatio) 21:01, 16 Augusti 2022 (UTC)

Images

@JimKillock: There are certain images that appear in the middle of a paragraph. I am inclined to think that putting them on the wrong page (the next page) is better than introducing new paragraph breaks. (The real answer probably involves getting floating images to work, which involves messing with the CSS in ways I'm not sure how to do.)

https://en.wikisource.org/wiki/Template:Img_float#Image_in_centre

Thoughts? Persimmon and Hazelnut (disputatio) 21:41, 16 Augusti 2022 (UTC)

I think it's painful / confusing to future editors or checkers to place the images on the wrong page within the book, but I also agree it is best to avoid new paragraph breaks where possible. Sometimes the images are mid-sentence etc, so there is simply no perfect way to do this.

We can use {{pt|text|text}} to hide text if we need, which might be a good workaround. Let me know where there is an issue and we can see what the best workaround is. --JimKillock (disputatio) 21:45, 16 Augusti 2022 (UTC)