Hyperides work numbers (tlg0030)

[helmadik]

https://raw.githubusercontent.com/PerseusDL/catalog_data/master/citecoll/Greek.csv has ordering following TLG canon, which in turn follows the Teubner edition: 001 = against Demosthenes 002 = defense of Lycophron This was also followed in the LSJ markup, way back when.

But https://catalog.perseus.org/?utf8=✓&utf8=✓&per_page=100&search_field=all_fields&q=Hyperides has 001 = Lycophron, 002 = Philippides, etc.. following the Loeb edition ordering, against the Teubner. I hope that Greek.csv represents what the data are going to be presented as.. For now I will go with that assumption. I attach the overview from the Loeb volume. Thanks in advance for clarification!

[AlisonBabeu]

Hi @helmadik you are quite right regarding this discrepancy. The Greek authors.csv file was generated from a spreadsheet (and hasn't been updated since the first catalog data generation back in 2016) and at the time it was generated followed the TLG Canon.

As I was later creating catalog metadata for Hyperides I discovered that the canonical identifiers that had been used for the Hyperides editions in Perseus were not the same as the TLG numbers, but instead had followed the Loeb numbering. We discussed this issue here a number of years ago (https://github.com/PerseusDL/canonical-greekLit/issues/26) and ultimately decided just to stick with the work numbering used in Perseus since changing them to match the TLG IDs would have made us have to destabilize a lot of identifiers that had been used to create CTS-URNs and all the links that had been made to them.

[helmadik]

Thanks so much for the quick response, Alison! I had checked my version of the LSJ, and given that the numbering scheme is tlgxxxx.tlgxxx with no indication that there is going to be a difference, I guess I'm on the other side of the earlier consensus. I like how with Gorgias Helen and Palamedes the 1st1K collection uses a tlgxxxx.1st1Kxxx number to make clear they're not following the tlg canon.. But I can see how this is influenced by the sunk cost - my old old Perseus secondary sources still have the TLG numbers so for me the cost goes the other way:-(. For third-party users, -or just speaking for myself:-) - it would be really helpful to indicate in some way when Perseus is deviating from the tlg standard, like the 1st1k project does with Gorgias.