International Support - Italy (request for the next pilot)

[fpagnoncelli]

Hi @bpepple, I'm the Italian user who offered to help populate the database with Italian comics. The International Support discussion (#515) is closed, so I'm opening this thread with the information you asked for.

1. Distribution: two channels, not one. Italian comics are split between newsstands (edicole, the classic periodici such as Bonelli's monthlies and Topolino, with no preorders and monthly cover dates) and comic shops (fumetterie), where publishers like Panini run preorder catalogs (Anteprima, Direct, Preview) with order deadlines and variant covers, functionally similar to the US Direct Market. So Metron's foc_date/store_date and variant fields are relevant for shop-distributed titles and simply stay empty for newsstand ones, while cover_date and price work as-is for both.

2. Translations. No site localization needed. For content metadata, ISO 639-1 it on the planned Series.original_language / Issue.language fields would cover Italian editions (single-language, no regional variants). I support the model changes discussed in #515 (language fields, Publisher alias).

3. How other databases handle it. Following up on the GCD comparison from #515: GCD models Italian publishers with a country field (Italy) and per-issue language (Italian), with prices recorded in the historical currency per issue. Nothing exotic, it fits Metron's model with a country plus a language field.

4. Why Italy fits the gated criteria. Astorina (1962), Sergio Bonelli Editore (1941, home of Tex, Dylan Dog and Zagor) and Panini Comics, publisher of Topolino, Italy's weekly Disney magazine starring Mickey Mouse (Topolino) and running since 1949, are long-established publishers with full metadata available elsewhere (GCD, Comic Vine and the publishers' own archives). I can provide information and data for a pilot.

5. A data consideration: historical currencies. Italy used the Lira until 2002, so long-running Italian series have decades of lire cover prices. django-money already defines ITL (non-decimal), so it's storable, but the current Issue.price field (max_digits=5) caps at 999.99, which lire prices exceed. Worth planning for when enabling EUR: recording historical prices in their original currency (like GCD does) rather than converting.

Happy to provide any further information needed, let me know what would help the decision.

[bpepple]

2. Translations. No site localization needed. For content metadata, ISO 639-1 it on the planned Series.original_language / Issue.language fields would cover Italian editions (single-language, no regional variants). I support the model changes discussed above (language fields, Publisher alias).

So, I did a bit of thinking on the site localization and think it is worth adding support for, since it might be useful for users to use their native language to browse publishers from their country, since based on my brief research, they usually have solicitations in their native language.

Now, since I don't speak Italian, I'm having an AI handle the translation of the *.po for now, so I can do local dev test. Once I've pushed out initial support, I'll setup the infrastructure (Weblate/Transifex/Crowdin) to have native speakers handle it whenever we add new countries down the road.

Now beyond that it sounds like we need to add an alias for for pubisher's name, and add a language field. Does the language field also need to be an array like the alias field?

5. A data consideration: historical currencies. Italy used the Lira until 2002, so long-running Italian series have decades of lire cover prices. django-money already defines ITL (non-decimal), so it's storable, but the current Issue.price field (max_digits=5) caps at 999.99, which lire prices exceed. Worth planning for when enabling EUR: recording historical prices in their original currency (like GCD does) rather than converting.

So, if I'm reading this right, we would need to increase the maximum number of digits? And if so, what would be a reasonable value?

[fpagnoncelli]

Thanks @bpepple, great to see the localization angle landed! Whenever the Weblate-style platform is up, I'd be happy to contribute to the Italian translations as a native speaker.

On the language field as an array: I'd say no, a single value is enough. Following the GCD model: each country edition is its own series under its own publisher (e.g. the Italian "Dylan Dog" vs. its foreign editions), and every issue is in a single language. So Series.original_language and Issue.language as single ISO 639-1 codes cover the use case, and an ArrayField would add filtering/validation complexity without a real scenario. Rare bilingual editions could be a follow-up later.

On the price digits: yes, lire prices (up to ~20,000 lire for deluxe newsstand editions) overflow the current 999.99 cap. A reasonable target is max_digits=8, decimal_places=2 (up to 999,999.99): covers lire comfortably and future currencies like the Japanese yen and Korean won. If you want extra headroom, max_digits=9 also works.

[bpepple]

I've pushed out a very limited slices of the site in Italian (navbar, footer, and issue detail), and with luck I can get the rest of the site done today. You'll have to tell me when I'm done, how well the AI did on the translations, since it's my first time using AI for an actual task.

Now, the changes I need to enable support for Italian comics would be:

• Add alias Arrayfield to the Publisher model.
• Add language field to the Series model. It shouldn't be necessary to also add it to the Issue model, right?
• Modify price fields max_digits to 8.
• Add EUR and ITL to CURRENCIES in settings.py

Does that look correct?

[fpagnoncelli]

Yes, that sounds good. I agree on all points.

Regarding the translations: I took a look at the .po file, and the translations are good.
I noticed a few minor issues (not incorrect translations, but ones that could be improved in terms of phrasing).
Only one string isn't translated correctly: "Future" in the navbar (next to "This Week"/"Next Week"), it was left untranslated, and should be "In arrivo".
Everything else is fine 👍

[bpepple]

What I'll do is (hopefully) push out the rest of the sites translations tonight, and then whatever Weblate-like service we go with, we can fix any errors the AI translations made.

I'm not sure if I'll have time tomorrow to make the model changes we outlined above, but should be able to get to it over the weekend.

[fpagnoncelli]

Sounds great! I'll give the full set of translations a final review once everything is out. And if there's anything else I can help with, just give me a shout.