Update Italian translation (RPU)

[NarakuITA]

Italian translation updated based on the local version and converted using msg2po, as suggested. Italian text corrected; English msgids left unchanged to maintain correspondence with the template.

[burner1024]

Why'd you drop all the accented characters?

[NarakuITA]

Why'd you drop all the accented characters?
Hey! Good question — the missing accents are intentional, not a bug or a loss of quality. A bit of background:Engine limitations (originally): the original Fallout 2 engine and its fonts were not built to display accented characters reliably. Depending on the setup, accented letters could show up as garbled or missing glyphs in the dialogue boxes. That's why the historical convention for the Italian translation was to replace them with a plain apostrophe: e' instead of è, a' instead of à, and so on.The original Italian translation: Ragfox's first Italian translation of Fallout 2 followed exactly this convention. For Italian players, that apostrophe style is the voice of Fallout: "E' una specie di casco..." reads as natural to us as proper typography would to anyone else.Consistency first: even if later patches (RPU/sfall) may have improved character handling — I'm not entirely sure of the current state — I chose to continue the original philosophy for this update. Mixing accented and non-accented sentences in the same dialogue would look inconsistent, and suddenly switching style after twenty years would feel like a different translation by a different team.So in short: no accents = deliberate choice, in line with both the engine's original limitations and the tradition of the Italian translation. If the project ever decides to move to full accented typography, it should be a project-wide decision (and with the new .po pipeline it would just be a one-pass conversion). But until then, the apostrophe style remains the canon for Italian Fallout.Hope this clears it up!

[burner1024]

I agree that inconsistency is not great.
However, I think that greater fidelity of accented characters in general is preferable (and that's what other languages use).
Character display does not depend on sfall. It's purely a font file.

I'm not entirely sure of the current state

So let's be sure first. What is the current state? Do the accented characters display properly in Italian? Do you have numbers (how many accented vs non-accented there are now)?

Secondly, what's this about? There are multiple changes like this:

Is that actually what you intended to do?

[NarakuITA]

I agree that inconsistency is not great. However, I think that greater fidelity of accented characters in general is preferable (and that's what other languages use). Character display does not depend on sfall. It's purely a font file.

I'm not entirely sure of the current state

So let's be sure first. What is the current state? Do the accented characters display properly in Italian? Do you have numbers (how many accented vs non-accented there are now)?

Secondly, what's this about? There are multiple changes like this:

Is that actually what you intended to do?

As for accented characters, they aren't displaying correctly in the game—at least not for Italian.
It’s been like that from the start. An example? Among the various unreadable characters, they appear like this: è
Regarding the special character "-", I don't know why it turned into a "?"; I never touched it. I'll check right away.

[NarakuITA]

Hi, two follow-up notes on the Italian translation PR:

1. About the "?" characters you spotted (e.g. "Shi?Huang?Ti", "Jella?Slot")
   Those were not intentional edits. During a local cleanup session the .po accidentally went through a UTF-8 → ANSI (cp1252) → UTF-8 round-trip in a text editor. Characters that don't exist in cp1252 — in practice the non-breaking hyphen U+2011, used in strings like "Shi‑Huang‑Ti" and "Jella‑Slot" — were silently replaced by "?" at that point, which is why lines I never touched showed up as modified in the diff.
   I then ran a full scan of the file against the pristine Weblate export: 46 entries were affected. All of them have been restored byte-for-byte to the original characters, so those lines are now identical to the base and should disappear from the diff entirely (see the latest commit on this branch).
   One project-wide consideration, while we're here: U+2011 (and a couple of other exotic characters) can never survive in the cp1252 .msg files anyway — unpoify's cp1252 output replaces them with "?" regardless of what the .po contains. It might be worth defining an explicit mapping in the chain (e.g. U+2011 → "-") so that the .po and the generated .msg stay consistent, and reviewers don't get phantom diffs in the future.
2. game/pipboy.msg is missing from the .po chain
   While rebuilding the Italian .msg tree I noticed that game/pipboy.msg is not covered by the .po at all: it has no entries, and unpoify never regenerates it. Older Italian dats (and my local build) carried a translated pipboy.msg, so the Pip-Boy UI texts do exist in Italian but live completely outside the Weblate/po workflow. In my community preview dat I kept the old translated file so the Pip-Boy doesn't fall back to untranslated text.
   If pipboy.msg was excluded intentionally, it might be worth documenting the reason; otherwise, adding it to the chain would keep the Pip-Boy UI translatable in future updates.
   Thanks again for the review!

[burner1024]

Now you added an extra PO file.

[NarakuITA]

Now you added an extra PO file.

Found it: one upload landed at the repository root instead of data/text/po/. The stray root-level italian.po has been deleted, and data/text/po/italian.po now carries the final version with the restored non-cp1252 characters. The diff should now show only intentional translation changes. Sorry for the noise, and thanks for the sharp eye!

[burner1024]

And now the whole file is rewritten:

I bet it's CRLF instead of LF.

[NarakuITA]

And now the whole file is rewritten:

I bet it's CRLF instead of LF.

You were absolutely right — my Windows tooling saved the file with CRLF endings throughout. Converted back to LF and pushed a fix commit; the diff should now show only the actual translation changes. Thanks for the catch!