[translation] Update translation from Crowdin - #864
Conversation
[ci skip]
[ci skip]
|
The Polish update here needs a few fixes on Crowdin before this goes in.
"Scope Request" in The
Also |
|
@Kuchta2100, please improve the quality of Polish translation from the Crowdin side. |
|
All done, now is 100% fixed and correct. |
[ci skip]
[ci skip]
Most of the sync is an improvement and stays. Polish settles on "framework" where the file had been split ten lines using "szkielet" against twelve using "framework", including the adjacent modules_incompatible and modules_incompatible_legacy pair; ten Polish plural sets that had been stubbed one/other gained the few and many forms the language needs, so "Włączono moduły: 3" becomes "Włączono 3 moduły"; the daemon's notifications lose two calques of "on user X"; and Japanese store_downloading and logs_part drop shapes that read ambiguously for the "1.2 MB of 3.4 MB" the source comment asks for. None of that is touched. What is undone falls in five groups. Meaning. settings_doh_bypassed had stopped saying that the proxy is the thing resolving names, which is the only reason the string exists, and launcher_install_conflict replaced "before this one will fit" with "so that this key matches", which the source never says. Polish and Japanese logs_dropped both read "longer than the viewer indexes" as a plural noun rather than a verb, leaving "longer than the viewer's indexes" in two languages independently; the English is ambiguous enough to be worth rewording at the source. Japanese logs_loading gave up "Indexing…" for a generic loading message, and it is the one string that says why the wait exists. Grammar. home_people_count_plain rendered "2 osób", where the nominative few form is "osoby"; logs_dropped disagreed in case across "Najstarsze %1$d wierszy", and now says "linii" like every other line reference in that file; and logs_rotate_unreachable left "z Daemon" uninflected after a preposition governing the instrumental. That last is a symptom - the sync had swapped "usługa" for an indeclinable English "Daemon" in seven places across strings_logs.xml and strings.xml, so all seven go back. Mood. home_history_loading, home_window_changed and language_system_summary had become imperatives, telling the reader to reach further back, to display saved activity and to choose a language, where the app is narrating a spinner, a state and a preference. store_tab_readme went furthest and named a file with the imperative "Przeczytaj"; it keeps the sync's intent to name the file rather than paraphrase it, as "Readme". Register. The four framework status words share a headline row with the wordmark, and the sync left two lowercase verb phrases, one clause and one gerund there, so "Vector jest aktywny" sat beside "Vector Wersje się nie zgadzają". All four read as labels again, none of them a finite verb phrase. store_all_current is lowercase again because it renders after a "·" in the store header, as it does in every other locale; retry matches settings_doh_retry rather than disagreeing with it; modules_batch_all_already_on stops collapsing into modules_batch_already_on and matches the "off" variant beside it; and nav_home is "Start" again, "Strona główna" being thirteen characters in a four-item navigation bar. Consistency. store_info_scope_installed arrived new in eight languages and five placed it correctly. Turkish said "Kancalar" beside "Hook'lar", Arabic "العناصر المُعترضة" beside "الارتباطات", and both Chinese variants 目标/目標 beside 对象/對象, each on the same tab as the label it disagrees with. Placeholders were sound throughout the sync and stay sound: none is added, dropped or reordered here.
0f73171 to
f416804
Compare
About the last commit on this PRBefore merging this sync I ran an AI review over it, and then reverted part of what it flagged. I want to be upfront about how that was done, because most of this is volunteer work and it deserves an explanation rather than a silent revert. This review is purely AI. I read English, French and Chinese. That is the whole list. For the Chinese strings below I can confirm the change myself; for Polish, Japanese, Turkish, Arabic and everything else, I am taking a model's word for it and I have no way to check whether it is right. I merged its conclusions because the reasoning it gave was specific and checkable against the source strings — not because I can read the result. So please treat every point below as a question, not a correction. If you speak one of these languages and the review got your string wrong, just say so and I will revert it. You do not need to argue the case or convince me; I would far rather trust a person who speaks the language than a model that does not. Most of the sync is a real improvement and is merged as it stands. What went back was a smaller set where the review argued the string had changed meaning, broken grammar, or turned a narration into an instruction. One thing held up everywhere: not a single placeholder was added, dropped or reordered in the whole sync, across fifteen files. That is the part that actually breaks the app, and nobody broke it. PolishThis file got the most work, and a lot of it was overdue. Kept: the file finally settles on "framework" — it had been split ten lines using "szkielet" against twelve using "framework", including the neighbouring Reverted:
That is a long list against one language, and I am conscious it reads harshly. It is long mostly because Polish is the only file that got a full rewrite in this sync — the others changed a line or two. The rewrite also carried the single biggest improvement in the PR. JapaneseKept: Reverted:
One left alone: TurkishOnly the one new string. ArabicOnly the one new string, same shape of issue. It came in as "العناصر المُعترضة" while Chinese (Simplified and Traditional)Only the one new string in each: 挂钩目标 / 掛鉤目標 sat directly beneath 挂钩对象 / 掛鉤對象. Unified on 对象 / 對象. This is the one I can vouch for myself. Russian, Ukrainian, Korean, Persian, VietnameseNothing to change. The new string landed correctly and consistently with its neighbours in all five. Thank you. |
|
@Kuchta2100 Please check the above review |
|
Some fix in Ukrainian translate done. |
No description provided.