Repository navigation
Terminology: seven glossary rulings (from action-translation#78 + cours/chapitre) #42
Description
Activity
@Honaminto — apologies: items 1–6 above had already been answered by you, with sources, on QuantEcon/action-translation#78 on 16 July, and your rulings are in the French glossary (v1.1). This issue should not have asked them again — please ignore them.
Only two questions remain open, and there's no rush on either:
- Lecture — « cours » or « chapitre »? The edition currently has 109 « cours » and 4 « chapitre ».
- Package — you used « package » in
oop_intro(to match the series) and « bibliothèque » for NumPy inpython_by_example, while the French Python docs use « paquet ». When is each right, and does « gestionnaire de paquets » stay?
- For “lecture”, I’d keep « cours » as the default. In French educational usage, a cours refers more naturally to the teaching unit itself, while a chapitre is usually a subdivision of a larger course, book, or document. That also fits the way QuantEcon uses “lecture”, since each lecture is essentially a self-contained teaching unit.
It also keeps us consistent with the existing French edition, where « cours » is already used overwhelmingly (109 occurrences versus 4 for « chapitre »). So I’d standardize lecture → cours and check the four remaining uses of « chapitre » individually. If they refer to a QuantEcon lecture, I’d change them to « cours »; if they genuinely refer to a chapter within a larger work, I’d keep « chapitre ».
- For “package”, I’d use « paquet » when we mean a Python package in the technical sense. This matches the official French Python documentation, which uses « paquet » for Python packages:
https://docs.python.org/fr/3/tutorial/modules.html#packages
I’d keep « bibliothèque » when the English source says “library” or refers more broadly to a software library, rather than specifically to the Python package concept.
So I’d use:
- package → paquet
- Python package → paquet Python
- library → bibliothèque
- package manager → gestionnaire de paquets
That means I wouldn’t automatically replace every reference to NumPy as a « bibliothèque ». If the English source calls NumPy a “library”, « bibliothèque NumPy » is appropriate; if it specifically refers to the NumPy “package”, then I’d use « paquet NumPy ».
And yes, I’d keep « gestionnaire de paquets ».
@Honaminto, thank you for these answers. They are recorded in #82, which also proposes a way to review the remaining lectures.
Part of #40 (Phase 0). Decision item for @Honaminto.
Question. Six glossary questions were proposed in action-translation#78 and merged 2026-07-22 marked review needed; none has had a native speaker's ruling. Answering them improves all 27 lectures at once, including the 25 nobody has read yet, so this is the highest-leverage item in the project.
The terms. Please reply here with agree / prefer X for each (verbatim from #78, plus one deferred since):
Standard normal→loi normale centrée réduite(the conventional French statistical term). Neither model produced it — both said "loi normale standard". Is the conventional term what we want?f-string— keep in English (proposed), orchaînes f?Broadcasting— keep in English (proposed), or gloss asdiffusion (broadcasting)?Index→Indice— confirm? Matches the existingTail index → Indice de queue, thoughindexis common in French tech writing.Tuple unpacking—déballage de tuple(proposed) ordécompression de tuples?frame— not pinned; the models splittrame/cadreand it is probably context-dependent (stack frame vs dataframe). Pin it at all, and if so to what?lecture—glossary/fr.jsonv1.1 pinsCours;chapitrewas raised and deliberately left to you in action-translation#241.What happens with the answer. Each ruling is pinned in
glossary/fr.json(a PR in action-translation, which we will open and link here), and the terminology flows into every future translation automatically.Done when