Template talk:hu-conj-unified/doWork

From Wiktionary, the free dictionary
Jump to navigation Jump to search

Acceleration[edit]

@Tacsipacsi: Thank you for adding acceleration to the conjugation template. An example for the formatting of the first-person singular forms for second-person objects would be látlak. I'm not sure what you meant by "acceleration is not applicable to the noun form, as it’s not a non-lemma form". Noun forms are non-lemma forms and the declension templates already have acceleration. Would you mind adjusting the order of parameters? In the current verb forms that were created before acceleration, we always followed the mood - time - def/indef sequence. Your implementation is time - mood - def/indef. Another subject: Would you be willing to help the Hungarian editors in future programming projects here in the English Wiktionary? Thanks. Panda10 (talk) 17:41, 13 January 2021 (UTC)[reply]

@Panda10: I don’t like the example of látlak—it states both first-person and second-person (I think the individual parameters should be correct descriptions on their own). This idea didn’t come into my mind before, but it would be okay for me if it said second-person-object (with two hyphens), although spelling out this in the parameter input doesn’t look right for me. Isn’t there a shortcut like 2 is for “second-person”?
Noun forms are actual nouns, and nouns are lemma forms. To stay at the example of lát, the noun form is látás, which is categorized as a lemma, and has its own non-lemma forms that aren’t (and should not be) present at lát.
Parameters swapped, thanks for noting this (I just used the order in which they came into my mind, so no particular opinion on the order).
In general I’m happy to help in technical stuff, although I can’t promise that I will be available/willing to help in each individual case. Is there any meeting point for users dealing with Hungarian entries? —Tacsipacsi (talk) 18:42, 13 January 2021 (UTC)[reply]
@Tacsipacsi: Thanks for swapping the parameters. I could ask for a new parameter for second-person-object but there is no guarantee that other editors will agree. The template we use is for all languages.
Ah, I see, we usually call látás a verbal noun and noun forms would be its declined forms such as látásnak, etc. It's just a question of terminology.
There is no specific meeting point for Hungarian entries. All editors use a community portal called Grease pit to discuss technical questions related to any language. I certainly would not want to burden you with extensive programming projects. We are all volunteers. But I appreciate your reply. Panda10 (talk) 20:21, 13 January 2021 (UTC)[reply]
@Panda10: Uselessness in other languages should not be an argument against describing Hungarian appropriately. Either a qualifier useless in most languages should be accepted generally if it’s useful in at least one language, or it should be possible to define allowed qualifiers per language. Actually, providing this qualifier is possible if one writes out the full word in the parameter: {{inflection of|hu|lát||second-person-object}} → second-person-object of lát (with the irrelevant qualifiers removed)—but I don’t know whether it’s accepted by the community. —Tacsipacsi (talk) 22:40, 13 January 2021 (UTC)[reply]

@Tacsipacsi, Panda10 I, for one, absolutely agree (and thank you for your help!), though I think the spelling "second-person-object" indicates an adjective so I miss a noun from the end. What if we added "form" afterwards (before "of")? (For látok and látom, the terms "definite" and "indefinite" are supplied, which can be nouns by themselves.) However, "definite" shouldn't be used here because this form is neither definite, nor indefinite (different grammars classify it differently). If I look up e.g. sings, it says "form", so I think it'd be fitting here too. Adam78 (talk) 15:03, 14 January 2021 (UTC)[reply]

@Tacsipacsi, Adam78: I updated látlak: added second-person-object form and removed def. If this looks good to both of you, I will request a new parameter for second-person-object form. Panda10 (talk) 16:37, 14 January 2021 (UTC)[reply]

@Panda10, Tacsipacsi Great. I suppose it's deliberate that we don't link "second person", to avoid confusion with the first term "first-person", right? Adam78 (talk) 16:49, 14 January 2021 (UTC)[reply]

@Tacsipacsi, Adam78: If we link at all, we should link the entire text second-person-object form and explain it in the Glossary. Panda10 (talk) 16:51, 14 January 2021 (UTC)[reply]

@Panda10, Tacsipacsi It might look like a sum-of-parts compound, but I tried to summarize it in the Glossary to make it clear that it can deserve a definition on its own. Indeed, it could be linked for clarity. The phrasing can be improved, of course. Now I think we're all set. :) Adam78 (talk) 17:44, 14 January 2021 (UTC)[reply]

@Adam78, Tacsipacsi I've requested the parameter in Grease pit. Panda10 (talk) 18:50, 14 January 2021 (UTC)[reply]
@Tacsipacsi The name of the new parameter is 2o (o as in object). I updated látlak. Would you help us to add the acceleration? Thanks. — This unsigned comment was added by Panda10 (talk • contribs).
Done. —Tacsipacsi (talk) 21:16, 15 January 2021 (UTC)[reply]
@Tacsipacsi Thank you! It works great. Panda10 (talk) 21:41, 15 January 2021 (UTC)[reply]

Acceleration testing[edit]

@Adam78 I started testing the new conjugation acceleration. Some look different than our previous layouts. This may not be a problem, I just wanted you to know. I will update the list as I go. Panda10 (talk) 19:22, 15 January 2021 (UTC)[reply]

  1. neveltek, nevelnék - this is a new layout implemented in some of the other languages
  2. neveljük - the two forms are separated with a slash, not in two lines as above
  3. neveltem, nevelnénk, nevelnétek - the indefinite/definite forms are separated with a slash
  4. nevelt - no separate Participle section
  5. nevelnem - compare the definition (first-person singular infinitive of nevel) with tudnom (first-person singular of tudni); which one should be kept?
  6. nevelendő - has to be reworked similar to adandó, replace Verb with Participle, replace {{inflection of|hu|nevel||future participle}} with {{participle of|hu|nevel||future}}
  7. nevelve - has to be reworked similar to tudva, replace Verb with Participle, replace {{inflection of|hu|nevel||adverbial participle}} with {{participle of|hu|nevel||adv}}
  8. nevelhet - compare with tudhat: in this case, I like the new format (conjugation has to be added manually, I guess)
  9. nevelte - since this form was already created at the time of testing, I tested ajándékozta instead: the participle section will have to be added by hand.
  10. nevelő - since this form was already created at the time of testing, I tested ajándékozó instead: replace Verb with Participle, replace {{inflection of|hu|nevel||present participle}} with {{participle of|hu|nevel||pres}}
  11. alszik - acceleration is missing

@Panda10 Thanks for the note. I have nothing against the layout at neveltek but I don't like the lack of the participle section (I suppose they'll need to be added manually), nor the slash. "Subjunctive/imperative" could be separated by a slash but not indicative and subjunctive. Can we do something about these two points? Adam78 (talk) 19:30, 15 January 2021 (UTC)[reply]

@Adam78 Unfortunately, I don't know. @Tacsipacsi Would you be able to modify the code as Adam described above? Panda10 (talk) 19:35, 15 January 2021 (UTC)[reply]
@Adam78, Panda10: Not really, this is how acceleration works. Maybe we could hack it around somehow (e.g. I can imagine some smartly placed <nowiki/>s), but that would result in incomprehensible template code and garbage in entries’ wiki code. If you compare the accelerated URLs of nevettek and nevetnétek, they’re almost the same:
Both have an accel1 and an accel2 parameter with similar structures, which simply describe the forms. All collapsing magic happens after clicking the link, at which point the template has no control over the process anymore, everything is coded in the gadget (which in fact can be edited only by interface admins, so neither of us could directly edit it even if we wanted). —Tacsipacsi (talk) 21:16, 15 January 2021 (UTC)[reply]
@Tacsipacsi Thank you. Panda10 (talk) 21:47, 15 January 2021 (UTC)[reply]
@Adam78 We'll have to go over each item in the list and address the issues. E.g.: Should we revert back to Verb instead of Participle? Or manually adjust the accelerated form by adding Participle? Panda10 (talk) 21:47, 15 January 2021 (UTC)[reply]

In this case, I'd rather we skipped it; it doesn't really seem to suit us. Even if we were to accept everything as it is now, the etymology and the pronunciation (and the participle, if there is one) need to be added manually. – In fact, I think the best way would be to have the complete entries mass-created with the help of a bot, in a format that we all like and find preferable. All things considered, it would take less work than adding etymology and pronunciation (and whatever else) and then saving each entry manually every time. Now we have 4.000 verbs (Magyar értelmező kéziszótár has about 16.000), each with a hundred forms or so (or nearly two hundred, including the potential forms, plus the prefix-separated forms if we ever get to implement it) – if you do the math, even a fragment is just inconceivable to handle, even by acceleration. It's drudgery in this quantity. How did German speakers manage it in the German-language Wiktionary? As I looked into some random articles, their edit summary says "autoedit/checkpage X.XX" (version number), source. Even for English verb forms (altogether one hundred thousand of them at the moment), the past tense (-ed), the 3rd-person present (-s), the gerund (-ing), and sometimes even archaic -eth, est forms must have been created by means of some automatism, weren't they? Adam78 (talk) 21:54, 15 January 2021 (UTC)[reply]

@Adam78 I think we should keep the acceleration and still should go through the items to see if we can benefit from the differences. And then apply the unified approach. At the same time, I'm all for creating the forms with a bot after we cleared our minds about what we want. The bot operator will need good instructions from us and the bot will have to be run from time to time as we add new verbs. Panda10 (talk) 22:06, 15 January 2021 (UTC)[reply]

OK, the new format in nevelnem and nevelhet is good. We can gain some useful insights from their comparison, it's true. The rest is as I described above; I still don't like the slash but we could perhaps ask a non-Hungarian reader, which way helps them interpret the information more easily. And how are we going to add usage examples to them? Each example will need to be clarified which sense they belong to out of the two options around the slash. Also, I'd like to keep the participle separately: we had a fairly long and detailed discussion over it. Adam78 (talk) 23:37, 15 January 2021 (UTC)[reply]

@Adam78 I don't like the slashes, either. I'd prefer to keep the different forms in two lines. The "Adding new rules" section of Module:accel#Adding_new_rules says: "The general parts are defined in this module, while the language-specific rules are handled by submodules." In the "Entry" section of Module:accel#Entry, there is a parameter called pos_header which could be applied to use Participle instead of Verb. If we examine the existing subpages Special:PrefixIndex/Module:accel/ for other languages, maybe we can figure out how to construct a Hungarian subpage for some specific functions. This might be a lengthy process and I'm not saying that all of the above problems can be solved. Also, I've just noticed that not every conjugation template has the acceleration. E.g. alszik doesn't have it. About running a bot, I assumed you wanted someone else to run it, but then I remembered that you said you did have a bot in Wikipedia. So you know how to do it. There are other editors who ran bots to create form entries, they could allow to copy the bot and modify it for Hungarian. Just an idea. I don't think we can accomplish adding all the elements by bot. It is too complicated. There are exceptions for pronunciations, we don't have an automated hyphenation template specifically for Hungarian, etc. Panda10 (talk) 21:06, 20 January 2021 (UTC)[reply]
alszik: the answer is simple—it doesn’t use this template. It uses {{hu-conj-szom-ud-usz}}, which isn’t accelerated. —Tacsipacsi (talk) 21:50, 20 January 2021 (UTC)[reply]
Thank you! Panda10 (talk) 23:50, 20 January 2021 (UTC)[reply]
@Adam78 I thought about this some more. It's good to know that there are options but it would take a lot of time and effort for us to create the customization. I'm glad we have the acceleration function for verbs even it it is not perfect. We can always manually apply the current standard just like we do for nouns. Panda10 (talk) 23:50, 20 January 2021 (UTC)[reply]
Retrieved from "https://en.wiktionary.org/w/index.php?title=Template_talk:hu-conj-unified/doWork&oldid=61617634"