Vanilla 1 is no longer supported or maintained. If you need a copy, you can get it here.
HackerOne users: Testing against this community violates our program's Terms of Service and will result in your bounty being denied.

Translation: Female/Male forms of verbs and adjectives

edited April 2007 in Vanilla 1.0 Help
Unlike English, Polish, Spanish and several other languages use different forms of verbs and adjectives when talking to or describing a male of female subject. In English you could see a difference between “her posts� and “his posts�. In other language it works the same way, but also with verbs and adjectives. For example, “Signed in� would translate as “Zalogowana� for female and “Zalogowany� for male subject. Without ability to do that, a translator has to use a formal, gender-insensitive form which just doesn't fit with the rest of Vanilla and gives an unpleasant impression. This is more or less general issue for many languages. To solve this, Vanilla would have to know every user's gender and have male and female forms in the definitions.php file. How about adding support for that?


  • lechlech Chicagoland
    The formal / gender neutral verbage is probably best in most cases, especially this one. Because you can't really know the users gender until they sign up, and if they do, having them reveal their gender is completely optional and not a requirement unless you make it so. If you were to make it mandatory, it would have to be done as an extension, and in the end, you would be nearly doubling the translation workload.
  • edited June 2006
    Because you can't really know the users gender until they sign up -- anonymous users don't write posts having them reveal their gender is completely optional and not a requirement unless you make it so -- you can leave it optional and then use male form by default. Users of my forum have gender-aware nicknames anyways, so gender privacy isn't an issue here, my users would tell their genders. I disagree with translation doubling workload. And I'm the one who does the translation, so I'd be dooming myself in the first place. Neutral translation is really difficult sometimes. You're Polish, so tell me how to translate “Last comment by <nickname>� neutrally?' - “Ostatni komentarz napisano <nickname>� is incorrect - “Ostatnio pisano <nickname>� is incorrect - “Ostatnio komentarz autorstwa <nickname>� is incorrect - “Ostatnie słowo <nickname>� is incorrect - “Ostatni komentarz: <nickname>� is awkward and another-software-that-can't-handle-translation-properly Any other ideas?
  • Why do foreign languages always have to bring gender into things. No wonder english people are so lazy at learning foreign languages.

    It does seem like this might be a good idea for an extension though, I'm pretty sure mark has no plans to add 'features' before v1 and this wasnt highlighted as an issue when language abilities were first put into vanilla...
  • edited June 2006
    English people know English already, don't they? And since they do, there's no need to bother... Minisweeper, I don't want to scare you, but there's also a translation issue concerning numbers. That's an issue for another discussion. Instead of “1 apple�, “2 apples�, we have more forms and quite complex rules on which form to use. I have a coded routine already, but I don't know how to extend it to a general mechanism. Back to the gender topic: As Vanilla claims to be the sweetest forum on the Web, gender issue is necessary. All the computer software seems so _rude_ with those formal translations. Having gender-aware translations would be very, very sweet, believe me.
  • Yeah I'm not saying it wouldnt be a good thing, and i suspect Mark would certainly consider implementing it into a future version, but I also think he's really trying to restrict himself from changing anything else besides bugs before he releases v1. It's for him to decide how important he feels it is to resolve this issue immediately or just in future though. I daresay it wouldnt be a small task.
  • edited June 2006
    Having this even in a “very future� release, Vanilla would still be the first. I can imagine that it's not a priority in scope of the global Vanilla development. I also believe that Vanilla's design (unike other forum engines') is strong enough to sustain such “foreign� demand. If not today, then tomorrow.
  • edited June 2006
    One thought. Maybe a mini-markup would be a good way to express the gender differences. Let's take three forms: male, female and neutral. There's an expression to translate, and only two words differ. There's no need to write the same thing three times just to make one word have different forms. Maybe something like this {male|female|neutral}, for example: “Użytkowni{k|czka|ctwo} napisał{|a|o}.� It's just an example, I don't say that curly brackets are the best. It's about having a small markup for translations.
  • lechlech Chicagoland
    Well, I grew up learning english as my primary, and I also speak (and can read some but not enough to write) in polish. My mom is always on the verge of smacking me upside the head because of the way I talk to her (her fault) but with that, I've found that having femenine and masculine forms are annoying to no end. It will likely take a long time until everyone just speaks formal or generally without implying gender to make it easier. However, it does help, To clear things up for you though, mini, in most languages by having femenine and masculine verbage, you can literally say less and mean more just by how things are said: ie implying who it's for, based on the verb/adjective gender. In english, the word 'cunt' or 'bitch' might be a good example (no offense to any ladies reading this), because most often, one will think it's in refference to a female if not being used to emasculate a guy. So in that sense, it makes things easier as the language is sometimes descriptively condensed or limited in some ways but richer in others. Given with that, automatthias, I can see where you're coming from, but if you do this, you'll want to format it as an extension. It will quite likely be neat if you could somehow also account for other language that have similar language styles. And like mini said, it probably would not be a small task as it wasn't thought of in detail until now. My guess is that on top of specifying gender, you would be looking for dictionary entries and simply rewriting them according to that specified gender, correct? Also, on a completely different side-topic: Taking all of this into account, would this imply that all websites in polish (or similarly sexed languages) are written incorrectly towards the specific audience if it was initially meant for everyone? I'll ask around in the meantime to my other friends who actually know polish better than I do on any translations.
  • jedrekjedrek New
    edited June 2006
    doot doot.

    lech got me in here, i haven't been around in forever.

    there's quite a vocabulary for gender neutral forumspeak. instead of saying 'Zalogowan(y/a)', forums will use 'Użytkownik zalogowany', 'Ostatni komentarz:' functions instead of 'Last commeny by' and so on. look at phpBB's translation, it's pretty much the de facto standard.

    sure, it'd be nice if vanilla was able to do gender specific translations, but to a large degree, this problem has been solved. with polish, you want to stay away from exceptions/new conditions as much as possible - it's a fucking minefield.
  • @lech: The entries should be defined by hand. It's impossible to write a function that changes the form of the verb or adjective. It's better just to hard code all of them. And having to hard-code, it's best to be able to only code the difference, not the whole string again.

    Having a lightweight markup for gender specification, the English translation could stay as it is, it would be only the foreign strings that'd be parsed and transformed.

    I'd also like to mention that it's not only about Polish. The issue concerns also Spanish (the third language in the world), Italian, Russian, French (I think...), certainly all the Slavic languages (Czech, Slovak, Bulgarian, etc.), and possibly many others.

    All the webistes in Polish and all the software translated into Polish either struggles to use gender-neutral and formal language, or uses just incorrect forms (“Jane went to a shop and bought himself a bra.�)

    @jedrek: I wouldn't say the problem was solved, but walked around. I've been in contact with it for years and years and never stopped having the feeling that it's awkward. Have you ever seen the “Dr Who and the Daleks� movie? It's just a Dalek voice in my ears.

    I believe that carefully designed translation engine would do the job.
  • disclaimer: my knowledge of word gender comes entirely from my crapass highschool spanish education.

    would it be possible to work around this for now by adding a preference in the profile to specify your gender, and then in the language file add a php if statement to determine which strings to add to the dictionary?

    i know for dead certain that this would double the translation workload, unless there's a third or fourth gender to worry about hiding someplace... but i guess it could be a workaround hack in the interim.
  • Sorry, that I digg this old topic, but... I was traslating Vanilla (and some other things) and I think that there is two worse things in translating Vanilla than lack of support for gender forms:
    - date! you can change date format, but what with the names of months? it should be placed in translation file
    - in all Slavic languages there are different forms of nouns, which depends of nunmber of it. In English there are only two forms: single and multiple (comment and comments), but in (for example) Polish we say: 1 komentarz, 2,3,4 komentarze, 5 to infinity komentarzy. I know some problems with it are also in Japaneese.
  • Plural forms are already accounted for in gettext. That's why I tried once to push toward gettext shift for Vanilla.
    From gettext manual:
    In Polish we use e.g. plik (file) this way:
    1 plik
    2,3,4 pliki
    5-21 pliko'w
    22-24 pliki
    25-31 pliko'w
    and so on (o' means 8859-2 oacute which should be rather okreska, similar to aogonek).
    Plural-Forms: nplurals=3; \ plural=n==1 ? 0 : \ n%10>=2 && n%10<=4 && (n%100<10 || n%100>=20) ? 1 : 2;
    I already mentioned in another thread that it's NOT very much work to shift toward gettext. I've done it as a test once. Currently my forum still uses standard strings, but I'm considering to upgrade it in the future.
    Maybe I could release an add-on. Don't know if and when I'll find the time.
  • For me the simplest solution (and I always strive for the simplest solution) would be a very short extension that has an optional gender to female checkbox in profile and if you have marked this, your female. Then you have two language files, one polish_default, and one polish_female. Where the language file is included, you do a simple if statement. I am aware that this doesnt cover all bases (it wont cover references to other female activity for instance), but it will make the forum speak to the user in a proper way at least.
This discussion has been closed.