“Gondi, Adilabad” [wsg] appears to be missing from the list of languages in 6.1.112 (and 6.1.110, for that matter). It used to be there…where could it have gone?
Hi Fraser, welcome to the Language Software Community site!
The short answer: use ‘gon’.
The long answer involves macrolanguage codes and how the computer industry deals with them. “gon” is the macrolanguage code for Gondi (gon | ISO 639-3) and it encompasses “esg”, “ggo”, “gno”, “wsg”. Current practice in computing is that one of these encompassed languages is chosen as the representative of the macrolanguage and is assigned the macrolanguage code. So, although “wsg” is still valid, “gon” is the preferred language tag to use for Adilabad Gondi.
But if this doesn’t answer your question, please ask a follow-up one.
Hi Fraser,
We should probably chat about this.
I do know that you all have been using wsg, and there are books uploaded to bloomlibrary with that tag.
But, going back to Bloom 5.6, I don’t see that tag coming up in the language picker.
Of note is that the new picker, which will release in version 6.2, does have it.
David, it sounds like you disagree with the approach we have taken, which is to direct people to use the specific language codes rather than macrolanguage tags. We should probably discuss that (likely in a different format).
Thanks, @drowe and @andrew_polk. Happy to have a conversation about this when it’s convenient.
My response was addressed at why “wsg” wasn’t in the list of languages. It seems that you have the means to use “wsg” in Bloom (at least in version 6.2) and that that is the recommendation from Bloom. I agree that the more specific tag is less confusing. You should be aware, however, that in other contexts “gon” may be substituted for “wsg”.
Current practice in computing
The Bloom project manages one of the larger collections of minority language materials in the world. From that position, we have developed an opinion on what is helpful when it comes to tagging materials. We strongly discourage the use of ambiguous codes where possible. There are two instances of this where we encourage doing better than standard practice:
- avoid macrolanguage codes, just use the specific ISO 639-3 code
- for languages without an established code, avoid raw codes like
qaa-qtx. Instead, always start withqaaand use the rest of the tag to make it descriptive, e.g.qaa-AQ-x-foo" (which means “Foo in Antarctica”).
Our new Language Picker is based on these two practices.
–John
Thanks for the work done on the new language picker. I think it does an excellent job of displaying things progressively, and includes enough options even with a vague search, without overwhelming the user with too many things to choose from.
I also like the fact that this interface actively discourages the use of macro language codes:
So I’m confused by a statement earlier in this discussion: although “wsg” is still valid, “gon” is the preferred language tag to use for Adilabad Gondi. I can’t see why it would be better to use ‘gon’ rather than a specific code like ‘wsg’. And who gets to decide which single sub-language gets to represent all the other languages within ‘gon’? That seems a bit of an arbitrary choice - and certainly one that outsiders wouldn’t/shouldn’t make.
If people are creating content to be uploaded to the BLOOM library, we need to encourage as much detail as possible (ie. drill down to the sub-language codes, and scripts, and even region(s)). But if the language picker is being used to search for content then we need to make sure that searching for a single macro-code ALSO provides results from the various sub-language codes below it. I hope I’m not just stating the obvious.