We are working to correct the implementation of Greek on digital platforms. Showing Unicode and OEMs that this matters for many readers is critical, so please read and sign this petition: www.openpetition.eu/gr/petition/... - ThreadSky

gerryleonidas.bsky.social • 7 days ago

We are working to correct the implementation of Greek on digital platforms. Showing Unicode and OEMs that this matters for many readers is critical, so please read and sign this petition:

https://www.openpetition.eu/gr/petition/online/digital-greek-typography-is-broken-improve-standards-and-demand-fixes-in-all-software

Comments

enablelanguages.bsky.social•6 days ago

With respect to case mapping issues, are there examples you can point to? source -> incorrect casing -> correct casing?

Quite a few of the issues don't seem to be specifically Unicode issues, most seem to be related developers being bad at internationalisation.

enablelanguages.bsky.social•6 days ago

I tend to work a lot with Python, but its string methods are language and locale insensitive, so default methods will not work correctly with Greek. I'd need import pyicu and use locale aware methods. But that said, there seems to be edge cases like Νεράιδα where it doesn't roundtrip.

enablelanguages.bsky.social•6 days ago

I know of https://www.w3.org/TR/grek-gap/#issue15_transforms, any other discussions?

enablelanguages.bsky.social•6 days ago

I think there are a number of approaches. One is developing information resources aimed at developers, discussing internationalization with a Greek focus. The second is something akin to W3C I18N WG's gap analyses. It identify Greek i18n issues and map them across key software. File bugs.

gerryleonidas.bsky.social•6 days ago

Indeed we have been circulating resources for literally decades. The W3C gap analysis is full of “needs research” for information we have shared with OEMs and developers similarly. Those “needs research” entries should be followed by “hm, let’s Google for Greek experts”. I/We are not hard to find.

enablelanguages.bsky.social•5 days ago

One of my long term projects is to document Python internationalisation. There really is limited information available, and what there is, is scattered. For instance, if I search for information on how to do Greek case conversions in Python I'm not likely to find anything. The defaults will not work

enablelanguages.bsky.social•5 days ago

But the average developer will not know this. As you indicate the issues have been discussed for decades. Some resources remain, some are long gone. There is currently no single source to refer to, and more importantly there are no resources to teach developers how to handle Greek properly.

enablelanguages.bsky.social•5 days ago

And I have been aware of your work, and a few others over many years. But focusing the discussion. Unicode has over time improved Greek support, I guess the question is, are there aspects of Unicode's handling that are still wrong?

gerryleonidas.bsky.social•5 days ago

Short answer: yes, we outlined only some of the problems. Characters not encoded properly, normalisation wrong (and breaking the char-glyph model). That’s all on Unicode. Then shaping engines apply rules wrong (bundling Greek with Latin and Cyrillic pushes a lot of assumptions on both Gr and Cyr)

enablelanguages.bsky.social•5 days ago

I understand. Its interesting going through old issue trackers and following the discussions. Although many of the resources pointed to are now dead links. Part of the transitory nature of the internet.

Comments

Posting Rules

Reply