Russian localization

uli

CityAceE <2> + <SHIFT> must type "№". But the help displays “2”, but actually prints a space.

That seems to happen when the font doesn't contain the character. I had the same issue when trying to make a PETSCII layout. Gotta look into that. (Please open a bug report so I won't forget.)

Dmian

If any of the fonts needs improving, just let me know. Alternatively, an issue can be opened in the repos:
https://github.com/damianvila/font-lexis

https://github.com/damianvila/font-bescii
(BESCII doesn't have Cyrillic. I'm planning to add it soon).

CityAceE

Dmian If any of the fonts needs improving, just let me know.

Thank you!

Right now, I see three problems with the Liexis font:

the letters “Й” and “й” have the same lettering and are visually indistinguishable. Examples: “ТРОЙКА, тройка, Йогурт, йогурт, Йод, йод”
there is no “No” character, which in the Russian layout is typed using SHIFT + 3. Example: “Пункт №1”.
The letters “Ё” and “ё” are different in appearance, but both have the same height and look about the same. Example: "МОЛОДЁЖЬ, молодёжь, Ёлка, ёлка, Ёж, ёж".

I suggest as a reference point the 6*8 font from the popular Soviet computer Mikrosha (Микроша).

Dmian BESCII doesn't have Cyrillic. I'm planning to add it soon

I'll look forward to it as it's kind of the most popular retro fonts.

https://en.wikipedia.org/wiki/Code_page_866

Dmian

CityAceE Oh! Ok. Thanks a lot for the reference. They're very handy. I'll try to improve the fonts as soon as possible.

Yes, with uppercase letter with diacritics (like Á, Ó, Ñ, etc.) it's very difficult to fit them in the 6x8 or 8x8 space. In the latin alphabet, since there's distinction between uppercase a lowercase (N and n, A and a, etc) I can do with making a smaller uppercase, and inserting the diacritic on top. But Cyrillic lowercase glyphs are basically the same as the uppercase ones (but shorter), so this solution clearly lends itself to confusion.

That's why having a local reference it's extremely important, since I can see how native readers solved the problem, and that guides me in creating a better version. Again, thanks for the references as these are extremely helpful. ☺️

Dmian

@CityAceE I'm now working on the glyphs for Bescii. Here's a little preview. I wanted to check with you to see if I'm on the right path. If you see something that looks weird or needs to be changed, let me know.

I used a couple sources as inspiration, so I'm not creating anything new, but even native speakers sometimes don't solve the problems presented with quite the same skill.

It may still take a bit, because I'm re-creating the font in a different (cheaper) version of the program I used that can't open the original files, and then I need to convert to TrueType, since this version doesn't export in that format, and check that everything works as expected. At the same type, I'm expanding the font with Greek, and reviewing a few characters I didn't like much (I will probably change the game glyphs quite a bit). I'm also adding something Ulrich asked me too, that was missing (the white bullet). 😄

So, this will most probably be version 2 of the font. I'll keep you updated on the progress.

CityAceE

Dmian I'm now working on the glyphs for Bescii. Here's a little preview. I wanted to check with you to see if I'm on the right path. If you see something that looks weird or needs to be changed, let me know.

Thank you! Everything looks familiar and readable to me. I don't see anything I'd like to fix.

Dmian

Please, try the beta of v2.0 of Bescii and tell me if you see anything wrong with it:

https://github.com/damianvila/font-bescii/releases/tag/v2.0-beta

I had to re-create the whole font, so I took the opportunity to fix a few things and add the requested features. But I used a different workflow, so I hope nothing's messed. I kindly appreciate if you can help me by checking that everything is ok.
Thank you!

uli

As a reminder (Have I actually mentioned that anywhere?): You can use LOAD FONT <filename$> to load TTF fonts and select them with FONT <name$> SIZE <x>, <y>. (The name is the file's base name without extension.)

Dmian

uli Oh! Right. Yes, yes you did. And I totally forgot… 😝.
I’m going to check it, and if everything goes ok, I’ll make the official release. (I already modified a few things I discovered).
Next weekend I’ll do Lexis. Cheers!

Dmian

Everything seems to be in order. We go with the release. 😊

Dmian

Tested with the desktop version, works well there too:

Btw: @uli While trying to generate this second image, I configured the system with the Spanish keyboard layout, but seems to be mapped wrong?
I checked TKeyboard.cpp (and the sdl one) and think some of the mappings are wrong. But I see some U and backslashes around that I have no idea what they do, and I'm afraid to break anything if I modify it. How can I help you correct it? So far I've identified 4 wrong keys.

uli

I picked up the new font and the keyboard layout, but they might only show up in tomorrow's build.

Dmian Btw: @uli While trying to generate this second image, I configured the system with the Spanish keyboard layout, but seems to be mapped wrong?
I checked TKeyboard.cpp (and the sdl one) and think some of the mappings are wrong. But I see some U and backslashes around that I have no idea what they do, and I'm afraid to break anything if I modify it. How can I help you correct it? So far I've identified 4 wrong keys.

Since the Spanish layout happens to be the last of the built-in layouts, we could simply drop it and add a keyboard layout file like the Russian one instead...

CityAceE

Here's what I got. I think it looks good. I still don't really like the way the capital letters "Ё" and "Й" look, but I realize it's almost impossible to do anything with them given the limited height. But those are letters that are pretty rarely used as capitals, so I think it's okay to ignore them.

Thanks again for your work!

@uli Please add Russian keyboard layout to next release.

Dmian

CityAceE I still don't really like the way the capital letters "Ё" and "Й" look, but I realize it's almost impossible to do anything with them given the limited height.

Exactly. There’s not much you can do with a 8x8 grid. I tried my best.

Dmian

uli Ok, perfect. I'll check the repo and look at the Russian file to see if I can make a similar Spanish file. I'll let you know. Thanks!

Dmian

@uli Ok, now I'm actually confused. I found what looked like a nice tool to check the scan codes: https://kbdlayout.info/kbdsp/scancodes
But when I compared it with the info: https://wiki.libsdl.org/SDL2/SDL_Scancode
They don't match at all. 😣
It makes things a bit confusing because all documentation is for ANSI 104 keyboards, and here we use ISO 105 keyboards...
For example, my keyboard layout looks something like this:

I have no idea what the scan code is for the key marked in red (that doesn't appear in ANSI 104). Also, I don't know how to map the deadkeys in green. Can you help me solve this problem?

CityAceE

Dmian But when I compared it with the info: https://wiki.libsdl.org/SDL2/SDL_Scancode

For Russian keyboart layout I used that table.

Dmian

Ok, this is what I have:

spanish.zip

417B

I think the code for the missing key is 100 (0x64).
I put the deadkeys with a backslash before them. They're the two first characters at 0x2f and 0x34.
If I need to do something else, let me know.

uli

Dmian I put the deadkeys with a backslash before them. They're the two first characters at 0x2f and 0x34.
If I need to do something else, let me know.

I just sat down to spend five minutes on implementing dead keys, but I quickly realized that that may not be entirely trivial after all...

We don't just need the dead keys themselves but also all combined characters that can be formed using the dead keys. I took a look at the Linux key map definitions, and they simply have values like "dead_acute" that magically do the right thing, i.e. the semantics of these keys are implemented "somewhere else", possibly hard-coded.

So I have (at least) several questions:

Are the combined characters that are formed for a specific dead key (e.g. acute, grave, circumflex) the same for every keyboard layout in the entire universe, or are there exceptions? (I have not looked into that, but I would bet on the latter...)
If there are exceptions we need to encode all the combined characters in the key map. Any suggestions on how to go about that?
How do those keys in the layout above work that are dead keys but also have AltGr bindings (e.g. [)? I assume that those are semi-dead keys that are only dead for certain modifier states, right?

Dmian

@CityAceE Ok, I'm with the Lexis (6x8) characters now.
This is what was previously:

And this is the new version:

I used the reference you sent me. I tried to make the "Ё" and "Й" different between caps and lowercase as much as I could. In the case of "Й" that's the way it was in the reference. Tell me what you think, and if you prefer a specific character from the old version, rather than the new.

CityAceE

@Dmian Thank you! I like the new version a lot. I wouldn't change a thing about it.