EDIT : The most important tests (for now) are already done. Thanks to those who posted test results. I'll be using the results to help me make some better language support in a compatible test patch, in a week or two (i hope) depending on how long it takes to do the updates and a few other fixes...

--- original post ---

This thread is for people who get the wrong characters appearing in LFS when they try to type some of their special national characters. I will ask you to run a quick test to help me support certain languages correctly, when your computer is set to the correct input mode...

For example : Cyrillic, Czech, Greek, Japanese Katakana
- these languages can be supported with single-byte code pages

Not including : Chinese, Japanese Kanji, Arabic, Hebrew
- why not : http://www.lfsforum.net/showthread.php?p=22360#post22360

More info and results of a Katakana test : http://www.lfsforum.net/showthread.php?t=1848

TEST 1 : If you use Cyrillic, Czech, Greek, Polish, Lithuanian, etc...

Please run the small exe in the attached ZIP file.
I would like to know the information that appears on the screen.
A screenshot is an easy way to show the info (attached : example for UK English).

TEST 2 : LFS keypress tests to help me check which codepage is being used...

Cyrillic - What character appears in LFS when you type this character :

CAPITAL LETTER YA (backwards R)

Greek - What characters appear in LFS when you type these characters :

CAPITAL LETTER ALPHA WITH TONOS
SMALL LETTER ALPHA

Czech - What character appears in LFS when you type these characters :

CAPITAL LETTER A WITH OGONEK
SMALL LETTER C WITH CARON
SMALL LETTER S WITH CARON

TEST 1: (attached)

TEST 2: i dont exactly understand what means OGONEK or CARON ...because that are not czech words.. but i guess OGONEK=carka and CARON=hacek...because we dont have any other special character for these letters...so in lfs the A with ogonek is correct and S with caron is also ok...but C with caron isnt ... (attached pic)

EDIT: i looked at wikipedia and CARON is indeed what i thought, but OGONEK seems to be something different http://en.wikipedia.org/wiki/Ogonek ....but we dont have any letter like that in cz

Thanks Don.

I think it's clear which code page you are using (Windows 1250) and you can see the OGONEK under some letters :
http://www.microsoft.com/globaldev/reference/sbcs/1250.mspx

You can compare with this (Windows 1252 - like LFS) and see how you get the wrong characters :
http://www.microsoft.com/globaldev/reference/sbcs/1252.mspx

So... that's covers Czech. I don't think we need to hear from any more Czech people unless they have different results...

Your Wikipedia link says that these characters are used in Lithuanian and Polish.

So... i guess the Windows 1250 codepage covers Lithuanian and Polish as well? I'd be pleased to see the results of running the test exe on a Lithuanian and Polish computer!

The attached screen has only the A ogonek glyphs, but as far as I noticed, most characters are pretty messed up.

Quote from noizz :

The attached screen has only the A ogonek glyphs, but as far as I noticed, most characters are pretty messed up.

OK

The "ANSI code page" is the important thing here, same as Czech... good news! And that is consistent with your A OGONEK being replaced with the Yen sign and the superscript 1. So page 1250 (known as Central Europe) definitely covers Polish and Czech.

Hello Scawen, nice to talk to you here is the result of the exe and the second screen shows first letter capital greek "A" with tone ,small a and capital A.
bye

Thanks psy00. Your in-game keys screen shot, though quite blurred, seem to show what i would expect, because your ANSI code page is shown as 1253 - the Greek one. So, this is working out consistently as well.

Good news, we now have consistent results for :

Japanese (page 932)
Czech (page 1250)
Polish (page 1250)
Greek (page 1253)

We still need the results for :

Cyrillic (probably page 1251)
Lithuanian (probably page 1250)

I don't understand exactly why, but the first test gives me different results, but it seems to give the same on the second test, although I'm not so sure because of the blurry small a of psy00.

Paranoid you have the old codepage to show the greek chars on old dos programs whice is not used any more...
bye

Here is Cyrillic test. I'd like to notice that in Cyrillic there is also a letter Ё (yo ) which is usually put aside of the other 32 letters, please, don't forget it.

[edit] added ё and Ё in the last attachment.

Quote from Paranoid Android :

I don't understand exactly why, but the first test gives me different results, but it seems to give the same on the second test, although I'm not so sure because of the blurry small a of psy00.

Hmm... that's a problem.

I was hoping to use the ANSI Codepage result, to detect how the incoming bytes should be interpreted. but in your case, that would not help at all because you have Windows 1252, standard Latin-1.

That means if i carry on as planned, your characters would still be wrong in the new version.

Doesn't make much sense to me... do you have the same problem (wrong characters) in Notepad, or does it work correctly in Notepad?

Quote from detail :

Here is Cyrillic test. I'd like to notice that in Cyrillic there is also a letter Ё (yo ) which is usually put aside of the other 32 letters, please, don't forget it.

No problem, yours is consistent with Windows Codepage 1251 : http://www.microsoft.com/globaldev/reference/sbcs/1251.mspx

Everything on that page should be included...

Quote from Scawen :

Hmm... that's a problem.

I was hoping to use the ANSI Codepage result, to detect how the incoming bytes should be interpreted. but in your case, that would not help at all because you have Windows 1252, standard Latin-1.

That means if i carry on as planned, your characters would still be wrong in the new version.

Doesn't make much sense to me... do you have the same problem (wrong characters) in Notepad, or does it work correctly in Notepad?

No, notepad and word have all the correct characters. Actually this problem has showed up in other translated programs too. I don't know what causes it though.

I'm pleased you say that it has gone wrong with other programs. Because it's really quite strange.

On your computer :

Language : American English
Locale : American English
ANSI Codepage : Windows 1252 (Latin 1)
OEM Codepage : US
Keyboard Layout : Greek

The Latin 1 ANSI codepage is correct for an American computer, i guess, but not what you would expect for a Greek one.

Your computer seems to be a blend of American English and Greek...

Quote from Paranoid Android :

No, notepad and word have all the correct characters. Actually this problem has showed up in other translated programs too. I don't know what causes it though.

Check the "Language Settings" under "Regional Options" in your control panel and make sure that Greek is set to Default - that way non-Unicode supporting programs will default to using the Greek codepage.