• Re: UTF vs specific

    From Karel Kral@2:423/39 to Eugene Subbotin on Sun Sep 20 07:51:26 2026
    Hello Eugene!

    19 Sep 26 09:35, you wrote to me:

    What must be done: 1) add CP895 into iconv and 2) get Golded+
    working with 1byte against UTF and back conversion

    No one will ever add this encoding to iconv or the Windows
    distribution: itīs used too rarely. Reversible encoding should work,
    so this encoding can be used with a conversion table.

    Somebody told me "never say never" ;-) In use of linux - why I could not add module in gconv by myself? Not sure about Windows nor OS/2.

    Again, thank you for your work.

    Karel

    --- GoldED+/LNX 1.1.5-b20240209
    * Origin: Plast DATA (2:423/39)
  • From Karel Kral@2:423/39 to Michiel van der Vlist on Sun Sep 20 07:54:22 2026
    Hello Michiel!

    19 Sep 26 11:46, you wrote to me:

    That will be a waste of time and energy. Just skip that fase and go directly to UTF-8. That is the future for Fidonet. For the rest of the digital world UTF-8 already is the present.

    You are (as usual) right.

    Majority in r42 agreed to wait a bit longer. But that is OK.

    Karel

    --- GoldED+/LNX 1.1.5-b20240209
    * Origin: Plast DATA (2:423/39)
  • From Eugene Subbotin@2:5075/35 to Karel Kral on Sun Sep 20 11:19:32 2026
    Hello Karel!

    Sunday September 20 2026 07:51, you wrote to me:

    What must be done: 1) add CP895 into iconv and 2) get Golded+
    working with 1byte against UTF and back conversion

    No one will ever add this encoding to iconv or the Windows
    distribution: itīs used too rarely. Reversible encoding should
    work, so this encoding can be used with a conversion table.

    Somebody told me "never say never" ;-) In use of linux - why I could
    not add module in gconv by myself? Not sure about Windows nor OS/2.

    Again, thank you for your work.

    Of course in Linux, you can write whatever you want for yourself. But I'm not sure that code like this will ever be accepted into the upstream repository. In Windows and OS/2, the set of included charsets is determined by Microsoft and IBM.

    Eugene

    ... It's full of stars!
    --- GoldED+/BSD 2.0.0-b20260912 (NetBSD 11.0 Intel Xeon Icelake)
    * Origin: FireFox Station (2:5075/35)
  • From Eugene Subbotin@2:5075/35 to Karel Kral on Mon Sep 21 18:28:16 2026
    Hello Karel!

    Sunday September 20 2026 07:54, you wrote to Michiel van der Vlist:

    That will be a waste of time and energy. Just skip that fase and
    go directly to UTF-8. That is the future for Fidonet. For the
    rest of the digital world UTF-8 already is the present.

    You are (as usual) right.

    Majority in r42 agreed to wait a bit longer. But that is OK.

    What's there to wait for? All you need to do in R42 is accept a document that clearly states that the primary encoding used is UTF-8. The fallback encodings are CP852 (for DOS) and LATIN-2 (for Unix). All of these encodings support Czech and Slovak characters, so there shouldnīt be any problems. The old CP895 is for read-only use-thereīs no future in it.


    Eugene

    ... It's full of stars!
    --- GoldED+/BSD 2.0.0-b20260912 (NetBSD 11.0 Intel Xeon Icelake)
    * Origin: FireFox Station (2:5075/35)
  • From Karel Kral@2:423/39 to Eugene Subbotin on Sun Sep 27 06:59:03 2026
    Hello Eugene!

    21 Sep 26 18:28, you wrote to me:

    What's there to wait for? All you need to do in R42 is accept a
    document that clearly states that the primary encoding used is UTF-8.

    Let say we have bad experience just accepting signed documents since Munich 1938 ;-) (just joking)

    1) coding from Kamicky brothers is better than others (text is readable even if you do not suppor it) - 2) we (in r42) agreed if anything we move to utf8 - where old setups which are used are not compatible. Most of people here.

    And honestly? There are not so many people there. Active players even less. I am happy that we can communicate within region.

    Karel

    --- GoldED+/LNX 1.1.5-b20240209
    * Origin: Plast DATA (2:423/39)