I will start with qouting Gunnars points on the ML.
He suggested to do the programming discussion here.
> 1. An exact description of the function.
The module should choose the right language file
according to the configuration file. There will be a
variable for the default language and if this language
should allways be used or if the script chooses the
appropiate language according to the browsers
references. For the last idea there will be a fallback
to the default-language. I would try to write a
IMO the language-file should look like
etc, it's a simple format and it the language files
can be written easly. The module now reads this
language file and returns a big hash.
The script doing that should do some syntak checking
Every line that doesn't start with a number is ignored.
If the number isn't followed by a colon it will be
ignored as well.
If there is nothing behind the colon a space should be
The linefeed at the end of each line should be cut off.
I planned to cut double spaces to one space.
I planned to cut spaces before and behind the number.
I'm still not sure if we should allow all letters or
just the standard letters a-z and A-Z. If we allow
extra letters a letter to html-tag converter needs to
be written. Otherwise the language file authors have
to write the html-tags into the language file.
> 2. Considerations as regards *maintaining* the
language files in the
> future in connection with program changes and/or
The goal should be to at least maintain 3-4 major
The programmers need to maintain the English language
> 3. Exchange of ideas about possible programming
I'm already writing on some points from #1.
> 4. Decision on the programming solution.
Up to now my code is still flexible. :)
> 5. Writing the code including creation of the
I already started doing the first part.
> 6. Testing.
Up to now I'm just doing it for my subs.
> 7. Release of a new program version.
> 8. Creation of non-English files.
Has anyone further suggestions and/or comments for the