Ulf Vesterlund wrote:
In my opinion the # 1 priority is to solve the problem with
programflow. Now directed by the value of the buttons, i.e. the
text on the button is the name of the subroutin and if they don't
match the program don't work.
...
I don't know what Gunnar means with *cleaning up the code*.
Are you changeing it to avoid this problem?
Well, that's two different things. "Cleaning up" is about making the
program pass 'use strict;' and '-w' tests, i.e. a form of validation of
the code. But please don't worry about the buttons; that issue is
addressed, and I have already prepared a solution to it.
In my opinjon this is the main points:
1. Make program flow language independet.
2. Move the text out from the program itself.
3. Translation should not require any special knowledge.
Agree.
As regards the rest of your posting, Ulf, I think we should distinct
between
1) the overall design of the language feature
2) the format of the language files
3) other aspects of customization.
My idea about the overall design includes:
- An option to choose a default language in rlconfig.pm for the
system.
- An option for the ringmaster to choose some other default
language for a ring.
- An ability to determine the output language via the language
preference settings of the visitor's browser.
Accordingly, future releases should include a number of different
complete language files.
The format of the lanugage files is another aspect, which has been
discussed in the most recent postings on this topic. You are right in
that much of the output is a mixture of text strings and various
variables, and that has to be taken into consideration.
In a few weeks I hope to be able to release a test version of Ringlink,
with a limited translation feature built-in. That would simplify the
discussion about different possible design solutions.
I don't think we should make this discussion even more complicated by
involving the possibilities to customize the layout of the output. You
can already do a lot of customization by editing the existing templates
(mainhtml.pm, mailhtml.pm, and admhtml.pm), and an expansion of the
customization options has lower priority IMO.
And it is only a short step to make it possible for each
ringmaster to redesign the output for his ring, even in a
different language.
I disagree on this one. A ringmaster who wants full control over the
layout of the output should rather install and run his/her own copy of
Ringlink.
SandridgeFarm@xxxxxxx wrote:
Will they click a Translation button on say the main admin or
ringadim page, select a language from a long list, click submit,
and then have all files translated into that language, or
what...........???
Basically, yes. Doesn't it sound like a nice feature?
Wouldn't it be easier to just have the Ringlink program
available for download in several different languages, with all
files including the index.html file already translated into
that language?
No, Pete, that wouldn't be easier. One idea with making use of (some
kind of) language files is that there will only be one version of the
actual program to maintain.
Ulf Vesterlund again:
Who will decide what languages and who will do the translations?
Will RSM native to non official Ringlink-languages be forced to
use one of the aproven languages?
IMHO I think that it is important that translation can be done
by *anybody*. So anybody, including my friend
Imo-Dokzyxm-Ixticon of the Xmtipoxika-nation kan translate och
run Ringlink in his native language Xklopitkinsa, if he wants to!
Honestly, I don't see the problem here. Of course future releases of
Ringlink should be distributed with different language files.
(Personally I will only engage in the most important languages: English
and Swedish. ;-) But at the same time, I fully agree that it should be
technically easy to translate into further languages.
Who will do the translations, you ask. How about Ringlink users around
the world? Don't forget that Ringlink is an open source program, which
means that all translations have to be publically available. I hope and
believe that the number of distributed language files will increase
gradually. And when your friend is finished with his translation into
Xklopitkinsa, I'll be happy to include that file as well in the Ringlink
distribution.
I have one concern in this respect: How do we ensure the quality of the
translations? Probably we don't. But that problem may not prevent us
from developing Ringlink to a multilingual application.
/ Gunnar