[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [zonecheck-locale] New to the "zonecheck-locale" mailing list, Germa
Re: [zonecheck-locale] New to the "zonecheck-locale" mailing list, German offered.
Sun, 29 Jun 2003 21:55:48 +0200
On Sun, Jun 29, 2003 at 02:18:47PM +0200, Purodha Blissenbach wrote:
> Hello zonecheck-locale readers,
> Being subscribed to this list since more than a week silently
> listening reading nothing ;-) I'm kind'a loosing temper and just
> introduce myself.
Good, you are the first to post to the mailing list. Champagne!
> Hi, I'm Purodha, senior software deveolper and consultant, co-founder
> of cooperative ISP http://hostsharing.net/ (website in German language
> only), maintainer of several web sites, etc.
> I'd like to add German language to zonecheck.
We will be more than happy to support a new language.
> Having recently checked out the cvs tree, and read through it,
> the process of adding a language seems pretty much straightforward
> due to good design and modularization. Well Done !
I was fearing that the splitting in so many files was a little
excessive and not using standard .po file a handicap.
It doesn't seem so, thank you.
> I've not started any activity yet, because:
> - I first wanted to make sure, noone else is already working on
> the task;
To my knowledge you are the first, I have added a new task
in savannah and if you agree i'll make you the personne
dealing with the german localisation.
> - Check back about how to submit additions (easy anyways, since
> it's simply a tree of additional files)
> - Is there a test case avaliable, that merely shows every message
> at least once, so as to support one in checking them in context ?
> Which is way better than translating isolated phrases, or, worse,
> groups of words!
Unfortunately there is no such test case:
- all the messages in the locale/test directory are used to display
information about a test (failure, name, success, ...) and
use the same model. It should not be too difficult to see one of them
in its different contexts by turning on the 'Output' or 'Report error'
options in the html form.
- testname, error, ok: should try to fit in 70 characters
(so it fits in one line when used in the console)
- explain: carriage return (\n) should be used to split the
text around the 70th charecter (later, splitting will be
automatically performed as it is currently done for details)
the first line should be the reference (tag 'ref')
or advice (tag 'adv')
- details: no particular constraint here.
- the cli.* for now only hold the help message (zc --help)
- the gtk.* is not fully designed/finished (so translation for it could wait)
- the zc.* the last half part are exceptions (error messages)
so they should be as short as possible to fit in one line
(the keyword ERROR: is displayed in front of it, and the %?
are replaced by a value)
> - I've seen few typing errors or minor grammatical glitches in
> English texts. How should I deal with them as encountering them ?
if you could make a patch and send it to this mailing list
or use the savannah patch manager.
Eventually if it only concern one or two errors you could
mention it and I'll fix them.
> - I want to let you know that, I'm not good @ French, so reading
> French original texts wouldn't help me much understanding what
> intended meanings are.
The original version is in english, but as it was written by a french
it can be ambiguous/wrong some times...
Let us know the translation you are not sure about and we will
see what we can do (it will certainly mean make the english
version more understandable)
> One thing that I didn't like in zonecheck: the html (main) exists
> once in every language. Since maintaining it means repeating design
> chancges in every language, I'd suggest to separate layout/design
> from text for these also. I know that, for some langage groups,
> like latin/character-based, arabic, far-east/chinese, you might
> wish to have different layouts as well, so well, that should not
> be a hindrance. I've two suggestions to choose from :-)
> - The individial language html files could be generated from a
> template and a language file, and then stored (in cvs or only)
> for the web server to read them.
> - Even the main html could be scripted.
Agreed but I haven't had time for that yet and the quickest way
was to have html files to show how it should be done to talk
with the cgi.
What was planned was to use the same localisation files to generate
the GTK interface and the HTML pages, unfortunately the GTK interface
is not a high priority task and all the needed strings
haven't been identified yet.
I've opened three bugs so I remember to do it.
> Had I written the html, I'd choosen lowercase tags, since my
> servers all use on-the-fly gzip compression (which really saves me
> traffic i.e. money) and lower case just compresses better.
why not, will see to make the changes.
(and i think xml tags are in lowercase, so it will be ready for xhtml)
> Btw. Accessing the web archive of th emailing list did not work
> for me - broken links.
it's a bug in mailman, when the archive is empty it doesn't display
a web page, just a broken link...