h a l f b a k e r yA hive of inactivity
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Please log in.
Before you can vote, you need to register.
Please log in or create an account.
|
A website that allows any project in any programming
language, to do i18n (internationalization) translation of the
user interface and experience to various (human) languages,
including the graphics and locations of objects for BIDI code.
The output is source code that the project managers
and
programmers can use to immediately have a translation out
of
the box.
Since it will be open source and free, all the programs in any
environment will be happy to use it.
Since it will be working with many programs, many of the
words can be automatically translated with an initial
translation. [thanks scad]
A dynamic "screenshot" of the messages, and the software
itself with translated parts inserted in it will be available to
the manager as well, and code to run the program and have
it
tested for language readability and intuitiveness.
What started me off on this...
https://trac.videol...wiki/HowToTranslate
[pashute, Nov 11 2014]
[link]
|
| |
// words can [be] automatically translated // |
|
| |
I'm not 100% sure what the flow you intend to use
is. My first impression was that the translations
would be made real time by the application
querying for the desired translation, but on a
second reading I don't think that's what you
meant. |
|
| |
Is it a system of tools for translation that requires
the tools to be run online? The advantage would
be that the translations made by users end up
being stored online, forming a database so that
future software translation with similar text can
be translated somewhat automatically. |
|
| |
Also, if the tools are used to generate several
"official" translations included with the software
package, a user in a non-supported language could
request an auto-translation from the online
system. The auto-translated version probably
wouldn't be "good", but might be marginally
usable, and users could tweak the automatic
translation, fixing grammar, adjusting graphics
positions, etc. the modified auto-translations
could be stored online for the benefit of other
users. There would of course need to be some
kind of voting system and a way to suggest
changes to an existing translations (branch
translation and modify, then send a message).
The software author could branch a good user-
submitted translation (branching will automatically
give credit to the contributors), set it as the
default translation for a particular language, and
include it in future releases. The author would
have the option to flag (but not remove)
translations that were deemed to not be serious
translations. That would allow people to have
their fun, but avoid causing trouble for people
who actually want to find a usable translation to
get work done. Maybe the rating system needs a
five star rating for quality/usefulness and a
separate 5 star rating for humor. |
|
| |
I can imagine automated translation of source code
between different languages. However, I'm not
convinced that the result would be very easy to
understand, update or maintain. |
|
| |
Aha. So if I had a program I'd written in C++, and
wanted it in Python or Silkworm or Ptarmigan or
whatever, would I get back something I could
understand? |
|
| |
It's easy enough to do a dynamic translation including character sets, and pretty much impossible to get it so that they'll all look right, on the page. |
|
| |
//program written// nah, human languages: auto-translating programs from one computer language to another never work with any semblance of sanity. |
|
| |
In fact I have to the best of my knowledge a deep
understanding of internationalization in software. I lead
internationalization of the first VOD Phone in the world, at
IBM, and then participated in translation of several open
source packages. I also was part of the giant effort of
translation of Google Translate, and of Facebook. |
|
| |
And no, I was not proposing programming language
conversion such as conversion from ADA to COBOL, or from
java to C# (just change the capital letters to lowercase and
vice versa). |
|
| |
There are various ways to achieve internationalization.
Most software is "closed" in DLLs etc. so that the user
cannot easily change the foundations such as the menus,
the texts and the warning messages. The "strings" are
usually kept in a "hardcoded" table or in a database and
loaded dynamically into the program. Translators are
given a table with the words to translate, and the graphics. |
|
| |
Things start getting really complicated when dealing with
BIDI (bi-directional) languages written from right-to-left,
with English or other left-to-right languages inserted in
between. Some BIDI languages are Japanese, Hebrew and
Arabic. |
|
| |
For example, in an online two page book, the cover of a
RTL book is on the BACK of the book with LTR.
But the cover page is printed in one peace (for the cover,
the back and the side). So all graphics have to be re-
aligned. The size of the text is also a consideration. The
font and the shape are many times part of the graphic... |
|
| |
And the worst part is that literal translation is not always
what is desired. When the unfortunate PM of Argentine
visited Brasil in an historical event, he translated from his
native Castillian to Portuguese and said: "We've had our ups
and downs". Only to hear the hundreds of thousands of
Brazilians burst out laughing. |
|
| |
so umm "wouldn't it be nice if..." |
|
| |