You are viewing a plain text version of this content. The canonical link for it is here.
Posted to derby-dev@db.apache.org by David Van Couvering <Da...@Sun.COM> on 2005/08/30 23:49:25 UTC
Message translation question
I am curious. What process is followed when someone adds a new message
to messages.properties. How do folks who want these messages translated
find out what new messages need to be translated in each release?
Thanks,
David
Re: Message translation question
Posted by Daniel John Debrunner <dj...@debrunners.com>.
Myrna van Lunteren wrote:
> And, who translates the messages?
Same as any other piece of work in an open source project, anyone who
wants to, anyone who needs it, anyone who has that itch.
Dan.
Re: Message translation question
Posted by Myrna van Lunteren <m....@gmail.com>.
On 8/30/05, David Van Couvering <Da...@sun.com> wrote:
>
> I am curious. What process is followed when someone adds a new message
> to messages.properties. How do folks who want these messages translated
> find out what new messages need to be translated in each release?
>
> Thanks,
>
> David
Thx for asking this - I've been wondering the same thing...
And, who translates the messages?
(my reason: I looked at the patch for DERBY-117 - because I wanted to
address DERBY-385)
Re: Message translation question
Posted by Rick Hillegas <Ri...@Sun.COM>.
Thanks for the clarification, Dan. With community backing, I suppose
that a release coordinator could tighten this up.
Cheers,
-Rick
Daniel John Debrunner wrote:
>Rick Hillegas wrote:
>
>
>
>>Thanks, Andrew. Now let me compound the confusion: The Supported
>>Territories section of the Derby Reference Manual lists some nine
>>languages which Derby claims to support. A reasonable person might
>>suppose this means that, if you use one of these locales, then you'll be
>>able to understand your first syntax error rather than puzzling over a
>>"Message ID not found" diagnostic.
>>
>>o Is the release coordinator responsible for rounding up translators for
>>all nine languages?
>>
>>
>
>No. If no-one has volunteered to translate new messages into, say
>French, that so be it. It's exactly the same as if no-one had
>volunteered to fix a (non show-stopper) bug. The release would go out
>with the bug and similarly the release would go out with a few errors
>that would display English text when in a French locale.
>
>Maybe we could be more proactive and have a page (automatically
>generated) that showed which messages had not been translated, thus
>allowing volunteers to submit such translations.
>
>
>
>>o Or might it be helpful to add some explanation to the Supported
>>Territories section to adjust users' expectations? If so, what should we
>>say? What distinguishes supported from unsupported locales?
>>
>>
>
>I think supported means that a locale file exists, and unsupported means
>it doesn't and you will get error messages in English.
>
>Dan.
>
>
>
>
Re: Message translation question
Posted by Daniel John Debrunner <dj...@debrunners.com>.
Rick Hillegas wrote:
> Thanks, Andrew. Now let me compound the confusion: The Supported
> Territories section of the Derby Reference Manual lists some nine
> languages which Derby claims to support. A reasonable person might
> suppose this means that, if you use one of these locales, then you'll be
> able to understand your first syntax error rather than puzzling over a
> "Message ID not found" diagnostic.
>
> o Is the release coordinator responsible for rounding up translators for
> all nine languages?
No. If no-one has volunteered to translate new messages into, say
French, that so be it. It's exactly the same as if no-one had
volunteered to fix a (non show-stopper) bug. The release would go out
with the bug and similarly the release would go out with a few errors
that would display English text when in a French locale.
Maybe we could be more proactive and have a page (automatically
generated) that showed which messages had not been translated, thus
allowing volunteers to submit such translations.
> o Or might it be helpful to add some explanation to the Supported
> Territories section to adjust users' expectations? If so, what should we
> say? What distinguishes supported from unsupported locales?
I think supported means that a locale file exists, and unsupported means
it doesn't and you will get error messages in English.
Dan.
Re: Message translation question
Posted by Rick Hillegas <Ri...@Sun.COM>.
Thanks, Andrew. Now let me compound the confusion: The Supported
Territories section of the Derby Reference Manual lists some nine
languages which Derby claims to support. A reasonable person might
suppose this means that, if you use one of these locales, then you'll be
able to understand your first syntax error rather than puzzling over a
"Message ID not found" diagnostic.
o Is the release coordinator responsible for rounding up translators for
all nine languages?
o Or might it be helpful to add some explanation to the Supported
Territories section to adjust users' expectations? If so, what should we
say? What distinguishes supported from unsupported locales?
Thanks,
-Rick
Andrew McIntyre wrote:
>
> On Aug 31, 2005, at 11:02 AM, Rick Hillegas wrote:
>
>> Who rounds up translators? Is localization (to languages other than
>> American English) left to productizing groups outside the community
>> after Apache Derby posts a release?
>
>
> Exactly. For instance, IBM recently translated the new messages in
> 10.1, and I am looking at having these translation contributed back to
> Derby. It shouldn't be a problem this time around, because there
> haven't been any changes to the translated messages by others.
>
> In the future, though, there is a question of how not to stomp on
> somebody else's contribution if they provide a bug fix with translated
> messages, or a full set of translated messages. Whose translation is
> correct? How would one arbitrate the decision of who's message
> actually goes in?
>
> andrew
>
>
Re: Message translation question
Posted by Daniel John Debrunner <dj...@debrunners.com>.
Andrew McIntyre wrote:
>
> On Aug 31, 2005, at 3:05 PM, Francois Orsini wrote:
>
>> You will always run into the situation where noone can review some
>> translated messages - it should work the same as getting code changes
>> reviewed - Can code changes be checked-in without getting reviewed? ;)
Yes, code can be checked in without review. Derby is following
Commit-then-Review, if a committer has confidence in a change it can be
committed.
http://www.apache.org/foundation/glossary.html#CommitThenReview
> Ok, but let's say that I have a set of translations for the 10.1
> messages (which I do). Can I not check them in because the community on
> derby-dev lacks the language expertise to review all the translated
> messages?
>
> I trust that the translations are correct. My employer paid for the
> translations and I have reason to believe that the quality of the
> translation is high. That doesn't mean that you should necessarily
> trust that the translations are correct. But, neither of us has the
> expertise to verify that they are all correct.
As a committer you can check the changes in. Translations follow the
same rules as the code.
Dan.
Re: Message translation question
Posted by Andrew McIntyre <mc...@gmail.com>.
On Aug 31, 2005, at 3:05 PM, Francois Orsini wrote:
> You will always run into the situation where noone can review some
> translated messages - it should work the same as getting code
> changes reviewed - Can code changes be checked-in without getting
> reviewed? ;)
Ok, but let's say that I have a set of translations for the 10.1
messages (which I do). Can I not check them in because the community
on derby-dev lacks the language expertise to review all the
translated messages?
I trust that the translations are correct. My employer paid for the
translations and I have reason to believe that the quality of the
translation is high. That doesn't mean that you should necessarily
trust that the translations are correct. But, neither of us has the
expertise to verify that they are all correct.
Do I sit on these translated messages forever more, and not
contribute them? Or can we all agree to trust the translations, even
though we can't personally verify their quality?
andrew
Re: Message translation question
Posted by Francois Orsini <fr...@gmail.com>.
You will always run into the situation where noone can review some
translated messages - it should work the same as getting code changes
reviewed - Can code changes be checked-in without getting reviewed? ;)
On 8/31/05, Andrew McIntyre <mc...@gmail.com> wrote:
>
> On 8/31/05, Daniel John Debrunner <dj...@debrunners.com> wrote:
> > David W. Van Couvering wrote:
> >
> > > I would suspect that translations should be reviewed and committed
> like
> > > any other patch, and svn functionality would be used to catch
> conflicts.
> > > Reviewing messages does assume someone other than the submitter knows
> > > the language :)
> >
> > Exactly, it's just like code really.
>
> With the exception that everyone here can read code and not everyone
> here can read, say, Korean.
>
> What if somebody checks in a change to a previous translation in a
> language where noone in the community is capable of reviewing it? Do
> you just assume it is ok, until a bug is reported against the message?
> For expediency's sake that seems like the thing to do, but we might
> want to enshrine that as policy, so that if the question arises in the
> future, somebody can point back and say the decision was made to
> accept all contributions of translated materials even if there is no
> one capable of actually reviewing the content.
>
> Just a thought... :-)
>
> andrew
>
Re: Message translation question
Posted by Andrew McIntyre <mc...@gmail.com>.
On 8/31/05, Daniel John Debrunner <dj...@debrunners.com> wrote:
> David W. Van Couvering wrote:
>
> > I would suspect that translations should be reviewed and committed like
> > any other patch, and svn functionality would be used to catch conflicts.
> > Reviewing messages does assume someone other than the submitter knows
> > the language :)
>
> Exactly, it's just like code really.
With the exception that everyone here can read code and not everyone
here can read, say, Korean.
What if somebody checks in a change to a previous translation in a
language where noone in the community is capable of reviewing it? Do
you just assume it is ok, until a bug is reported against the message?
For expediency's sake that seems like the thing to do, but we might
want to enshrine that as policy, so that if the question arises in the
future, somebody can point back and say the decision was made to
accept all contributions of translated materials even if there is no
one capable of actually reviewing the content.
Just a thought... :-)
andrew
Re: Message translation question
Posted by Daniel John Debrunner <dj...@debrunners.com>.
David W. Van Couvering wrote:
> I would suspect that translations should be reviewed and committed like
> any other patch, and svn functionality would be used to catch conflicts.
> Reviewing messages does assume someone other than the submitter knows
> the language :)
Exactly, it's just like code really.
Dan.
Re: Message translation question
Posted by "David W. Van Couvering" <Da...@Sun.COM>.
I would suspect that translations should be reviewed and committed like
any other patch, and svn functionality would be used to catch conflicts.
Reviewing messages does assume someone other than the submitter knows
the language :)
David
Andrew McIntyre wrote:
>
> On Aug 31, 2005, at 11:02 AM, Rick Hillegas wrote:
>
>> Who rounds up translators? Is localization (to languages other than
>> American English) left to productizing groups outside the community
>> after Apache Derby posts a release?
>
>
> Exactly. For instance, IBM recently translated the new messages in
> 10.1, and I am looking at having these translation contributed back to
> Derby. It shouldn't be a problem this time around, because there
> haven't been any changes to the translated messages by others.
>
> In the future, though, there is a question of how not to stomp on
> somebody else's contribution if they provide a bug fix with translated
> messages, or a full set of translated messages. Whose translation is
> correct? How would one arbitrate the decision of who's message
> actually goes in?
>
> andrew
>
>
Re: Message translation question
Posted by Andrew McIntyre <mc...@gmail.com>.
On Aug 31, 2005, at 11:02 AM, Rick Hillegas wrote:
> Who rounds up translators? Is localization (to languages other than
> American English) left to productizing groups outside the community
> after Apache Derby posts a release?
Exactly. For instance, IBM recently translated the new messages in
10.1, and I am looking at having these translation contributed back
to Derby. It shouldn't be a problem this time around, because there
haven't been any changes to the translated messages by others.
In the future, though, there is a question of how not to stomp on
somebody else's contribution if they provide a bug fix with
translated messages, or a full set of translated messages. Whose
translation is correct? How would one arbitrate the decision of who's
message actually goes in?
andrew
Re: Message translation question
Posted by Rick Hillegas <Ri...@Sun.COM>.
Who rounds up translators? Is localization (to languages other than
American English) left to productizing groups outside the community
after Apache Derby posts a release?
-Rick
David Van Couvering wrote:
> Great, sounds reasonable, thanks.
>
> David
>
> Daniel John Debrunner wrote:
>
>> David Van Couvering wrote:
>>
>>
>>
>>> I am curious. What process is followed when someone adds a new message
>>> to messages.properties. How do folks who want these messages
>>> translated
>>> find out what new messages need to be translated in each release?
>>>
>>
>>
>> Couple of options:
>>
>> 1) The translator keeps a copy of the english messages.properties
>> (either directly or through an svn version number) and performs a diff
>> of the orignal and the new file when they want to translate.
>>
>>
>> 2) Just for added messages, one can write a simple Java program that
>> shows which messages exist in the english file but not in the translated
>> file. This is easy because the message files are properties files, thus
>> they can easily be loaded as properties objects and compared.
>>
>> Dan.
>>
>>
>>
Re: Message translation question
Posted by David Van Couvering <Da...@Sun.COM>.
Great, sounds reasonable, thanks.
David
Daniel John Debrunner wrote:
>David Van Couvering wrote:
>
>
>
>>I am curious. What process is followed when someone adds a new message
>>to messages.properties. How do folks who want these messages translated
>>find out what new messages need to be translated in each release?
>>
>>
>
>Couple of options:
>
>1) The translator keeps a copy of the english messages.properties
>(either directly or through an svn version number) and performs a diff
>of the orignal and the new file when they want to translate.
>
>
>2) Just for added messages, one can write a simple Java program that
>shows which messages exist in the english file but not in the translated
>file. This is easy because the message files are properties files, thus
>they can easily be loaded as properties objects and compared.
>
>Dan.
>
>
>
Re: Message translation question
Posted by Daniel John Debrunner <dj...@debrunners.com>.
David Van Couvering wrote:
> I am curious. What process is followed when someone adds a new message
> to messages.properties. How do folks who want these messages translated
> find out what new messages need to be translated in each release?
Couple of options:
1) The translator keeps a copy of the english messages.properties
(either directly or through an svn version number) and performs a diff
of the orignal and the new file when they want to translate.
2) Just for added messages, one can write a simple Java program that
shows which messages exist in the english file but not in the translated
file. This is easy because the message files are properties files, thus
they can easily be loaded as properties objects and compared.
Dan.