[rfc][icedtea-web] Add new PL localized messages

[Jacob Wisor]

On 06/24/2014 02:38 PM, Jiri Vanek wrote:
> On 06/23/2014 05:48 PM, Jiri Vanek wrote:
>> On 06/20/2014 01:58 PM, Jacob Wisor wrote:
>>> Hello there!
>>>
>>> Sorry for delaying this for so long, but it's finally here.
>>>
>>> As always, the comments in Messages_pl.properties are for your convenience
>>> during review only and
>>> will get removed before push into repo.
>>>
>>> Thank you for reviewing
>>>
>>> Jacob
>>
>>
>> Hi!
>>
>> Attached are results of the testruns.
>>
>>  From quick glance I have no objections against messages as they are.
>>
>> Feel free to fix+push. discus/fix/push, discus/push  ir just push how time
>> permits you.
>>
>> J.
>>
>> //hmmm the files are somehow strange, I hope you will be able to read them :(
>
> Here are fixed and cleaned files.
> The files which are redundant in 1.5 can be skipped (but well can be safely
> deleted)
> Other errors should be fixed for head and 1.5.

Although these test result files are difficult to read, my overall impression is 
that for the most part everything is okay. The errors that are reported for head 
are due to even newer additional messages. My intention so far has been to only 
get localization up to date with 1.5 (as initially requested). So, the proposed 
PL patch seems to be up to the task and does not require any technical fix 
(please feel free to correct me if I am wrong).

As a side note: The overall quality and usefulness of the test could be 
increased by making the results (warnings/errors) more legible (comprehensible) 
and by only doing comparison tests regarding default (EN) vs non-default 
languages. DE vs CS or PL vs DE tests bear little or no useful information at 
all, or are even rather counter productive because the produced results only 
cause confusion.
The test should check mainly only for (1) completeness, that is whether all 
messages in the default language have /existing/ equivalents in every 
non-default language, and for (2) inequality of message values for the same 
message (assuming unequal message values as having been localized). There also 
may be a third test for (3) pruned messages where no messages exist for the 
default language but "orphan" messages exist for the non-default language.

Such simple tests with readable results would greatly ease the task of keeping 
track of new, forgotten, overlooked, or pruned messages.

Jacob