Closed (fixed)
Project:
Project
Version:
5.x-1.x-dev
Component:
Miscellaneous
Priority:
Normal
Category:
Bug report
Assigned:
Reporter:
Created:
29 Jun 2007 at 23:05 UTC
Updated:
28 Jul 2007 at 15:27 UTC
Jump to comment: Most recent file
There are many (not counted) untranslatable strings in the module. Add t() to the missing strings, please.
| Comment | File | Size | Author |
|---|---|---|---|
| #24 | project_ui_bugs_2.patch | 12.69 KB | hass |
| #22 | project_ui_bugs_1.patch | 13.05 KB | hass |
| #21 | project_ui_bugs_0.patch | 13.06 KB | hass |
| #19 | project_ui_bugs.patch | 12.62 KB | hass |
| #10 | project-release-history.patch | 1.03 KB | hass |
Comments
Comment #1
hass commentedI will add this POT file generation errors to this case, while this may be one to-do:
Comment #2
hass commentedHere you get a patch for many bugs i don't like to split for every file. takes to much time... POT file will be updated in a few seconds.
What does it fix?:
1. contains fix from http://drupal.org/node/155709
2. Fixes an HTML layout issue with "Automatically generated path alias"
3. renames "Go" info "Filter" what tells more what behind the button and is overtaken from content type filter
4. A wrongly used t() in project.inc
5. Many invalid BR tags (XHTML validation)
6. fixes all POT generation errors (except one non fixable)
7. fix value saved for project_release_unmoderate (may need an additional fix in .install file for updating older versions, but no harm without)
8. small text corrections (commas, etc.)
9. removes "#!/usr/bin/php" from package-release-nodes.php
Comment #3
hass commentedComment #4
dwwa) I hate monolithic patches.
b) #!/usr/bin/php is important in package-release-nodes.php -- it's a CLI script, not part of Drupal. That should stay.
c) Some of this is clearly an improvement, thanks. However, (a) and (b) makes it complicated, so I can't just add the improvements.
d) I'll take a look at the rest of this once the D6 code freeze hits and I have time.
Comment #5
hass commenteda) well, but with many files stupid work.
b) aha, like the perl shebang... not required on windows, but on unix it should stay there, sorry.
however here follow splitted patches... :-(
Comment #6
hass commented#2 file
Comment #7
hass commented#3 file
Comment #8
hass commented#4 file
Comment #9
hass commented#5
Comment #10
hass commented#6
Comment #11
hass commentedAny idea how to make the names of "sort methods" (date, name, category, etc) translatable?
Comment #12
drewish commentedhass, I think rather than breaking the patch up by file you need to break it up by goal. all the
<br />changes should go into one issue title XHTML validation, likewise the string changes should go into their own issue. that way each patch can correct the same type of bug in several files.i'm not going to review different 6 patches on one issue because it's too hard to figure out what each one is supposed to change and if it does so. i'd suggest splitting up these changes and marking this issue as won't fix.
Comment #13
drewish commentedComment #14
hass commented@drewish: dww made me to split up the files. I don't like this, however i've done it. if you don't like to review the small patches, review the big one in http://drupal.org/node/155727#comment-268528. This patch is all what is in the smaller ones, too.
It's took much time to review the module for this translation bugs and i don't like to create for every small glitch a different patch. This BR tags are bugs inside the translatable strings, therefor absolutely related to the translatable strings. We can start discuss, why BRs are inside a translatable string... well they shouldn't! Why should i not fix this XHTML bugs if i need to edit the same line with bugs in the translatable string. Don't be so fussy, please.
The bugs needs work, this patch not really.
Comment #15
dww@hass: I said "a) I hate monolithic patches."
That doesn't mean I like 1 issue with N tiny patches. Drewish is right. 1 issue per goal, please.
Comment #16
drewish commentedhass, i was trying to explain why dww told you to split them up, it he didn't want to review the changes to each file by themselves, he wanted to review the changes to correct each type of bug by itself. mixing up different bug fixes makes it harder for both reviewers and committers to see what's going on and verify that "yes, this patch does what it's supposed to".
it's a bit of a pain but you're the one trying to get changes made, so the burden is on you to present it in the way the committer wants it.
Comment #17
drewish commentedwhoops, cross post
Comment #18
hass commentedNow i understand what a monolithic patch means :-).
Well, i explained hardly in heavy and too much details what have been changed, but this is all translation / output related. Not so much XHTML related. check out the patch. I change translatable strings containing a BR, and additional saw BR does not contain a slash. Now if i create a patch and split this things up, the second patch will kick out my first patch. Patches that require another patch are not very easy to commit. you need to keep in mind the correct order... this makes things only complicated. Do you like this?
Comment #19
hass commentedReworked patch with the real goal.
Comment #20
dwwa) Doesn't follow coding style for string concatenations.
b) This is a step backwards:
c) This doesn't seem to have anything to do with what you're talking about, so I don't see why it's in this patch:
d) I thought this wasn't good form:
Core always seems to put the
<p>and</p>tags outside of t() calls, instead of smashing everything together into a giant t().Ahh, the joys[sic] of 1 big patch trying to do too many things...
Comment #21
hass commenteda) string concatenations fixed
b) What's wrong with this "char"? I'm not a native speaker, but isn't a "letter" a piece of paper you send by offline mail? A char is what the 'x' is... not sure if there is a difference between British / American English?
c) this fixes an UI issue, too. If this DIV is not surrounded the node page looks like
With the patch it looks like this:
d) yup, in general - yes. But this should be an exception... i cannot say why, but core have many strings containing HTML if it comes to complex text's. It's very difficult to prevent HTML at all - think about UL LI and such things. check out the .po files from core... big blocks are build in this way, too. however it should be avoided wherever possible!
i think - only if this is one very long text, without breaks in the text... if you need many breaks an exception should be in place.
Maybe "Mr international" can say why...
20 single lines very easy to review isn't very much, isn't it? there are no complex dependencies... only one line from the prior line what shouldn't be a big trick.
Comment #22
hass commentedmissed one older string concatenations bug, patch rebuild.
Comment #23
aclight commentedI agree with dww. First of all, it's "character", not char. I don't think an abbreviation should be used, since my guess is that many non-native English speakers might not know what a "char" is.
Secondly, I think "letter" is more clear than "character", at least in American English (but you are right, a letter is also a form of written communication). See http://www.answers.com/letter?cat=health
Comment #24
hass commentedrolled back the letter string, patch attached
Comment #25
dwwa) mostly done, still missing a few, but whatever, i'll fix those myself.
b) i'm a native speaker. yes, english is one of the worst languages on earth, no question. it's true that a "letter" is both what you make words out of ("letters" of the alphabet) and something you can make out of words (when you send a "letter" through the mail). go figure. and no, this isn't a difference with british vs. u.s. english, it's just one of the millions of insane aspects of this crappy language i grew up speaking. i suppose "character" would be ok here, though that's both an ASCII entity, and someone's personality. english sucks, what can I say? however, "char" isn't a word in english (well, it is, but it means to burn something, but that's not what you're talking about), it's a data type in some computer languages. that won't do. i'd rather leave it as "letter", and let people translate it however they want based on the context.
c) ok, sure, i see that now that i look closely (it's even on d.o), but this is strange, since i knew it didn't used to be like this when i first added the functionality. IMHO, this is a perfect example of a small, self-contained change that should have just been a separate issue with a trivial patch, since then it would have gone in immediately (if there was a clear description of the problem and why this fixes it). that said, i tend to prefer using #prefix and #suffix for wrapping form elements in a div so that it's easier to alter if someone wants to add a class or something.
d) i sent email to the translation list for input on this.
Side note: it's shocking that you care about translation and XHTML validation of project_release_update.php. ;) This is a one-time script, mostly for converting d.o from 4.7.x-1.* to 4.7.x-2.*. It requires manual editing for non d.o sites that were using the old "release" functionality in project* before the dawn of project_release.module. I don't think anyone cares about this script, and no one needs it (though it's possible there are a tiny handful of sites that were using the old release functionality that need a copy of it for their upgrade to D5).
Comment #26
hass commenteda) in my patch? i'd like to learn :-)
c) for sure, #prefix and #suffix is perfect...
About project_release_update.php - i only looked on the plain code, not if a script is a required script or not :-). should i re-roll the patch for C? sounds like you are working on the fixed and i don't need to do!?
Comment #27
dwwregarding d): Gabor has spoken: http://lists.drupal.org/pipermail/translations/2007-July/000400.html
Splitting into smaller t() and keeping the HTML out of it is the way to go.
However, yes, since this issue is bugging me, I'll just fix the remaining problems with it and get it in.
Comment #28
dwwHelp text changes moved into new patches for 2 separate issues, 1 new, 1 old:
http://drupal.org/node/57667
http://drupal.org/node/159334
Review and test those.
UI fixes committed separately:
http://drupal.org/node/159321
http://drupal.org/cvs?commit=73533
http://drupal.org/cvs?commit=73537
http://drupal.org/cvs?commit=73538
Where appropriate, each of these was backported to the right branches.
Remaining t() and XHTML changes fixed and committed:
http://drupal.org/cvs?commit=73554
http://drupal.org/cvs?commit=73558
Again, where appropriate, backported to the right branches.
@hass: This just took me almost 2 hours to properly deal with the mess you created with this issue (granted, I ran into the bug at http://drupal.org/node/57667 while testing, so that was a minor side-track, but still). If you ever want me to review any more patches from you in the future please read the commit messages, and study the diffs carefully for everything I just posted above. If you learn from this experience, you could become a productive contributor in the project* issue queue, and that's always welcome. If you don't, you're going to drive me insane and I'm going to just start closing any issue you open without comment. ;)
Cheers,
-Derek
Comment #29
(not verified) commented