Time |
S |
Nick |
Message |
11:03 |
|
kados |
paul: are you around? |
11:31 |
|
paul |
kados : yep |
11:46 |
|
kados |
paul: did you get my message? |
11:46 |
|
kados |
paul: asking about translations? |
11:47 |
|
paul |
yes. |
11:47 |
|
kados |
ahh, good |
11:47 |
|
paul |
sorry to have missed answering |
11:47 |
|
kados |
it's ok, I'm glad you got it, email has been flakey lately |
11:47 |
|
kados |
:-) |
11:47 |
|
kados |
so what do you think? is there a solution? |
11:47 |
|
paul |
my internet connexion is "flakey" as well... |
11:47 |
|
kados |
hehe |
11:48 |
|
paul |
(can I use it anywhere ?) |
11:48 |
|
kados |
hehe |
11:48 |
|
kados |
yes |
11:48 |
|
kados |
it's safe :-) |
11:48 |
|
kados |
so for unimarc it seems we need non-utf-8 |
11:48 |
|
kados |
but for marc21 we need utf-8 :( |
11:49 |
|
paul |
I think the best/only solution is to add a parameter on the commandline. |
11:49 |
|
paul |
and comment this part (that tries to detect encoding) |
11:49 |
|
kados |
but it means we will have css_opac_utf8.pot and css_opac_8859.pot ? |
11:50 |
|
paul |
mmm... not necessary probably |
11:50 |
|
paul |
we could just use utf8 for everybody, except for frenchies |
11:50 |
|
kados |
if it's OK with you, it's OK with me |
11:50 |
|
paul |
the "conflicting charset" comes from the fact that : |
11:51 |
|
paul |
in extracted strings, there are BOTH utf8 AND iso8859-1 |
11:51 |
|
paul |
because they are both in template, one being used by the <tmpl_if>, the other one discarded |
11:51 |
|
kados |
Conflicting charsets in msgid: <!-- vs UTF-8 |
11:51 |
|
kados |
I suspect that <!-- is unreadable |
11:51 |
|
paul |
but the translator don't understand what to do with those 2 strings. |
11:52 |
|
paul |
right. To solve this, i've commented the line & forced iso8859-1 in my local script |
11:52 |
|
paul |
(& I did not commit) |
11:52 |
|
kados |
ok, so if I force UTF-8 and commit it's OK with you? |
11:53 |
|
paul |
ok |
11:53 |
|
kados |
and you can handle french translations with 8859 |
11:53 |
|
kados |
ok, thanks |
11:53 |
|
paul |
ok. |
11:53 |
|
paul |
did you see my 2 new features commits on rel_3 |
11:53 |
|
paul |
? |
11:53 |
|
kados |
not yet, I'm behind on email :-) |
11:53 |
|
paul |
http://o15.bureau.paulpoulain.[…]c-tags_subject.pl |
11:53 |
|
paul |
http://o15.bureau.paulpoulain.[…]opac-topissues.pl |
11:54 |
|
kados |
wow, great! |
11:54 |
|
kados |
NPL will want this immediately |
11:54 |
|
paul |
less than half a day of work ;-) |
11:54 |
|
paul |
(has still to be improved, as explained in my commit) |
11:54 |
|
kados |
hehe |
11:55 |
|
paul |
(should commit in 1-2 hours) |
11:55 |
|
kados |
woot |
11:58 |
|
kados |
$ ./tmpl_process3.pl create -i ../../koha-tmpl/opac-tmpl/css/en/ -s po/css_opac.pot -r |
11:58 |
|
kados |
Removing empty file po/css_opac.pot |
11:58 |
|
kados |
msgmerge: /tmp/UPgXe0rX6o: warning: Charset "CHARSET" is not a portable encoding name. |
11:58 |
|
kados |
Message conversion to user's charset might not work. |
11:58 |
|
kados |
................................................... done. |
11:58 |
|
kados |
The create seems to be successful. |
11:58 |
|
kados |
paul: is that a safe warning? |
11:59 |
|
paul |
wow... |
11:59 |
|
paul |
I never saw this warning. or didn't remember it ! |
11:59 |
|
kados |
hehe |
11:59 |
|
paul |
a kbabel on the file would confirm things are OK |
11:59 |
|
kados |
I just added $charset_in = "UTF-8"; |
11:59 |
|
kados |
and took out the die |
13:11 |
|
kados |
tnb_: ! |
13:11 |
|
kados |
woops |
13:20 |
|
kados |
hey owen |
13:20 |
|
kados |
owen: how's vacation treating you? :-) |
14:50 |
|
kados |
_paul: are you avaialble? |
14:50 |
|
kados |
_paul: sorry to bother you, but default intranet-tmpl has many probs with translation script |
16:42 |
|
chris |
morning |
16:47 |
|
kados |
hey chris |
16:48 |
|
kados |
wow :-) |
16:48 |
|
chris |
yeah i completely misread my clock |
16:48 |
|
chris |
hehe |
16:49 |
|
kados |
did you see paul's new commits |
16:49 |
|
kados |
http://o15.bureau.paulpoulain.[…]c-tags_subject.pl |
16:49 |
|
kados |
subject cloud ^^ |
16:49 |
|
kados |
http://o15.bureau.paulpoulain.[…]opac-topissues.pl |
16:49 |
|
chris |
woo cool |
16:49 |
|
kados |
popularity I think is the second one |
16:50 |
|
chris |
should show david bigwood that, he was hot on tags |
16:50 |
|
kados |
I'd like to get that subject cloud up somewhere in production |
16:50 |
|
kados |
yea, but it's busted ... probably better fix it first :-) |
16:50 |
|
chris |
this is on 3.0 code? |
16:50 |
|
kados |
yea |
16:50 |
|
kados |
rel_3 |
16:50 |
|
kados |
we'll back-port it to dev_week |
16:50 |
|
kados |
once I get a spare moment :-) |
16:51 |
|
chris |
:) |
16:51 |
|
chris |
where does it fetch the subjects from? |
16:51 |
|
kados |
dunno, haven't looked at the code yet :-) |
16:52 |
|
kados |
if I were to write it I'd do it as a zebra scan |
16:52 |
|
chris |
ill have to have a look |
16:52 |
|
chris |
yep |
16:52 |
|
kados |
which tracks the count as well |
16:52 |
|
chris |
if its modular, it might be easy to get going for 2.2 .. using the bibliosubjects table |
16:53 |
|
kados |
ahh, yea |
16:53 |
|
kados |
that'd be very cool |
16:53 |
|
kados |
and easy |
16:53 |
|
chris |
yep, ill have to have a look at the code |
16:53 |
|
kados |
I'd like to do facets for rel_2_2 also |
16:53 |
|
chris |
cool |
16:53 |
|
kados |
maybe for 2.2.9 |
16:53 |
|
chris |
yep |
18:20 |
|
kados |
thd: hi there |
18:20 |
|
kados |
thd: how's things? |
18:21 |
|
thd |
kados: I have been up all night debugging some python code in Zope which had worked fine 2 years ago |
18:21 |
|
kados |
bummer |
18:22 |
|
thd |
I think that my hosting provider, a Zope specialist upgraded Zope and thus introduced a new bug |
18:25 |
|
thd |
kados: my hosting provider had rebooted the server and I lost my screen session a day ago |
18:25 |
|
thd |
kados: is the screen session not automatically preserved in a detached state if the machine reboots or even if power is lost? |
18:30 |
|
kados |
correct, it's not preserved |
18:31 |
|
kados |
the only way to preserve it is to set it up in your .screenrc |
18:31 |
|
kados |
then you can at least start up the same sessions as before |
19:07 |
|
thd |
kados: do you know what screen option allows recovery of an undetached screen session following a sudden reboot or after power loss? |
19:08 |
|
thd |
kados: I have checked the man pages and online documentation without finding an obvious .screenrc option for saving an interrupted screen session. |
19:17 |
|
kados |
there is no way to recover an undetacthed screen session fllowing a sudeen reboot |
19:17 |
|
kados |
it's the one caveat of screen |
19:24 |
|
chris |
a reboot of the machine you are running screen on eh? |
19:27 |
|
thd |
chris: yes, I guess the last time it happened I was disconnected from the server long enough for my session to be automatically detached |
19:28 |
|
chris |
well if the machine running screen reboots, ur screen will be dead |
19:28 |
|
mason |
my screen sessions hang/crash/freeze all the time :) |
19:28 |
|
chris |
mine dont :) |
19:29 |
|
chris |
but if you ssh to a machine, run screen on that, ... and you crash then you can get your ssh connection back up |
19:29 |
|
chris |
kill the old one, if it hasnt timed out |
19:29 |
|
chris |
and screen -r -d |
19:30 |
|
chris |
it will only be the zombie connection that is stopping you, so if you kill it, you're away again |
19:30 |
|
thd |
mason I can crash screen by doing some strange IO operations form the clipboard when running screen in an xterm |
19:32 |
|
mason |
i havent been able to work out what causes my screen weirdness |
19:32 |
|
thd |
s/form/from/ |
19:36 |
|
thd |
mason too much copying and pasting between an xterm console and an xterm GUI will crash my x windows session so I try to avoid pasting text between consoles and GUIs when I can. |
21:07 |
|
rch |
hey mason |
21:09 |
|
mason |
heya |
21:09 |
|
rch |
and hi chris |
21:09 |
|
rch |
gotta quick spiinelabel question... |
21:09 |
|
mason |
sure |
21:10 |
|
rch |
the pdf generated has name of script (ends in .pl) |
21:10 |
|
rch |
is there a way to get a .pdf? |
21:10 |
|
chris |
thats how it used to work |
21:10 |
|
chris |
we made a file, then redirected to it |
21:11 |
|
mason |
hmm, the orig. way was to write the file out to a new.pdf in a temp dir |
21:11 |
|
chris |
but that meant you needed a dir that was writable by the webserver |
21:11 |
|
mason |
yep, then redirect... |
21:11 |
|
chris |
and everyone kept screwing that up :) |
21:11 |
|
chris |
so it now returns the pdf on the fly |
21:11 |
|
mason |
yep |
21:11 |
|
rch |
easy to do the redirect? |
21:12 |
|
chris |
it could be a syspref eh mason? |
21:12 |
|
mason |
sure could |
21:12 |
|
chris |
that might be the best way to do it |
21:13 |
|
mason |
so the problem is that if you click on 'save' in acrobat or moz, it brings up a dialog box with 'label-print-pdf.pl' as a filename? |
21:13 |
|
mason |
instead of barcode.pdf? |
21:14 |
|
chris |
i wonder if you can do something with the header in cgi |
21:14 |
|
mason |
perhaps another way, is to keep the nice on-the-fly code, but set the filename to blah.pdf instead? |
21:14 |
|
mason |
(which is what chris is suggesting it think:) |
21:15 |
|
chris |
you cant call the script .pdf ... cos then apache will serve it instead of executing it, but im wondering if we can do something with the header |
21:15 |
|
rch |
in os x, it doesn't seem to recognize mime type |
21:16 |
|
rch |
so you either have to assocate .pl with acrobat |
21:16 |
|
rch |
or choose app every time |
21:16 |
|
rch |
unless i'm doing something stupid... |
21:16 |
|
chris |
what does the header set currently mason? |
21:17 |
|
chris |
this might be the ticket |
21:17 |
|
chris |
The -attachment parameter can be used to turn the page into an attach- |
21:17 |
|
chris |
ment. Instead of displaying the page, some browsers will prompt the |
21:17 |
|
chris |
user to save it to disk. The value of the argument is the suggested |
21:17 |
|
chris |
name for the saved file. In order for this to work, you may have to |
21:17 |
|
mason |
gimme a tic |
21:17 |
|
chris |
set the -type to "application/octet-stream". |
21:20 |
|
chris |
ahh doing it the hard way |
21:20 |
|
chris |
print STDOUT "Content-Type: application/pdf \r\n\r\n"; |
21:20 |
|
chris |
id make that line |
21:21 |
|
chris |
my $input = new CGI; |
21:22 |
|
chris |
print $input->header(-type=> |
21:23 |
|
chris |
'application/pdf', -attachment=>'barcode.pdf'); |
21:25 |
|
rch |
beautiful |
21:25 |
|
chris |
it might, or might not work :) |
21:25 |
|
chris |
but its worth a try |
21:25 |
|
rch |
works great. |
21:25 |
|
chris |
sweet |
21:25 |
|
chris |
commit that sucker |
21:28 |
|
rch |
thx chris |
21:28 |
|
mason |
they call him 'the surgeon'... |
21:29 |
|
rch |
:) |
21:31 |
|
chris |
heh |
21:32 |
|
mason |
kinda catchy aye |
21:33 |
|
chris |
ill put it on my business card |
22:07 |
|
rch |
mason: oops- wasn't paying attention. I double-committed. |
22:07 |
|
rch |
I'll clean up my mess |
22:07 |
|
rch |
:) |
22:18 |
|
rch |
mason: reverted... there _are_ two cgi objects created in that script. |
22:24 |
|
mason |
hiya |
22:24 |
|
mason |
you mean the 'print $cgi->redirect("/intranet-tmpl/barcodes/new.pdf");' |
22:25 |
|
mason |
on line 177 ? |
22:25 |
|
rch |
right, there's a $cgi and a $input |
22:26 |
|
mason |
aah, right |
22:26 |
|
mason |
the $cgi is an exisitng one from the orig ver., that i didnt completely pull out |
22:27 |
|
mason |
its not needed now it think. |
22:27 |
|
mason |
its not needed now i think. |
22:28 |
|
mason |
ill fix that now... |
22:31 |
|
mason |
that seems to work |
22:32 |
|
rch |
yep. looks good on my end. |
22:33 |
|
mason |
cool, ill try and sort out a syspref for this |
22:33 |
|
mason |
your pdf now opens up automagically? |
22:33 |
|
chris |
shouldnt need one now |
22:33 |
|
rch |
sorry to step on your toes... too much multitasking |
22:33 |
|
mason |
np :) |