Time |
S |
Nick |
Message |
13:07 |
|
kados |
hdl: you around? |
13:08 |
|
kados |
hdl: please verify that all your patches have been pushed up |
13:08 |
|
kados |
hdl: also, I have a question for you if you have a moment |
13:11 |
|
paul |
hi kados. |
13:11 |
|
paul |
'morning to you. |
13:11 |
|
paul |
my patches have been lost, or is there a problem with them |
13:12 |
|
paul |
(it was a patches.tgz, very important, as they contains the updated french translation) |
13:12 |
|
paul |
(and I have another update one to send... for both opac & staff) |
13:13 |
|
gmcharlt |
good morning #koha, paul, and kados |
13:13 |
|
paul |
'morning gmcharlt |
13:14 |
|
gmcharlt |
paul: when you have a chance, please look at the changes I made to rebuild_zebra.pl -- I'm wonder if the code to work with the config files is still needed |
13:17 |
|
paul |
gmcharlt: you have removed them already if I don't mind ? |
13:17 |
|
paul |
s/them/it/ |
13:18 |
|
gmcharlt |
gmcharlt: no, code is still there, just now need to supply --munge-config to invoke it |
13:18 |
|
gmcharlt |
paul, rather (need to stop talking to myself :) ) |
13:18 |
|
gmcharlt |
but if not needed, I can remove outright |
13:20 |
|
gmcharlt |
hehe |
13:20 |
|
paul |
lol |
13:24 |
|
kados |
paul: hi |
13:24 |
|
kados |
paul: did you get my email ... I had some problems with the patches, can you rebase and re-submit as a tgz? |
13:25 |
|
paul |
ok, I do it immediatly (and, unless I missed it, I haven't got a mail from you) |
13:26 |
|
paul |
kados : I send the mail to jmf or patches mailbox ? |
13:26 |
|
kados |
send to rmkoha.org this time |
13:27 |
|
kados |
but normally, patches@ |
13:27 |
|
kados |
:-0 |
13:27 |
|
paul |
done |
13:27 |
|
kados |
paul: also, when it's conveneint, I have a question |
13:29 |
|
kados |
paul: do any of the french libraries use the 'accounts, fines' features in Koha? |
13:30 |
|
paul |
(on phone) |
13:31 |
|
paul |
kados : I don't think so |
13:31 |
|
hdl |
kados: back |
13:31 |
|
paul |
mmm... no, i'm wrong. sanop use them if I don't mind |
13:31 |
|
hdl |
maybe some local libraries. |
13:32 |
|
kados |
in france, how many decimal place for the currency (euro)? |
13:32 |
|
kados |
just two like in the US? |
13:32 |
|
paul |
2 |
13:35 |
|
kados |
ok, thanks |
13:36 |
|
kados |
paul: have any plans for the holiday? |
13:39 |
|
arsenic |
guys, do you still using Bugzilla to report bugs? |
13:39 |
|
kados |
arsenic: bugs.koha.org |
13:39 |
|
kados |
arsenic: yep |
13:40 |
|
arsenic |
are you using rel_3_0 for, well, Koha 3.0 bugs? |
13:40 |
|
kados |
HEAD until the release |
13:40 |
|
kados |
but rel_3_0 will work |
13:40 |
|
kados |
paul: adding _() to strings for translator |
13:40 |
|
arsenic |
ok. thanks kados |
13:40 |
|
kados |
paul: does this patch improve things, i'm unsure based on your commit message |
13:41 |
|
kados |
paul: 'Note that it DOES not work as expected. It seem that the translator misses the _() string inside expandable button' |
13:41 |
|
paul |
kados : 2 commit later, I say that in fact it works fine |
13:41 |
|
kados |
hehe, ok |
13:44 |
|
kados |
paul: Civility is incorrect, Title is preferred |
13:44 |
|
paul |
it's an unsolvable problem for translation ! |
13:44 |
|
paul |
title in french is "Titre" (=biblio title) and "Civilité" (=ppl title) |
13:45 |
|
paul |
so we need 2 different terms. |
13:45 |
|
kados |
I see |
13:45 |
|
kados |
hmmm |
13:45 |
|
paul |
(we can use "titre" for civility, but only for king, duke or things like that) |
13:46 |
|
paul |
(not for Mister, Miss...) |
13:48 |
|
kados |
we could use Salutation |
13:49 |
|
arsenic |
that i see almost everywhere |
13:49 |
|
kados |
hmmm, not quite right |
13:49 |
|
kados |
that would be more like the first line in a letter 'Dear X' |
13:50 |
|
kados |
paul: I'll approve the patch and will think of a better term in the meantime |
13:50 |
|
arsenic |
Well, it's actually Dear [M./Miss/Dumb] X, so that is why salutation is used |
13:55 |
|
kados |
paul: all patches pushed |
13:55 |
|
kados |
arsenic: ok, maybe that is the right term |
13:55 |
|
arsenic |
here an awesome example: http://inlibro.com/index.php?o[…]leforms&Itemid=60 :) |
13:57 |
|
arsenic |
but it seems that Civilité is used on many french (from France) web site... and Salutation in french (from Quebec/Canada)'s one |
13:57 |
|
kados |
ok, salutation it is |
13:58 |
|
arsenic |
We do have multiple inconsistencies between our french, so I guess that a day or another, I'll have to sponsored a French (Canada) translation. |
14:02 |
|
arsenic |
The koha database must be created with which Collation? I used utf8_general_ci, is that correct? |
14:02 |
|
paul |
that's what I use |
14:07 |
|
arsenic |
There is something I find strange in Koha 3. When creating a new MARC record, the ISBN field lets me enter only 2 characters. Do i miss something? |
14:08 |
|
kados |
arsenic: sounds like a bug |
14:11 |
|
arsenic |
Well, that what I though at first. Then I wait a while, but since a few weeks, nobody complaints. Is that because most people are not using the z3950 feature to import marc record? |
14:11 |
|
kados |
arsenic: I'd say it's because no-one noticed or tested it |
14:11 |
|
owen |
I've tested the z3950 search more recently than weeks ago |
14:11 |
|
kados |
oh ... I stand corrected |
14:11 |
|
kados |
hey owen |
14:12 |
|
arsenic |
owen: did you use the ISBN field? |
14:12 |
|
owen |
Yes, no problem. |
14:13 |
|
owen |
Could it be a framework issue? |
14:13 |
|
arsenic |
Is there a field length setting set in the framework? |
14:15 |
|
arsenic |
I'm using the sample data provided with the installation. I looked at paul's demo and the problem is there too. |
14:16 |
|
kados |
arsenic: unimarc or marc21? |
14:17 |
|
arsenic |
marc21 |
14:17 |
|
arsenic |
you see how weirdo we are in quebec... we speak french but using marc21... :) |
14:20 |
|
kados |
:) |
14:20 |
|
arsenic |
kados: do you have the problem? |
14:21 |
|
owen |
arsenic: Do you know why it only allows 2 characters? Is there a "maxlength" attribute attached to the input tag? |
14:22 |
|
arsenic |
that what I looked at but there is no lenght attrib... looking at the .tmpl file now. |
14:22 |
|
hdl |
arsenic : two characters are for indicators. |
14:22 |
|
hdl |
not for data. |
14:23 |
|
hdl |
click on tag description to input data. |
14:24 |
|
arsenic |
hdl: that must be it... |
14:25 |
|
arsenic |
hdl: that's it, thanks. I knew i was missing something... |
14:26 |
|
hdl |
don't mention it |
14:26 |
|
arsenic |
Shouldn't we put an icon beside the field to add data? it's really not obvious that we have to click on the description... |
14:33 |
|
owen |
It might be confusing to a brand new user, but it's easily remembered once someone understands it |
14:33 |
|
owen |
I wonder if the alternative would clutter the interface in a way that outweighs its usefulness |
14:34 |
|
arsenic |
of course that once you know it, it's something easy to remembre. Maybe we could just put a not in the top of the page instead |
14:35 |
|
arsenic |
... a note ... |
14:50 |
|
fbcit |
g'morning koha |
14:50 |
|
kados |
morning fbcit |
14:50 |
|
fbcit |
owen arsenic: that got me the first time too... |
14:50 |
|
fbcit |
hi kados |
14:51 |
|
arsenic |
fbcit: hi, and thanks for making me feel better ;) |
15:02 |
|
kados |
arsenic, fbcit so what's this? |
15:02 |
|
kados |
arsenic: icon beside what field? |
15:04 |
|
owen |
The question is should there be some kind of indication that you have to click the tag text in order to expand the entry form in addbiblio |
15:08 |
|
fbcit |
I believe that expanded/contracted option is controlled in the framework setup, if I remember right. |
15:14 |
|
fbcit |
gmcharlt: fix-perl-path.PL worked great with your changes. |
15:21 |
|
gmcharlt |
fbcit: good to hear |
15:21 |
|
gmcharlt |
fbcit: so what dependent Perl modules are left? |
15:21 |
|
fbcit |
ZOOM which depends on YAZ |
15:23 |
|
fbcit |
most of the XML modules |
15:23 |
|
fbcit |
Schedule::At |
15:24 |
|
fbcit |
which will require some patching up to work either with Win32 'at' or task scheduler |
15:24 |
|
fbcit |
the Win32 port of YAZ does not provide the libs that ZOOM wants |
15:24 |
|
fbcit |
the ZOOM module author had not additional insight |
15:24 |
|
gmcharlt |
fbcit: nice thing about Schedule::At is it's only used by one script so far; so if there's a portable scheduling module on CPAN, would be easy to switch over |
15:25 |
|
fbcit |
so I'm going to try compiling YAZ under cygwin and moving the libs over |
15:25 |
|
fbcit |
good idea on S::At |
15:26 |
|
fbcit |
I was working on YAZ/ZOOM last night some |
15:27 |
|
fbcit |
some of this stuff will probably require us maintaining a separate package of Win32 ports |
15:27 |
|
fbcit |
just to keep things in the realm of sanity |
15:28 |
|
gmcharlt |
fbcit: yep, although I think the indexdata folk would be amenable to taking suggestions re the YAZ package if we make a good case |
15:31 |
|
fbcit |
gmcharlt: I also posted to the Net::3950::ZOOM list per the author's suggestion, but received no response to date. |
15:31 |
|
fbcit |
the current Koha Win32 port does not use ZOOM, correct? |
15:32 |
|
gmcharlt |
fbcit: dunno, sorry |
15:32 |
|
fbcit |
kados ? |
15:36 |
|
fbcit |
hdl around? |
15:36 |
|
hdl |
yes |
15:37 |
|
fbcit |
hdl: I found the NoZebra issue |
15:37 |
|
hdl |
what was it ? |
15:37 |
|
fbcit |
in buildQuery, when there was no $query and only $limit |
15:38 |
|
fbcit |
$query was set to $limit with a white space prefixed to it |
15:38 |
|
arsenic |
is it possible to get a report of all checked out books? |
15:39 |
|
fbcit |
hdl: one moment |
15:40 |
|
atz |
fbcit: I thought Koha used ZOOM on all platforms |
15:40 |
|
gmcharlt |
atz: including 2.2? |
15:41 |
|
gmcharlt |
atz: 2.2 for Win32, I mean |
15:41 |
|
fbcit |
hdl: because of this and a later regex, on the second pass through NZanalyse something to this effect was produced: |
15:41 |
|
fbcit |
EXECUTE : biblioserver, , NF at /usr/share/kohaclone1/C4/Search.pm line 1395. |
15:41 |
|
atz |
ah, i see... yeah, I'm thinking 3.0 |
15:41 |
|
atz |
don't know about earlier stuff |
15:42 |
|
fbcit |
hdl: so I added a conditional to buildQuery based on the existence of $query to determine whether or not to prefix the white space and |
15:42 |
|
fbcit |
corrected the regex in NZanalyse |
15:42 |
|
hdl |
atz: Koha now use ZOOM for any z3950 research. |
15:42 |
|
atz |
right |
15:43 |
|
fbcit |
hdl: see http://git.koha.org/cgi-bin/gi[…]9e112206a4af0f8ec |
15:43 |
|
fbcit |
although it accidentally deletes a line which kados added back for me later last evening |
15:45 |
|
hdl |
+ $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\///g; # we must not introduce spaces in place of these chars |
15:45 |
|
hdl |
I donot agree. |
15:45 |
|
fbcit |
? |
15:45 |
|
fbcit |
brb |
15:45 |
|
hdl |
take this : Benjamin franklin un américain à paris (1776-1785) |
15:46 |
|
hdl |
if you simply remove - and donot replace - by space, you will query 17761785 and will surely get NO results. |
15:46 |
|
fbcit |
ahh.. ok |
15:47 |
|
hdl |
This is a big bug. |
15:47 |
|
fbcit |
the main issue is with the parenthesis |
15:47 |
|
fbcit |
being changed to spaces |
15:47 |
|
hdl |
which parenthesis ? |
15:47 |
|
hdl |
In operands ? |
15:48 |
|
fbcit |
yes |
15:48 |
|
hdl |
Again, Benjamin franklin un américain à paris (1776-1785) |
15:48 |
|
hdl |
you have to search for Benjamin franklin un américain à paris 1776 1785 |
15:49 |
|
fbcit |
in the case where only a limit is passed in, no keywords, etc |
15:49 |
|
fbcit |
the limit becomes the search |
15:49 |
|
hdl |
yes. |
15:49 |
|
hdl |
and it is catched in : |
15:49 |
|
fbcit |
when the parens are replaced with spaces |
15:49 |
|
fbcit |
the query is parsed incorrectly |
15:50 |
|
hdl |
line 1209 |
15:50 |
|
hdl |
then parsed. |
15:51 |
|
hdl |
The files i sent you yesterday works : |
15:51 |
|
hdl |
See o10.hdlaurent.paulpoulain.com |
15:53 |
|
hdl |
So I donot think you patch will help. It rather will break things. |
15:53 |
|
hdl |
Hopefully I am wrong. |
15:55 |
|
fbcit |
hdl: |
15:55 |
|
fbcit |
Pas de réponse ! |
15:55 |
|
fbcit |
Aucun résultat ne correspond à votre recherche “� dans le catalogue de Musée du CNAM |
15:56 |
|
hdl |
what have you done precisely ? |
15:56 |
|
fbcit |
wait |
15:56 |
|
fbcit |
it appears that you have no itemtype DVD? |
15:56 |
|
hdl |
Yes. |
15:57 |
|
fbcit |
selecting itemtype of books works |
15:57 |
|
hdl |
No items DVD now |
15:57 |
|
fbcit |
but I still cannot get your Search.pm to work here :( |
15:57 |
|
fbcit |
[Fri Dec 21 11:46:11 2007] search.pl: itemtype / = / NF |
15:57 |
|
fbcit |
[Fri Dec 21 11:46:11 2007] search.pl: EXECUTE : biblioserver, itemtype, NF at /usr/share/kohaclone1/C4/Search.pm line 1337. |
15:57 |
|
fbcit |
[Fri Dec 21 11:46:11 2007] search.pl: mc / = / NF |
15:57 |
|
fbcit |
[Fri Dec 21 11:46:11 2007] search.pl: EXECUTE : biblioserver, mc, NF at /usr/share/kohaclone1/C4/Search.pm line 1337. |
15:57 |
|
fbcit |
fails on the second recursion |
15:58 |
|
hdl |
you have no index mc have you ? |
15:58 |
|
fbcit |
no |
15:58 |
|
hdl |
this is the reason. |
15:59 |
|
hdl |
mc or itemtype has been alternatively used by Liblime for itemtypes research. |
15:59 |
|
fbcit |
so should this not be installed by default? |
16:00 |
|
hdl |
by which default ? |
16:00 |
|
fbcit |
rather where is this index defined? |
16:00 |
|
hdl |
maybe but you can do it : |
16:01 |
|
hdl |
NoZebraIdexes |
16:01 |
|
hdl |
NoZebraIndexes in systempreferences |
16:01 |
|
hdl |
copy itemtype line and add it replaceing itemtype by mc. |
16:02 |
|
hdl |
then rebuild_nozebra.pl |
16:04 |
|
hdl |
btw : on the version you tested (which is 1 week old), I only had one index used for itemtypes : mt. and the search gave : |
16:06 |
|
hdl |
QUERY: mt=MON at /home/hdl/Koha/Code/installations/CNAM3//C4/Search.pm |
16:06 |
|
hdl |
QUERY CGI: at /home/hdl/Koha/Code/installations/CNAM3//C4/Search.pm line 912. |
16:06 |
|
hdl |
QUERY DESC: at /home/hdl/Koha/Code/installations/CNAM3//C4/Search.pm line 913. |
16:06 |
|
hdl |
LIMIT:mt=MON at /home/hdl/Koha/Code/installations/CNAM3//C4/Search.pm line 914. |
16:06 |
|
hdl |
LIMIT CGI:&limit=mt:MON at /home/hdl/Koha/Code/installations/CNAM3//C4/Search.pm line 915. |
16:06 |
|
hdl |
LIMIT DESC:and mt:MON at /home/hdl/Koha/Code/installations/CNAM3//C4/Search.pm line 916. |
16:06 |
|
hdl |
string : mt=MON at /home/hdl/Koha/Code/installations/CNAM3//C4/Search.pm line 1254. |
16:10 |
|
fbcit |
hdl: works after rebuilding |
16:11 |
|
fbcit |
we really need to patch syspref.sql to include mc along with your other patches, though. |
16:12 |
|
fbcit |
hdl: what did the EXECUTE look like for the above debug? |
16:16 |
|
Brooke |
howdy |
16:17 |
|
hdl |
Hi Brooke |
16:18 |
|
Brooke |
bonjour henri :) |
16:21 |
|
arsenic |
Concerning reports created from SQL queries, is there any limitation there (eg column names)? |
16:24 |
|
kados |
arsenic: from the report writer you mean? if so, the answer is no |
16:25 |
|
arsenic |
kados: I'm trying to create a simple query like SELECT * FROM issues WHERE returndate is NULL; |
16:26 |
|
kados |
arsenic: use the report writer |
16:27 |
|
arsenic |
kados: what is the report writer exactly? |
16:27 |
|
kados |
arsenic: reports -> guided report writer -> circulation |
16:27 |
|
kados |
arsenic: if you want to do the issues query |
16:28 |
|
kados |
arsenic: or you could also type in the SQL directly |
16:28 |
|
kados |
arsenic: there's extensive help files written in there, take a look at the [?] top right of the screen |
16:48 |
|
arsenic |
ok... i guess that (again) i'm may be missing something... (btw, thanks for pointing me the contextual help :) |
16:50 |
|
arsenic |
Ok. I'm building a new report. Step 1: Circulation, Step 2: Tabular, Step 3: ??? what should i use to get only the current issues...?! |
16:51 |
|
arsenic |
let's say I'd like to have the barcode and title of the currently issued items. |
16:59 |
|
paul |
time to leave for me. I'll be back after christmas. |
16:59 |
|
paul |
merry christmas to everyone here. |
16:59 |
|
owen |
Merry Christmas, paul! |
16:59 |
|
arsenic |
Bon temps des fêtes Paul. Repose toi bien |
17:00 |
|
Brooke |
Merry Christmas paul |
17:01 |
|
Brooke |
Joyeux Noël |
17:01 |
|
atz |
i trust you'll be eating well and enjoying the holiday, paul! |
17:11 |
|
hdl |
kados: about guided report, is it forecast to allow people to use Distinct for lists ? |
17:11 |
|
hdl |
is there any difference at the moment between tabular matrix and summary ? |
17:12 |
|
hdl |
will there be ? |
17:12 |
|
kados |
hdl: there will be in a future version |
17:12 |
|
kados |
hdl: no difference currently |
17:18 |
|
arsenic |
Is it possible to set holidays for all branches at once? |
17:20 |
|
kados |
arsenic: I think there used to be a 'default' branch ... is that gone now? |
17:20 |
|
arsenic |
kados: i guess so.. let me double check |
17:20 |
|
arsenic |
kados: yes, it's gone. |
17:21 |
|
kados |
arsenic: can you file a bug please? |
17:21 |
|
arsenic |
sure |
17:26 |
|
arsenic |
other than paul's one, is there any staff client demo for koha 3 availables? |
17:31 |
|
kados |
nothing official yet |
17:31 |
|
kados |
liblime will have one shortly after the release |
17:34 |
|
arsenic |
ok. It would be nice to have one created from every days latest version... |
17:35 |
|
kados |
I agree |
17:38 |
|
arsenic |
do you also encounter the following search problem. Launch a search in the advance search page. Use the back button of your browser. Press the OK button to re-launch the search. Nothing happens (but sadness) |
17:41 |
|
kados |
arsenic: staff client? or opac? |
17:41 |
|
kados |
arsenic: zebra or nozebra? |
17:41 |
|
arsenic |
opac, nz |
17:41 |
|
arsenic |
But i remembre of seeing it on paul's opac too |
17:43 |
|
kados |
arsenic: I don't have that problem |
17:46 |
|
hdl |
Merry Christmas every body. |
17:47 |
|
hdl |
See you. in 10 days. |
17:47 |
|
kados |
ciao hdl |
17:47 |
|
arsenic |
ciao hdl |
17:48 |
|
arsenic |
kados: hmm... are you using FF? |
17:48 |
|
kados |
yes |
17:50 |
|
fbcit |
Merry Christmas hdl |
17:51 |
|
arsenic |
kados: I'm experiencing the problem on http://zoomopac.liblime.com/search |
17:52 |
|
kados |
arsenic: sounds like a browser problem then |
17:58 |
|
arsenic |
kados: you are right. It seems to be related to FF on Vista...?! |
17:59 |
|
kados |
weird |
18:44 |
|
arsenic |
i remember that in 2.2, there was a problem with the circulation rules matrix. We had to fill every cells to get the desired result. It seems that it still the case with 3.0. Anybody can confirm? |
18:45 |
|
arsenic |
If I set 25,5 in the Total/Default cell, i can still loan a 6th book to a borrower without being notified. |
19:01 |
|
[K] |
*** join #kohaFreeNode: maktrix n=mak59.152.90.138 |
19:01 |
|
kados |
arsenic: really! |
19:01 |
|
kados |
arsenic: can you file a bug for that? |
19:06 |
|
arsenic |
kados: yep |
19:07 |
|
kados |
thanks |
19:30 |
|
fbcit |
their dropping like files... |
19:30 |
|
fbcit |
s/flies/flys/g |
19:32 |
|
kados |
hehe |
19:32 |
|
arsenic |
In which files the warn"xxxxx" from the .pl file are supposed to be logged? |
19:32 |
|
kados |
apache log files |
19:33 |
|
fbcit |
*grimace* |
19:33 |
|
arsenic |
hehe |
19:33 |
|
arsenic |
is there a setting that must be set to see them? |
19:34 |
|
kados |
no, just check your koha-httpd.conf file for where they are located |
19:34 |
|
fbcit |
$DEBUG |
19:35 |
|
kados |
oh |
19:35 |
|
kados |
right |
19:35 |
|
fbcit |
if you want to be overrun... |
19:36 |
|
kados |
that's just for the warns in C4/Search.pm though, right? |
19:36 |
|
fbcit |
there are several $DEBUG's floating around various files that I have seen |
19:36 |
|
fbcit |
though not all warn's are conditioned by it |
19:37 |
|
fbcit |
some files look at ENV $DEBUG |
19:38 |
|
kados |
yea, someone changed that and I've no idea how it's ever set |
19:38 |
|
kados |
atz: was that you? |
19:39 |
|
kados |
atz: how does one turn on $DEBUG now? |
19:39 |
|
atz |
that is my style, yes... but some DEBUG's pre-existed |
19:39 |
|
kados |
atz: a+ for style, but how does it work? :-) |
19:39 |
|
atz |
i want to write a front-to-back style guide for debug |
19:39 |
|
fbcit |
hehe |
19:40 |
|
kados |
heh |
19:40 |
|
fbcit |
maybe we can do $DEBUG_atz $DEBUG_fbcit, etc |
19:41 |
|
kados |
hehe |
19:41 |
|
atz |
mine are tripped by non-zero $ENV{DEBUG}, set in the BEGIN{} block to a local variable $debug |
19:42 |
|
atz |
separately, some scripts provide feedback on CGI param('debug') |
19:42 |
|
atz |
note: ENV should never be SET by CGI param |
19:43 |
|
atz |
there appears to be some hard coded debug stuff |
19:43 |
|
atz |
and separately, a user syspref (??) |
19:43 |
|
fbcit |
least intrusive |
19:43 |
|
kados |
works for me |
19:44 |
|
atz |
actually, there are valid reasons to use the others' styles too :\ |
19:44 |
|
kados |
there's a syspref that sets the debug level for Carp |
19:44 |
|
atz |
my style reflects that most of my work is at the command line |
19:44 |
|
atz |
owen might prefer the CGI param, for example |
19:44 |
|
kados |
*nod* |
19:44 |
|
atz |
the one I know least about the syspref |
19:45 |
|
fbcit |
module DEBUG..... |
19:45 |
|
atz |
i have not seen that used anywhere |
19:45 |
|
kados |
afaik it's only used for the carp stuff |
19:45 |
|
atz |
ah... ok, not much then |
19:46 |
|
kados |
would it make sense to appropriate that for debug level to trigger your BEGIN block? |
19:46 |
|
kados |
I'm all for ENV, but that's tricky to enable for testing when using the web interaces |
19:46 |
|
atz |
yeah, these are really separate ideas |
19:47 |
|
atz |
most of my module level stuff outputs to STDERR or warn |
19:48 |
|
atz |
meaning if you testing via web, the messages end up in the log |
19:49 |
|
atz |
so you don't want to enable the module level debug from a CGI param (though you might want the reverse) |
19:49 |
|
kados |
doesn't warn send output to stderr? |
19:49 |
|
atz |
yeah, same destination, different formatting |
19:50 |
|
kados |
I guess from my POV, it'd be darned useful to be able to trigger a module level debug from a cgi param |
19:50 |
|
atz |
part of the value of using ENV is that it has no dependencies (CGI or Context) |
19:51 |
|
atz |
kados: yeah, but not as a permanent feature of the system, where any user could start dropping 50 lines per request into the error log |
19:52 |
|
atz |
i would like to see more CGI-responsiveness to a debug param |
19:53 |
|
kados |
so what about : syspref turns on debug globally, including the ability to add cgi debug params |
19:53 |
|
atz |
but again, i think of this as related but distinct from the lower level debugging |
19:55 |
|
atz |
i'm not sure. |
19:56 |
|
gmcharlt |
kados, atz: from my point of view, need ability to turn on debugging (i.e., messages to Apache error log) without needing APache reload |
19:56 |
|
atz |
it would possible for that to work smoothly |
19:57 |
|
gmcharlt |
so perhaps something like $debug enabled if either env var is present or a syspref is defined |
19:57 |
|
atz |
gmcharlt: send ?debug=1 to memberentry.pl and see what happens :) |
19:57 |
|
gmcharlt |
atz: not useful in customer support situation |
19:58 |
|
atz |
how so/? |
19:58 |
|
gmcharlt |
i.e., shoudl be able to tell user to turn on syspref, then send customer support log |
19:58 |
|
gmcharlt |
if tell user to start munging URLs, more difficult, especically if trying to trace a multi-page transaction |
19:59 |
|
atz |
that makes sense |
20:01 |
|
kados |
yea, I like the syspref method |
20:01 |
|
atz |
some of my debug stuff has levels... like: ($debug>3) and warn "some big complex dump" |
20:01 |
|
gmcharlt |
so how about this: |
20:02 |
|
gmcharlt |
check for DEBUG env var and syspref is done ones, in C4::Context |
20:02 |
|
atz |
this would be useful to maintain |
20:02 |
|
fbcit |
atz: I have used that elsewhere too... works nicely |
20:02 |
|
atz |
any problem w/ a 0-9 convention ? |
20:02 |
|
gmcharlt |
other modules take debug setting from (say) C4::Context->debug(), or perhaps we even export a koha_debug_level() or something |
20:03 |
|
gmcharlt |
and per atz: depending on programmer, that passed $debug value can have levels |
20:03 |
|
gmcharlt |
0 = no debug info produced |
20:03 |
|
gmcharlt |
1-9 levels |
20:03 |
|
gmcharlt |
and if a particular module doesn't need to worry about levels of reporting, $debug and warn("xxx") still works just fine |
20:04 |
|
atz |
right, that's the idea |
20:04 |
|
fbcit |
we could add a page to ship the log to support as well |
20:04 |
|
kados |
yea, that's ideal |
20:04 |
|
gmcharlt |
and perhaps instead of doing a $debug and warn all the time, may just have C4::Context export a koha_carp function |
20:05 |
|
gmcharlt |
so that you would just do koha_debug("xxx"); |
20:05 |
|
gmcharlt |
(yes, I know that I just changed sub names) |
20:05 |
|
kados |
that would be very cool |
20:05 |
|
kados |
(:)) |
20:05 |
|
fbcit |
module DEBUG... |
20:05 |
|
atz |
gmcharlt: not sure I'd want to go that far |
20:05 |
|
arsenic |
I would go for a koha_trace("xxx"); |
20:05 |
|
kados |
fbcit: yea, we could use a debug module |
20:05 |
|
kados |
this discussion should be moved to koha-devel |
20:06 |
|
kados |
any volunteers for spearheading that? |
20:06 |
|
gmcharlt |
could allow several names -- koha_trace, koha_debug, etc., to express different levels |
20:06 |
|
arsenic |
with eventually a level param. koha_trace(3, "my level 3 warning"); |
20:07 |
|
gmcharlt |
atz: what would be too far? |
20:07 |
|
atz |
gmcharlt: so we'd be making a bunch of calls to koha_trace without knowing whether or not they will be used, because the debug level would be encapsulated? |
20:08 |
|
gmcharlt |
atz: yeah, but there would always be a runtime evalulation, either at the begging on koha_trace |
20:08 |
|
gmcharlt |
or in the context of a $debug and warn contruct |
20:08 |
|
gmcharlt |
so question is whether sub invocation is too expensive |
20:09 |
|
atz |
the latter is preferable only .... |
20:09 |
|
gmcharlt |
not sure that it would be, unless people start putting in a lot of these |
20:09 |
|
atz |
because the dump messages can be pretty big |
20:09 |
|
atz |
and shouldn't bother to be assembled unless they will be used |
20:09 |
|
gmcharlt |
atz: ah, I see, evaulation of the contents of the debug message |
20:10 |
|
atz |
right, assembling the message itself should be avoided in normal operation |
20:15 |
|
gmcharlt |
atz: we could always use a source filter to deal with this ;-) |
20:16 |
|
gmcharlt |
but seriously, it looks like it would have to be $debug and koha_trace("xxx") [or koha_trace("xxx") if $debug] |
20:16 |
|
gmcharlt |
so atz, will you raise this on koha-devel? |
20:20 |
|
atz |
the tactical side of me says write the style guide first :) |
20:21 |
|
kados |
fair enough, if you want to, write a style guide as a proposal, and release it as an RFC to koha-devel |
20:21 |
|
kados |
I'd use the koha wiki |
20:21 |
|
kados |
but my pref would be to keep it brief for now, in case someone objects ... ie, don't spend a ton of time on it :-) |
20:22 |
|
atz |
right |
20:22 |
|
kados |
'release early, often' |
20:22 |
|
atz |
any of this overhaul is after the release, obviously |
20:23 |
|
atz |
but we should have a stronger convention on debugging |
20:23 |
|
gmcharlt |
atz: we should have *a* convention on debugging ;-) |
20:24 |
|
atz |
well put |
20:25 |
|
[K] |
*** part FreeNode!#koha: maktrix n=mak59.152.90.138 |
21:34 |
|
arsenic |
fcbit, did you mention that just by setting the $DEBUG, i should get the warn message in my apache log file? (yeah, I still trying to figure out this) |
23:05 |
|
arsenic |
hey! I remember seeing a thread talking about if it was ok or not to be able to get the patrons' reading history in the staff client. What was the final conclusion about this? Is this an option? |
23:05 |
|
arsenic |
it was a question about the personal information protection, brough by paul if my memory serves me well |
23:07 |
|
kados |
it is an option, you can enable it in sysprefs |
23:09 |
|
arsenic |
I guess is either ReadingHistory or intranetreadinghistory. There are both set to true and I can see anything related to reading history in the patron's page... |
23:37 |
|
arsenic |
ok, it seems that the reading history is because that the intranetreadinghistory flag never goes through the template params... |
23:38 |
|
kados |
arsenic: sounds like another bug in 3.0 |
23:38 |
|
kados |
arsenic: works fine in 2.2 and dev_week |
23:38 |
|
arsenic |
why is there some hardcoded category_type? C and A seems to be reserved for children and adult... |
23:39 |
|
kados |
yes, that is a new 'feature' of 3.0 |
23:39 |
|
kados |
you have types of categories |
23:39 |
|
arsenic |
oh, is this the super categories? |
23:39 |
|
kados |
the types are hardcoded, the categories are not |
23:39 |
|
kados |
yes |
23:39 |
|
arsenic |
ok. I'll fill a bug for the reading history stuff. |
23:40 |
|
kados |
arsenic++ thanks |
03:28 |
|
fbcit |
kados: around? |
04:03 |
|
arsenic |
fcbit: still working on a friday night?! :) |