Time |
S |
Nick |
Message |
12:03 |
|
hdl |
thd : is z3950-daemon-options all right ? |
12:07 |
|
thd |
hdl: looks fine to me but I have no working copy for comparison |
12:20 |
|
hdl |
thd :here is mine |
12:20 |
|
hdl |
RunAsUser=hdl |
12:20 |
|
hdl |
KohaZ3950Dir=/home/hdl/Koha/koha-dev/koha/z3950 |
12:20 |
|
hdl |
KohaModuleDir=/home/hdl/Koha/koha-dev/koha/C4 |
12:20 |
|
hdl |
LogDir=/var/log/koha |
12:20 |
|
hdl |
KohaConf=/etc/koha/koha-head.conf |
12:22 |
|
hdl |
thd : are you trying some z3950 tests with slef ? |
12:25 |
|
thd |
hdl: no slef appeared, said good morning, then said broken, and then seemed absent |
12:29 |
|
thd |
hdl: I have a task switching problem and need to restart x-windows |
12:37 |
|
thd |
hdl: Is there a known set of diagnostic tests for the Koha z3950 that I might be checking? |
12:53 |
|
thd |
hdl: My z3950deamon options are the default options and look good to me |
12:53 |
|
thd |
RunAsUser=www |
12:53 |
|
thd |
KohaZ3950Dir=/usr/local/koha/intranet/scripts/z3950daemon |
12:53 |
|
thd |
KohaModuleDir=/usr/local/koha/intranet/modules |
12:53 |
|
thd |
LogDir=/usr/local/koha/log |
12:53 |
|
thd |
KohaConf=/etc/koha.conf |
12:55 |
|
thd |
hdl: unless you have a clue otherwise, I suspect the daemon process itself has a problem |
12:56 |
|
hdl |
try processz3950queue |
12:57 |
|
thd |
hdl: try what with processz3950queue? |
12:57 |
|
hdl |
thd : sorry : was away. |
12:57 |
|
hdl |
Try to launch processz3950queue in a console. |
12:57 |
|
hdl |
|
12:57 |
|
hdl |
Beware . |
12:58 |
|
hdl |
You must export KOHA_CONF and PERL5LIB. |
12:58 |
|
hdl |
And then you launch processz3950queue |
12:59 |
|
hdl |
But you must have a good z3950 perl install. |
12:59 |
|
hdl |
You could also search in this direction. |
13:01 |
|
hdl |
Pls Forgive me. |
13:01 |
|
hdl |
I go to lunch |
13:03 |
|
thd |
hdl: I have been told that eating is important :) |
13:16 |
|
sylvain |
I've a perl problem, I've 10 files to bulkmarcimport so I've written a text file with the 10 calls to bulkmarck import and then I ./myBat but I get a problem. When I type the call to bulkmarcimport by hand, it works correctly but when I use my "bat", the char_decode function only return what is given in entry without transcoding (unnimarc) |
13:28 |
|
sylvain |
shit ! in fact my line was ..... -c UNIMARC and the EOL was added to my argument :( |
13:31 |
|
sylvain |
I should haven't edited my file under windows before using it on linux |
13:31 |
|
sylvain |
stupid boy !!!! |
13:54 |
|
thd |
hdl: for your return console produced something for unfinished z3950 searches but no joy yet. |
13:55 |
|
thd |
hdl: I will return later |
15:18 |
|
slef |
hello |
15:27 |
|
rach |
hello |
15:34 |
|
slef |
mmm dinner |
16:55 |
|
thd_ |
slef: are you back from dinner? |
17:48 |
|
slef |
thd: yes |
18:27 |
|
thd |
slef: Did you mean to say ealier that koha z39.50 services are currently broken? |
20:03 |
|
slef |
I mean that your z39.50 daemon looks broken. |
20:05 |
|
thd |
slef: how can I fix it? |
01:58 |
|
hdl |
hi |
02:07 |
|
osmoze |
hello |
02:15 |
|
sylvain |
hi all |
02:16 |
|
sylvain |
is it normal that if I bulkmarcimport files one after one, even if all the files are the same size (10 000 records) it goes more and more slowly |
02:16 |
|
sylvain |
? |
02:39 |
|
hdl |
If you import things into the same database, it is logical : The DB gets bigger and bigger ;) |
02:39 |
|
hdl |
=> more and more mem used ;) |
02:44 |
|
sylvain |
in french sorry : Oui mais je pensais que mysql ne perdait pas tant que ça dés que la base augmentait ... Là c'est presque du simple au double pour 70 000 notics |
03:35 |
|
hdl |
Can you give some figures of computing time ? |
03:41 |
|
sylvain |
yes, just 2 seconds |
03:41 |
|
indradg |
kados, around? |
03:43 |
|
sylvain |
7 mo --> 35 minutes8 mo --> 55 minutes8 mo --> 1h 10 minutes9 mo --> 1h 20 minutes7 mo --> 1h 20 minutes8 mo --> 1h 25 minutes |
03:43 |
|
sylvain |
these are the figures hdl size --> time |
03:53 |
|
hdl |
Sylvain : thx. Increase seems to be logical since any addition must search database for doubles. But maybe you could ask on koha-fr for some performance on bulkmarcimport to compare with. Think about asking for configuration and amount of biblio imported. |
04:49 |
|
sylvain |
cette marc_word_table est vraiment chiante dès qu'il y a des gros volumes de données à gérer/charger |
04:52 |
|
hdl |
C'est un peu pour cela que l'on étudie le passage à autre chose... Zebra. |
04:52 |
|
hdl |
;) |
04:54 |
|
sylvain |
oui je sais :) mais j'avais tjs travaillé avec des relativement petites bases mais aujourd'hui, je souffre :) |
05:20 |
|
hdl |
sylvain : Paul avait fait un POC de table inverse pour avoir une efficacité accrue au niveau des recherches. Tu peux essayer de tester... Si le coeur et le temps te le permettent. |
05:24 |
|
hdl |
POC : Proof Of Concept... |
05:25 |
|
hdl |
En tout cas, lui, il avait eu un gain en perf assez impressionnant.... Après l'indexation des éléments qui prenait un certain temps toutefois... Et je ne parle pas de l'occupation mémoire de la base ;) |
05:35 |
|
sylvain |
oui hdl j'avais vu sa poc, c'était en effet assez impressionant |
05:35 |
|
sylvain |
mais pas trop le temps en ce moment de m'y pencher :( |
05:38 |
|
hdl |
C'était juste une suggestion;). dtf, puisqu'on se tourne vers Zebra, c'est maintenant plus histoire de dire que paul avait déjà bien pensé et même presque résolu le problème qu'autre chose. |
07:15 |
|
slef |
kados: when will 2.3.0 be? |
07:18 |
|
slef |
paul_away: can you buzz me when you get back, please? I have auto_install questions. |
08:00 |
|
thd |
hdl: I can run processz3950queue in console mode and get the previous incomplete lookup |
08:08 |
|
hdl |
thd : yes... do you get answers ? |
08:12 |
|
thd |
17403/10 : Processing isbn=0596003137 at UNIVERSITY OF CALIFORNIA MELVYL melvyl.cdlib.org:210 CDL90 USMARC (1 forks) |
08:12 |
|
thd |
17403/10 : 2 >> |
08:12 |
|
thd |
17403/10 : creating and 17403/10 : working on results entry 11 |
08:12 |
|
thd |
17403/10 : connected to UNIVERSITY OF CALIFORNIA MELVYL |
08:12 |
|
thd |
17403/10 : UNIVERSITY OF CALIFORNIA MELVYL : 1 records found, retrieving them (max 80) |
08:12 |
|
thd |
17403/10 : melvyl.cdlib.org:210 search done. |
08:14 |
|
thd |
kdl: yet no record in the breeding farm |
08:39 |
|
hdl |
thd : seems to me that you are trying to get a biblio that you already have in you base. |
08:40 |
|
hdl |
Have you verified ? |
08:45 |
|
thd |
hdl: verified? |
08:46 |
|
thd |
hdl: what and where is my 'base'? |
08:52 |
|
thd |
hdl: the isbn comes from my last attempt to find a record over z39.50 where the daemon was most likely not running |
08:53 |
|
thd |
hdl: does the su user for the daemon need a login? |
08:56 |
|
slef |
thd: it needs a user, existing in /etc/passwd or whereever your users come from. What do you mean 'login'? |
08:57 |
|
thd |
slef: no password for that user for security |
08:57 |
|
thd |
slef: user www |
08:58 |
|
slef |
does it have a shell? |
08:59 |
|
thd |
slef: yes, www did not formerly for security, but I obviusly had to ad one to try to run the daemon. |
08:59 |
|
slef |
that's not the problem |
08:59 |
|
slef |
then |
09:00 |
|
slef |
:-/ |
09:00 |
|
thd |
slef: any clue, something to test? |
09:03 |
|
slef |
no, normal tricks, read the log files, try running them with sh -x and so on |
09:04 |
|
thd |
slef: if you read the previous communication to hdl I can get processz3950queue to do something with the previous incomplete record search. |
09:05 |
|
thd |
slef: nothing relevant appears in the logs |
09:05 |
|
slef |
does anything appear? |
09:06 |
|
slef |
as in the daemon log, not apache logs |
09:06 |
|
thd |
slef: yes, but nothing related to z39.50 record searches, daemons, or otherwise |
09:07 |
|
slef |
how can there be nothing about daemons in the z3950 daemon log? |
09:08 |
|
thd |
slef: It seems that the daemon may never have run |
09:09 |
|
thd |
slef: I have tried to start the daemon launcher as root assuming the daemon will assume the www user id |
09:11 |
|
slef |
yes, that should work |
09:11 |
|
slef |
you can also su to www and try running the daemon directly |
09:11 |
|
thd |
slef: what process name should I see running ulimately if it does work? |
09:11 |
|
slef |
have you set the environment variables KOHA_CONF and PERL5LIB? |
09:12 |
|
slef |
perl :-/ |
09:12 |
|
thd |
slef: perl? |
09:12 |
|
slef |
yes, it appears in ps as perl, I think |
09:12 |
|
slef |
I've not got a working z3950 daemon at the moment to test (don't ask) |
09:13 |
|
thd |
slef: ok, I had not understood the question that was answered there |
09:14 |
|
thd |
slef: yours is broken too :- / |
09:14 |
|
slef |
yes, hasn't worked in 2.2 |
09:15 |
|
slef |
I guess I should just let jared have the installer and I'll go nuts on the biblio entry side |
09:15 |
|
slef |
I want to finish the basic cleanups first, though |
09:18 |
|
hdl |
thd : sorry I was on a hot potatoe and it didn't ring me when you wrote. |
09:38 |
|
thd |
slef: sorry lost connection for some minutes |
09:40 |
|
thd |
slef: and you have not solved your broken daemon or have not bothered? |
09:40 |
|
thd |
slef: Is there a difference, possibly unintended, in upgraded as opposed to newly created installs that may cause a problem? |
09:40 |
|
slef |
not solved... not tried for a little while |
09:41 |
|
slef |
no, a new test install didn't work either |
09:41 |
|
hdl |
thd: I said You may already have the notice in your database. have you verified if you had it already ? |
09:41 |
|
hdl |
look in your catalogue. |
09:42 |
|
thd |
hdl: for the biblio? |
09:42 |
|
hdl |
yes. |
09:42 |
|
thd |
hdl: It is not found searching the breeding farm |
09:42 |
|
hdl |
When you already have the biblio you try to get with z3950. It simply doesn't import it. |
09:43 |
|
hdl |
Not in the breeding form, in your catalogue. |
09:43 |
|
hdl |
Try to search for a book you are SURE you donot have. |
09:44 |
|
hdl |
Evangelium Vitae if you are not a religious library. |
09:45 |
|
hdl |
Or Learning PERL If you donot have computer science books. |
09:45 |
|
hdl |
But beware. Run procesz3950 before to see the log process ;). |
09:45 |
|
thd |
hdl: I get the same negative results for books I am sure I do not have as for those I was hoping to have. |
09:46 |
|
thd |
hdl: nothing ever appears in the z3950 log, ever |
09:47 |
|
hdl |
17403/10 : Processing isbn=0596003137 at UNIVERSITY OF CALIFORNIA MELVYL melvyl.cdlib.org:210 CDL90 USMARC (1 forks) |
09:47 |
|
hdl |
[15:12:48] <thd> 17403/10 : 2 >> |
09:47 |
|
hdl |
[15:12:48] <thd> 17403/10 : creating and 17403/10 : working on results entry 11 |
09:47 |
|
hdl |
[15:12:48] <thd> 17403/10 : connected to UNIVERSITY OF CALIFORNIA MELVYL |
09:47 |
|
hdl |
[15:12:48] <thd> 17403/10 : UNIVERSITY OF CALIFORNIA MELVYL : 1 records found, retrieving them (max 80) |
09:47 |
|
hdl |
[15:12:49] <thd> 17403/10 : melvyl.cdlib.org:210 search done. |
09:47 |
|
hdl |
seems to me there is an answer ;) |
09:47 |
|
thd |
hdl: I know that should be in the log but it never was |
09:47 |
|
hdl |
1 record found. |
09:48 |
|
hdl |
where hav you taken those lines then ? |
09:48 |
|
thd |
hdl: the daemon seems to have never been running |
09:48 |
|
thd |
hdl: those lines were from the console |
09:49 |
|
hdl |
Leave aside the daemon for now. You say it is z3950 client from Koha that is broken. Let's test it deeply. |
09:49 |
|
thd |
hdl: I piped them to a file so I could report them after confirming again that the log was empty |
09:49 |
|
thd |
hdl: ok |
09:50 |
|
thd |
hdl: how? |
09:50 |
|
hdl |
so Can you launch processz3950 in a console after having set KOHA_CONF AND PERL5LIB. |
09:51 |
|
hdl |
processz3950queue sorry... |
09:51 |
|
thd |
hdl: yes, hence the lines reported |
09:52 |
|
thd |
hdl: Is there a way of passing values to processz3950queue without using koha |
09:52 |
|
hdl |
so lets try a new search on a book you never had but are sure to get from the library you ask to. |
09:52 |
|
hdl |
thd: Don't think so. |
09:53 |
|
hdl |
In fact, it is just a loop that gets data from Koha and send it to the z3950 server through perl. |
09:54 |
|
hdl |
You can also test with perl... |
09:54 |
|
thd |
hdl: But, I must supply the ISBN within Koha |
09:56 |
|
thd |
hdl: since I cannot pass the ISBN directly to processz3950queue from the command line? |
09:59 |
|
hdl |
You can also pass a title or an author name ;) |
10:01 |
|
thd |
hdl: I was uncertain that the interface supported other than Isbn record searches currently, I was going to fix the issue if it was restricted to isbn only |
10:02 |
|
hdl |
It is not restrected to isbn. I can assure you. |
10:03 |
|
thd |
hdl: Good, I wondered what the special mark for ISBN was for. |
10:05 |
|
thd |
hdl: I should enter the isbn, author, or title in the add marc dialogue and the press the search button or then run processz3950queue from the command line? |
10:06 |
|
hdl |
First: run ONE processz3950queue on a command line. |
10:06 |
|
thd |
hdl: we know the ideal answer is the button, but for your test. |
10:06 |
|
thd |
hdl: using what user name? |
10:07 |
|
hdl |
Enter then one author and press the search button. |
10:07 |
|
hdl |
When using processz3950queue, it copes with user names form you daemon options. |
10:07 |
|
thd |
hdl: I run processz3950queue in advance of the search? |
10:08 |
|
hdl |
Yes. |
10:09 |
|
hdl |
It will wait, receive your search strings, fork, send search query, wait for reply... And you should have some answer. |
10:13 |
|
thd |
hdl: well, as I had seen previously it first works with the previous incomplete search. |
10:13 |
|
thd |
hdl: that is what happened just now, with the same ISBN I had been using to test. |
10:15 |
|
thd |
hdl: I will now run processz3950queue again. Now that the previous run has worked with the old ISBN. |
10:16 |
|
hdl |
thd : you can let processz3950 run in background ;)))) |
10:17 |
|
hdl |
But you really ought to test with an unknown bokk, sot that you could get an answer. |
10:17 |
|
hdl |
I had that problem once, couldn't get around. And I realized weeks later it already worked fine ;) |
10:17 |
|
thd |
hdl: I never get the command prompt back when when running processz3950queue as a foreground process. I had always used ^C to abort the process when the prompt did not return. |
10:18 |
|
hdl |
you can have multiple console with Konsole ;) |
10:19 |
|
thd |
hdl: Is konsole a KDE thing? : ) |
10:21 |
|
hdl |
yes, but anyway, you can have more than one console... tty1,2,3,4,5,6... |
10:22 |
|
hdl |
But always with Linux. |
10:25 |
|
hdl |
thd ??? |
10:29 |
|
thd |
hdl: just a moment, I had run processz3950queue with sudo and had to kill the process running in background seemed to cause a strange presentation of the password prompt for the sudo command. |
10:34 |
|
thd |
hdl: well, at least this is different. "still ?? requests to go" followed by "still 1 request to go" followed by "The requested URL /cgi-bin/koha/z3950/0 was not found on this server". |
10:35 |
|
slef |
hdl: screen does that for most people. |
10:39 |
|
hdl |
strange : It works fine for me. |
10:39 |
|
thd |
hdl: what works fine for you> |
10:40 |
|
hdl |
It shows me the results. |
10:41 |
|
hdl |
I don't use the deamon. But procesz3950queue shows me good results. |
10:42 |
|
thd |
hdl: why would anyone ever get "The requested URL /cgi-bin/koha/z3950/0 was not found on this server" as the final page appearing on the popup search window? |
10:44 |
|
thd |
hdl: that is more satisfying than a perpetual "still ?? requests to go" ; ) |
10:48 |
|
thd |
hdl: in console I get 3 records found now. I had only reported the popup before. |
10:49 |
|
hdl |
thd : I searched for java at Lyon : |
10:49 |
|
hdl |
here are my results : |
10:49 |
|
hdl |
19004/45 : Processing title=JAVA at LYON3 193.52.199.5:21210 ADVANCE MARC21 (1 forks) |
10:49 |
|
hdl |
19005/45 : Processing title=JAVA at KOHATEST localhost:9999 kohatest UNIMARC (2 forks) |
10:50 |
|
hdl |
19004/45 : 2 >> |
10:50 |
|
hdl |
19004/45 : creating and 19004/45 : working on results entry 71 |
10:50 |
|
hdl |
19004/45 : connected to LYON3 |
10:50 |
|
hdl |
@attr 1=4 "JAVA"19004/45 : LYON3 : 66 records found, retrieving them (max 80) |
10:50 |
|
hdl |
19005/45 : no connection at KOHATEST |
10:50 |
|
hdl |
19005/45 : localhost:9999 search done. |
10:50 |
|
hdl |
19004/45 : LYON3 : 193.52.199.5:21210 records retrieved 66 SPEED: 3 |
10:50 |
|
hdl |
19004/45 : 193.52.199.5:21210 search done. |
10:50 |
|
hdl |
|
10:50 |
|
hdl |
and the popup shows : |
10:50 |
|
hdl |
Résultats de la recherche Z39.50 |
10:50 |
|
hdl |
Notices trouvées |
10:50 |
|
hdl |
Titre Auteur ISBN venant de |
10:50 |
|
hdl |
Indonésie Élisabeth Richard, Antoinette Vicart 2700308018 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Le grand guide de Java 2070588300 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Programmation Java Jean-François Macary, Cédric Nicolas 2212089163 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Formation à Java Stephen R. Davis 284082177X Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
HTLM & Java L. Viegnes & S. Boix 2742907335 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
L'art de l'Asie du Sud-Est Philip Rawson 2878110889 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Java Gilles Clavel, Nicolas Mirouze, Emmanuel Pichon... 2729606564 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Lieux d'Islam dirigé par Mohammad Ali-Moezzi 2862605808 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Les langages de développement sur Web Imad Saleh 2866015428 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Java client-serveur 2212089384 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
CORBA, ActiveX et Java Beans Jean-Marie Chauvet 2212089503 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Java Michel Bonjour, Gilles Falquet, Jacques Guyot et André Le Grand 2841801349 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Constructing intelligent agents with JAVA 0471191353 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
JDBC et Java George Reese 2841770427 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Le programmeur Java 1.2 Laura Lemay et Rogers Cadenhead 2744005193 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Penser objet avec UML et Java Michel Lai,... 2225834288 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Java 1.1 Michel Martin 2736127587 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Le dictionnaire officiel Java 2 Patrick Chan 2212090897 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Java 2 Rogers Cadenhead 2744006971 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
UML, la notation unifiée de modélisation objet Michel Lai 2100050230 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
UML en action Pascal Roques, Franck Vallée 2212091273 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Understanding SQL and Java together 1558605622 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Rural industrialization in Indonesia by KÅosuke Mizuno 4258520314 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Programmation Java côté serveur Andrew Patzer 2212091095 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Sécurité en Java Scott Oaks 284177063X Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Penser objet avec UML et Java Michel Lai 2100053787 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
Java, la synthèse Gilles Clavel... 2100053795 Z3950-LYON3 Importer cette notice |
10:51 |
|
hdl |
XML et Java Jean-Christophe Bernadac, François Knab, Franck Lepoivre... 2212091486 Z3950-LYON3 Importer cette notice |
10:53 |
|
hdl |
and so on and so forth. |
10:54 |
|
hdl |
Maybe the problem comes form a UNIMARC flavor ??? |
10:54 |
|
hdl |
kados ??? |
10:54 |
|
hdl |
around ? |
10:55 |
|
thd |
hdl: kados is busy making business at the Chicago ALA convention. |
10:56 |
|
thd |
hdl: I have never seen the liblime demo do anything at all (no popup) when attempting a z39.50 search. |
10:58 |
|
thd |
hdl: Actually, I have not set a working z39.50 search on any koha demo server for years now. |
11:00 |
|
thd |
slef: why does processz3950queue have a 20 June date? Did you change anything in that file? I have not checked the diffs yet? |
11:03 |
|
unmadindu |
kados: ping |
11:04 |
|
hdl |
kados : pong... kados is busy making business at the Chicago ALA convention. |
11:05 |
|
unmadindu |
hdl: this is indradg chatting from a friend's comp |
11:05 |
|
hdl |
thd :I don't have this date on my processz3950queue. |
11:05 |
|
hdl |
hi indradg ;) |
11:05 |
|
unmadindu |
hdl: Koha Live CD for ALSA conference is ready, but need 350 MB to upload |
11:07 |
|
unmadindu |
s/ALSA/ALA |
11:08 |
|
unmadindu |
hdl: kados wanted to hand out the Live CDs |
11:08 |
|
hdl |
maybe you could put it on the koha-fr ftp server when it is uprgaded if kados wants to get it from there. |
11:09 |
|
hdl |
always the sams problems too busy busy ;) |
11:10 |
|
unmadindu |
hdl: ok thanks |
11:10 |
|
thd |
hdl: I got the latest version, more resent than 2.2.3 because the -s switch for the su command does not exist in BSD where I installed koha. slef removed the -s switch 20 June. That was a redundant to prevent using cshell, then the options would not be read correctly I presume. |
11:12 |
|
thd |
hdl: Removing the su -s switch should not have caused any problem. |
11:12 |
|
hdl |
I don't think so. |
11:21 |
|
hdl |
thd : have you investigated z3950/search.pl code ? |
11:21 |
|
hdl |
Line 103 : if numberpending = 0 then return 0.... and not a page. |
11:22 |
|
hdl |
You should take a closer look to checkz3950 subroutine. |
11:22 |
|
hdl |
As I will to. |
11:28 |
|
thd |
hdl: I assumed if it worked for you the code would not be the problem. Only my configuration and use would be at issue. |
11:30 |
|
thd |
hdl: Although, it is a little worrying that slef has not had a working z3950 since 2.2. Yet, he did not fully investigate the issue for himself. |
11:40 |
|
hdl |
thd... : try this : in z3950/search.pl code ? |
11:40 |
|
hdl |
[18:21:52] <hdl> Line 103 : if numberpending = 0 then return 0.... |
11:41 |
|
hdl |
change ? 0 by ? ""' |
11:42 |
|
hdl |
and retry. |
11:57 |
|
hdl |
unmadindu : do you want an ftp access ? |
11:57 |
|
thd |
hdl: in line 103 I changed '? 0' to '? ""'. Note the code quotation and all changes are enclosed in a single quote but no single quote was used in changes. |
11:59 |
|
hdl |
can you copy/paste the line you got? |