| Time |
Nick |
Message |
| 00:15 |
rfrasur |
jcamins: ty...will look into that. |
| 00:15 |
* rfrasur |
also calls it a night. |
| 01:21 |
* paxed |
wonders if he's doing something wrong. installing manually, doing the "configure, make, make install" results in cp: cannot stat ‘branding/*’: No such file or directory |
| 01:21 |
|
Mark__T joined #evergreen |
| 01:31 |
paxed |
also, INSTALL says cp Open-ILS/examples/apache_24/eg.conf ... but the apache 2.4 files are eg_24.conf and eg_vhost_24.conf |
| 02:04 |
|
Callender_ joined #evergreen |
| 02:06 |
|
eeevil joined #evergreen |
| 02:06 |
|
gdunbar joined #evergreen |
| 02:08 |
|
senator joined #evergreen |
| 02:08 |
|
csharp joined #evergreen |
| 02:10 |
|
jcamins joined #evergreen |
| 02:20 |
|
remingtron_ joined #evergreen |
| 02:35 |
|
bshum joined #evergreen |
| 02:40 |
|
pastebot joined #evergreen |
| 05:17 |
|
bshum joined #evergreen |
| 05:17 |
|
_bott_ joined #evergreen |
| 05:17 |
|
atheos joined #evergreen |
| 05:17 |
|
tater joined #evergreen |
| 05:20 |
|
csharp joined #evergreen |
| 05:20 |
|
pastebot joined #evergreen |
| 05:20 |
|
remingtron_ joined #evergreen |
| 05:20 |
|
senator joined #evergreen |
| 05:20 |
|
GeoffSams joined #evergreen |
| 05:20 |
|
SimonHM joined #evergreen |
| 05:20 |
|
ldwhalen joined #evergreen |
| 05:20 |
|
mtcarlson_away joined #evergreen |
| 05:20 |
|
phasefx_ joined #evergreen |
| 05:20 |
|
dbs joined #evergreen |
| 05:20 |
|
zxiiro joined #evergreen |
| 05:20 |
|
b_bonner joined #evergreen |
| 05:20 |
|
jeff joined #evergreen |
| 05:20 |
|
tsbere joined #evergreen |
| 05:20 |
|
fparks joined #evergreen |
| 05:20 |
|
leofseige joined #evergreen |
| 07:02 |
|
timlaptop joined #evergreen |
| 07:09 |
|
jboyer-isl joined #evergreen |
| 07:41 |
|
collum joined #evergreen |
| 08:11 |
|
akilsdonk_ joined #evergreen |
| 08:34 |
|
csharp joined #evergreen |
| 08:36 |
|
kbeswick joined #evergreen |
| 08:40 |
|
kmlussier joined #evergreen |
| 08:40 |
|
Shae joined #evergreen |
| 08:49 |
|
mmorgan joined #evergreen |
| 08:56 |
csharp |
netsplits-- |
| 08:58 |
dbs |
paxed: master, at least, doesn't say that. |
| 09:00 |
dbs |
as part of a2078f2213a |
| 09:00 |
pinesol_green |
[evergreen|Dan Scott] Update docs for Fedora install - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=a2078f2> |
| 09:00 |
|
ericar joined #evergreen |
| 09:23 |
paxed |
dbs: line 220? |
| 09:24 |
paxed |
err |
| 09:24 |
dbs |
http://git.evergreen-ils.org/?p=Evergreen.git;a=blob;f=docs/installation/server_installation.txt;h=9834628255b784dd02fb8a6fb9254b0138f04f64;hb=refs/heads/master#l220 looks good to me. |
| 09:25 |
paxed |
yea, i got confused by how the debian/ubuntu is for apache < 2.4, and the fedore is for apache >= 2.4 |
| 09:31 |
|
rjackson-isl joined #evergreen |
| 09:35 |
|
mrpeters joined #evergreen |
| 09:39 |
|
yboston joined #evergreen |
| 10:01 |
|
mllewellyn joined #evergreen |
| 10:02 |
|
Dyrcona joined #evergreen |
| 10:47 |
paxed |
aaaargh |
| 10:48 |
bshum |
Sounds about right today. |
| 10:48 |
paxed |
if it was only for today, it would be a relief. |
| 10:48 |
paxed |
unfortunately, this has been going on for days now. and unless this is solved, we will not be able to use evergreen. |
| 10:51 |
paxed |
fieldmapper gives english texts. seems like at random. |
| 10:51 |
paxed |
possibly a caching problem. |
| 10:53 |
dbs |
Are you using most recent OpenSRF master? https://bugs.launchpad.net/opensrf/+bug/1188195 seems applicable |
| 10:53 |
|
jdouma joined #evergreen |
| 10:53 |
pinesol_green |
Launchpad bug 1188195 in OpenSRF "Per-process default session locale for clients (Perl)" (affected: 1, heat: 6) [Undecided,Fix committed] |
| 10:57 |
|
zerick joined #evergreen |
| 10:57 |
paxed |
dbs: i think so, but i will redo our dev box to be sure |
| 11:15 |
|
dboyle joined #evergreen |
| 11:20 |
|
ElliotFriend joined #evergreen |
| 12:10 |
csharp |
gmcharlt++ # volunteering to teach IRC school |
| 12:11 |
gmcharlt |
er, more of a TA role was what I had in mind :) |
| 12:11 |
csharp |
heh |
| 12:16 |
|
smyers_ joined #evergreen |
| 12:33 |
paxed |
dbs: ok, i now definitely have that opensrf fix. and it doesn't help. |
| 12:34 |
paxed |
dbs: on linux, it usually(?) works. on windows, not at all. |
| 12:37 |
paxed |
would anyone be willing to test this for me? http://62.148.106.91/staff_client/ (use the _testing.exe one), and use that same server to log on. egadmin/egadmin |
| 12:41 |
paxed |
this isn't good. because we should be deploying in few months ... |
| 12:48 |
dbs |
paxed: um, what server? |
| 12:49 |
paxed |
^ |
| 12:50 |
paxed |
the ip |
| 12:50 |
csharp |
paxed: testing... what should I be looking for? |
| 12:50 |
paxed |
csharp: change the staff client language to finnish. see if patron registration is in english or in finnish. |
| 12:50 |
csharp |
(btw, it's weird in an awesome way to see all this finnish!) |
| 12:50 |
csharp |
ah - ok |
| 12:50 |
|
Rish joined #evergreen |
| 12:52 |
csharp |
okay - yeah, I see English in the staff registration |
| 12:52 |
paxed |
oh good, it's not just us. |
| 12:52 |
paxed |
and this is on windows, right? |
| 12:54 |
csharp |
yep Windows 7 |
| 12:54 |
csharp |
I'm creating a new profile to see if that changes anything... |
| 12:56 |
csharp |
yep - still English |
| 12:56 |
paxed |
both OpenSRF and Evergreen are current master |
| 13:00 |
csharp |
paxed: if I can help test anything else, just let me know |
| 13:00 |
paxed |
i have no idea how to proceed further. |
| 13:02 |
csharp |
sorry, I haven't needed to know this, so I just don't know... at what point should fi-FI be loaded? |
| 13:02 |
csharp |
is it done in the build process or on the fly somehow? |
| 13:06 |
bshum |
paxed: Tested with Linux side and it also doesn't play nicely. |
| 13:06 |
bshum |
(not that you need the extra confirmation I guess) |
| 13:08 |
csharp |
okay, so this - http://git.evergreen-ils.org/?p=Evergreen.git;a=blob;f=build/i18n/po/fm_IDL.dtd/fi-FI.po - is where the text is that *should* be loaded in the patron registration screen, right? |
| 13:10 |
* csharp |
reads the GNU gettext overview, which answers his questions |
| 13:11 |
csharp |
http://www.gnu.org/software/gettext/manual/html_node/Overview.html |
| 13:15 |
bshum |
@hate using the phone |
| 13:15 |
pinesol_green |
bshum: The operation succeeded. bshum hates using the phone. |
| 13:21 |
|
_bott_ joined #evergreen |
| 13:29 |
|
DPearl joined #evergreen |
| 13:35 |
|
Dyrcona joined #evergreen |
| 13:43 |
dbwells |
paxed: Maybe you already know all this, but I am trying to familiarize myself with how the translated IDL works. In testing, I could get the translated IDL labels to show in the Patron Reg screen in Firefox. |
| 13:43 |
dbwells |
Here is what I did: |
| 13:44 |
dbwells |
1) In about:config, I set the intl.accept_languages = fi-FI |
| 13:44 |
dbwells |
2) cleared cache, and browsed to http://62.148.106.91/eg/actor/user/register |
| 13:46 |
dbwells |
The 'IDL2js' Apache module looks first in the query string for a locale, then in the 'Accept-Language' header. Since I don't see the locale set in any IDL2js requests, I am assuming it is expecting the "Accept-Language" header to be set. |
| 13:47 |
dbwells |
I don't know how that is supposed to be set in the staff client, but you could use about:config to set it there if needed. |
| 13:48 |
paxed |
dbwells: setting the locale in the staff client sets intl.accept_languages |
| 13:49 |
dbwells |
paxed: where does one set the locale in the staff client? |
| 13:49 |
tsbere |
login screen |
| 13:50 |
dbwells |
ah, right, thanks |
| 13:50 |
phasefx |
dbwells: paxed: a heavy-handed hack I did before IDL2js got smarter with locales was to make an alternate base.tt2 for the patron editor to load, one that omitted the <script> tag for /IDL2js |
| 14:02 |
|
RBecker joined #evergreen |
| 14:03 |
dbwells |
paxed: If you hardcode 'fi-FI' as the default locale in OpenILS/WWW/IDL2js.pm, does that help? Just trying to narrow down the source of the problem... |
| 14:03 |
dbwells |
fg |
| 14:03 |
berick |
ctrl-z |
| 14:03 |
gmcharlt |
bg |
| 14:04 |
dbwells |
dang, where did vi go? ;) |
| 14:06 |
paxed |
dbwells: ok, i just hardcoded $locale = 'fi-FI'; in handler... |
| 14:08 |
paxed |
doesn't seem to help, but i don't know if that's cached or not. |
| 14:11 |
berick |
it's cached per locale |
| 14:12 |
paxed |
right, saw that. restarting eg |
| 14:15 |
paxed |
well, now i'm getting finnish on the field labels, but all the rest of it is still in english (the Examples, "Save" and "Save & Clone" -button, etc) |
| 14:16 |
|
dconnor joined #evergreen |
| 14:16 |
|
dconnor joined #evergreen |
| 14:17 |
|
dconnor joined #evergreen |
| 14:17 |
dbwells |
Those labels wouldn't be from the IDL, so that's a different issue altogether. |
| 14:17 |
dbwells |
Well, maybe not 'altogether different', but still different. |
| 14:19 |
paxed |
osrfsys.log: http://62.148.106.91/osrfsys.txt |
| 14:20 |
paxed |
that's live, and loglevel is 4, so... |
| 14:22 |
dbwells |
I am wondering if the "oils://remote" wrapper thingy doesn't send along the 'Accept-Language' header or something. But really, I am now just combining two things I know very little about :) |
| 14:25 |
dbs |
paxed: you've added alert() or console.log() stuff to dojo/fieldmapper/IDL.js to check on the value of OpenSRF.locale? |
| 14:26 |
paxed |
dbs: istr i did last week, and OpenSRF.locale was '' or something like that - but i could be mistaken |
| 14:27 |
dbs |
That would be a helpful piece. As would seeing if that block of code where the dojo.xhr() call is made is even ever invoked |
| 14:28 |
paxed |
dbs: ah, right - i think that piece of code was never executed in my tests. |
| 14:32 |
paxed |
just put an alert in IDL.js before the dojo.xhrGet() to show OpenSRF.locale - and it just doesn't happen. |
| 14:35 |
|
jbfink joined #evergreen |
| 14:35 |
jbfink |
hey people, help a sleep deprived dude out: |
| 14:35 |
jbfink |
in http://evergreen-ils.org/documentation/install/README_2_4.html section 9, in the eg_db_config thing, what is the right value for <port>? |
| 14:36 |
bshum |
jbfink: Most of the time it's usually 5432 (if the postgres is running on the default port) |
| 14:37 |
jbfink |
aces |
| 14:37 |
jbfink |
danke |
| 14:38 |
eeevil |
berick: do you recall the benefits of IDL2js over AutoIDL and IDL chunking? was there a serious speed increase? or was it only recently that AutoIDL learned about locale awareness? |
| 14:44 |
berick |
eeevil: IDL2js is smaller/faster than loading fm_IDL.xml. |
| 14:44 |
berick |
IDL2js does chunking, fwiw, but the staff client always loads the whole thing anyway |
| 14:46 |
berick |
i don't recall how much smaller/faster, though |
| 14:46 |
eeevil |
berick: ah, the xul interfaces in particular? (and those html ones not yet taught how to use AutoIDL, I guess) |
| 14:46 |
|
stevenyvr2 joined #evergreen |
| 14:48 |
berick |
yeah, mostly the xul interfaces or any interface where we know we need a wide array of classes so we just use the whole thing, since it's cached in the browser/xul |
| 14:48 |
jbfink |
does eg_db_config make changes to ejabberd? |
| 14:48 |
jbfink |
at the very end of step 9, something happens that makes my ejabberd not start. |
| 14:48 |
berick |
also, there are some cases where AutoIDL does not give you everthing you need. trying to recall the details.. stuff like var f = new fieldmapper.abc() |
| 14:48 |
berick |
will fail in a pure AutoIDL environmen |
| 14:48 |
berick |
t |
| 14:50 |
eeevil |
if the class def for abc hasn't been loaded yet, you mean? |
| 14:50 |
berick |
right |
| 14:50 |
eeevil |
right... you'd have to OpenSRF.IDL.load(['abc']) first |
| 14:51 |
eeevil |
or similar |
| 14:51 |
eeevil |
(however it's spelled ... I should look a the code ;) ) |
| 14:52 |
jbfink |
actually hang on a sec |
| 14:52 |
jbfink |
aright nevermind, got it |
| 14:54 |
jbfink |
holy god it works |
| 14:55 |
jbfink |
functioning dockerized evergreen |
| 14:55 |
bshum |
jbfink: What's "dockerized"? And cool! :) |
| 14:56 |
jbfink |
bshum: http://docker.io |
| 14:56 |
paxed |
berick, eeevil: i'd appreciate any pointers ... |
| 14:56 |
jbfink |
bshum: basically a kinda-sorta virtualization thing, except portable -- meaning I can give you my dockerized evergreen and you can run it exactly how it works for me. |
| 14:56 |
jbfink |
provided you have a working docker install. |
| 14:56 |
bshum |
jbfink: Ah, interesting. |
| 14:57 |
berick |
paxed: well, i'm only partly paying attention, but it sounds like the locale is not making it from the staff client to the IDL2js, correct? |
| 14:58 |
eeevil |
paxed: I'm trying to reason in the general area of the problem ... to see if there's some foundational logic problem that we're just now discovering. I think dbwells may be onto something, though, with the oils:// scheme theory |
| 14:59 |
* phasefx |
remembers testing the IDL2js patch for localization support and having it work, specifically with the patron editor |
| 14:59 |
* eeevil |
wonders if there's some way to log incoming headers that apache sees |
| 15:00 |
phasefx |
or, I vaguely remember that |
| 15:04 |
berick |
paxed: is the patron editor what you're looking at? |
| 15:18 |
paxed |
berick: yes. |
| 15:19 |
|
DPearl joined #evergreen |
| 15:20 |
paxed |
phasefx: istr i did the same thing (or whatever the change was), and patron editor worked then. although now i'm starting to suspect it was cached ... |
| 15:20 |
berick |
paxed: if you view source on the patron reg screen (under debug along top-right of xul window), does it show fi_fi for the html lang / xml:lang ? |
| 15:22 |
paxed |
berick: yes |
| 15:23 |
berick |
and just to clarify, what we're talking about now are the strings in the page, not the IDL labels, correct? |
| 15:23 |
|
_zerick_ joined #evergreen |
| 15:23 |
berick |
the conversation from 14:15 (eastern) says IDL labels are now working.. |
| 15:23 |
jbfink |
now gotta figure out a programmatic way to change ejabberd passes, reflect changes in opensrf_core.xml, and also everything else I assigned a very lame password to. |
| 15:24 |
moodaepo |
bshum++ |
| 15:24 |
paxed |
berick: the idl labels are working because i tested a hard-coding the locale in IDL2js.pm |
| 15:24 |
berick |
oh, ok |
| 15:24 |
|
18WAENSG7 joined #evergreen |
| 15:24 |
paxed |
berick: i'll remove that test, and restart eg &c to clear the cache |
| 15:24 |
berick |
paxed: mind testing this? user/berick/staff-web-ui-idl-locale |
| 15:25 |
berick |
@ working |
| 15:25 |
pinesol_green |
berick: Go away, or I'll replace you with a very small shell script! |
| 15:25 |
|
zerick joined #evergreen |
| 15:26 |
berick |
i don't know why accept-lang is not coming through, but that patch will take advantage of the already-parsed locale and relay it to IDL2js |
| 15:27 |
paxed |
berick: will do... |
| 15:36 |
paxed |
berick: hmm... well. the idl labels seem to work now. unfortunately this isn't good enough, because the rest is still in english. |
| 15:37 |
berick |
well, one technical hurdle at a time |
| 15:37 |
paxed |
yeah |
| 15:37 |
berick |
that's very odd that the template engine (EGWeb.pm) has the correct locale |
| 15:38 |
berick |
but it's not loading the correct template.. |
| 15:38 |
berick |
well, wait, do you have a po file for that UI? |
| 15:38 |
paxed |
berick: one other data point: for some reason everything works for me on linux. (that is, i get everything we've translated in finnish) |
| 15:39 |
paxed |
not just the idl stuff, but also the other texts on the page. |
| 15:40 |
berick |
so.. the problem is the windows staff client? |
| 15:41 |
paxed |
i don't know. i've randomly seen it not working on linux too - but it's very rare. |
| 15:42 |
berick |
paxed: same server in both cases, though? |
| 15:42 |
paxed |
... perhaps related to bug 1096871 |
| 15:42 |
pinesol_green |
Launchpad bug 1096871 in Evergreen "TPAC swaps between translated strings in search filter lists" (affected: 1, heat: 6) [Undecided,Confirmed] https://launchpad.net/bugs/1096871 |
| 15:42 |
paxed |
yes |
| 15:43 |
paxed |
and, i really must say this: if we can't solve the problem, we'll crash and burn. there's no way we could use evergreen... |
| 15:48 |
berick |
paxed: ah, you should save the big guns for the big problems! this is probably a lame, 2-line bug somewhere. |
| 15:48 |
berick |
paxed: mind setting your apache log level to debug and telling me what you see when you load the patron reg UI in the apache logs. something like: |
| 15:48 |
berick |
[Tue Sep 10 15:46:16 2013] [debug] EGWeb.pm(77): [client 65.190.183.162] egweb: messages locale = en_us |
| 15:48 |
berick |
[Tue Sep 10 15:46:17 2013] [debug] IDL2js.pm(83): [client 65.190.183.162] IDL2js using locale 'en-US' |
| 15:50 |
berick |
similarly, does it work if you load the UI from Firefox? https://hostname/eg/actor/user/register ? |
| 15:52 |
paxed |
berick: https://62.148.106.91/eg/actor/user/register |
| 15:52 |
paxed |
in firefox, the IDL are in english, the string already in the page are in finnish. |
| 15:53 |
paxed |
you mean apache logs or osrfsys.log or what? |
| 15:55 |
berick |
apache error logs (e.g. /var/log/apache2/error.log) |
| 15:58 |
berick |
that's odd the IDL is english in FF. the IDL link in the source is correct and returns finnish. wonder if a clear-cache is needed in the FF test. /me can't see the IDL strings in the page w/o a login |
| 15:58 |
paxed |
berick: egadmin/egadmin |
| 15:58 |
paxed |
it's a dev server |
| 15:59 |
berick |
phasefx: thanks.. i'm getting finnish for the IDL labels there. |
| 15:59 |
paxed |
/etc/apache2/apache2.conf:LogLevel debug ... but i can't see the egweb/IDL2js locale message. |
| 15:59 |
paxed |
in the apache log |
| 15:59 |
berick |
Tunnus, Vaihtoehtoinen käyttäjätunnus, Salasana... |
| 16:00 |
berick |
well that's odd |
| 16:00 |
berick |
re: logs |
| 16:01 |
paxed |
and my intl.accept_languages is fi-FI and i don't get finnish for the IDL labels. |
| 16:01 |
berick |
in FF, though, i'm not getting finnish for the page strings |
| 16:01 |
paxed |
so the total opposite for me. |
| 16:02 |
|
kmlussier joined #evergreen |
| 16:02 |
berick |
hm, ok, in FF i have Finnish [fi] as a locale |
| 16:02 |
berick |
but not fi-FI |
| 16:02 |
berick |
FF doesn't give me an fi-FI option |
| 16:03 |
paxed |
change accept_languages to just "fi" - and everything is in english. |
| 16:04 |
paxed |
hm. are the User setting names also IDL labels? |
| 16:04 |
paxed |
the default phone number, etc? because i get those in finnish. |
| 16:05 |
phasefx |
I'd expect those to be coming from i18n_core |
| 16:06 |
berick |
yeah, what phasefx said |
| 16:06 |
berick |
via config.org_unit_setting_type |
| 16:06 |
paxed |
...and wtf? i just reloaded the page, on both linux (my OS), and windows (running under virtualbox), and both now show everything in finnish?!! |
| 16:06 |
berick |
.label |
| 16:07 |
berick |
hm, ok, something getting cached somewhere? |
| 16:07 |
paxed |
ergh. |
| 16:07 |
dbwells |
berick: Not sure if it helps (not following closely), but I was earlier able to use about:config in FF to set the intl.accept_languages = fi-FI manually, if that helps for testing. |
| 16:07 |
berick |
ah, thanks dbwells |
| 16:07 |
paxed |
well, the linux one shows everything in finnish. the windows one doesn't show the page strings, only idl labels, in finnish. |
| 16:08 |
berick |
w/ that (fi-FI), it's all finnish for me (in FF) |
| 16:09 |
* berick |
tries a few shift-reloads |
| 16:09 |
* berick |
watches the bits flying across (under?) the Atlantic |
| 16:10 |
dbwells |
berick: yes, it was all Finnish that way for me earlier, too. Not sure if it was luck, or a staff client-specific problem, or something else. |
| 16:10 |
paxed |
so, basically, everyone gets different results? |
| 16:10 |
paxed |
should i perhaps turn off memcached or something? |
| 16:11 |
berick |
well, i'm consistently getting all finnish in FF on linux using fi-FI |
| 16:11 |
* berick |
tries swapping locales back and forth |
| 16:11 |
paxed |
what about the staff client on windows? |
| 16:12 |
berick |
two things I don't have handy |
| 16:12 |
paxed |
http://62.148.106.91/staff_client/ - can't help with the lack of windows, though... |
| 16:13 |
phasefx |
I'll poke at the client in a seoncd |
| 16:13 |
phasefx |
second, even |
| 16:16 |
paxed |
is there a way to turn off caching on the server end all together? would just turning off memcached do the trick? |
| 16:16 |
berick |
switching to en-US, everything *but* the IDL is english consistently (FF/linux). oddly, the template thinks the locale is fi-FI. |
| 16:17 |
phasefx |
still hard-coded? |
| 16:17 |
dbwells |
Earlier on, I was able to get it to load in FF fine, but the Staff Client wouldn't work. I verified that the intl.whatever setting as 'fi-FI' even in the staff client, but that didn't seem to matter. That's what made me wonder whether the 'oils://remote' wrapper was losing 'Accept-Language'. Just recapping... :) |
| 16:18 |
dbwells |
berick: I had suggested as a troubleshooting means to hardcode 'fi-FI' in IDL2js. Not sure if that is still in effect, or how hard-coded it was (i.e. just a default, or always not matter what). |
| 16:19 |
dbwells |
s/not matter what/no matter what/ |
| 16:19 |
paxed |
dbwells: i removed the hardcoded locale in there, the only change from master right now is working/user/berick/staff-web-ui-idl-locale |
| 16:19 |
dbwells |
paxed: ok, good to know. |
| 16:20 |
paxed |
except, whoops, i forgot to remove the change from base.tt2 |
| 16:20 |
paxed |
fixing... |
| 16:20 |
* dbwells |
should probably stop drifting in and out of this conversation |
| 16:20 |
paxed |
well, now FF works as it should, eveything in finnish. |
| 16:21 |
paxed |
client on linux works too. |
| 16:21 |
paxed |
client on windows: no worky. |
| 16:22 |
berick |
paxed: so, do you get anything in the apache error log? |
| 16:22 |
|
RoganH joined #evergreen |
| 16:23 |
phasefx |
paxed: one thing to watch out for; you may need to restart the client whenever you change locale |
| 16:23 |
jeff |
a pox on all vendors and software that doesn't comprehend the concept of a unique/database id vs a barcode. |
| 16:23 |
berick |
also, are you using syslog? if so, it might help to temporarily use a local log file for apache. i've seen oddness w/ apache+syslog |
| 16:23 |
paxed |
phasefx: i am aware of that, believe you me. |
| 16:23 |
dbwells |
jeff: MeLCat related? :) |
| 16:23 |
jeff |
indeed. |
| 16:24 |
berick |
paxed: and in windows it still shows lang fi_fi in the template source? |
| 16:24 |
paxed |
berick: yes. |
| 16:24 |
berick |
that's so freaky |
| 16:24 |
jeff |
NCIP acknowledges that the Visible ID and the Unique ID are or can be distinct values. |
| 16:24 |
berick |
because that's the locale used to load the template files |
| 16:25 |
paxed |
berick: i'm getting the successfull requests in /etc/apache2/other_vhosts_access.log - error.log shows nothing. |
| 16:25 |
berick |
paxed: not access log, eror log |
| 16:25 |
berick |
they're usually different |
| 16:25 |
berick |
oh, maybe you meant both, sorry |
| 16:25 |
paxed |
-rw-r----- 1 root adm 0 elo 4 06:25 error.log |
| 16:25 |
berick |
ok, yeah, that should pretty much never be empty |
| 16:25 |
berick |
if it's being used |
| 16:26 |
berick |
error logs must be going somewhere else |
| 16:26 |
jeff |
Unfortunately, testing with this implementation shows that the mandatory UniqueUserId/UserIdentifierValue is discarded and only the value of UserOptionalFields/VisibleUserId/VisibleUserIdentifier is stored, and used for subsequent requests (in the UniqueUserId/UserIdentifierValue field, of course) |
| 16:26 |
berick |
also, to confirm, the template source in the linux staff client shows the same lang/xml:lang of fi_fi ? |
| 16:27 |
paxed |
berick: yes |
| 16:27 |
berick |
crazy |
| 16:28 |
berick |
jeff: ncip is like trying to speak 2 different lanuages via morse code |
| 16:29 |
paxed |
berick: urgh, yes, eg errors go to syslog apparently |
| 16:29 |
berick |
i suppose ncip is not alone in that regard, though |
| 16:29 |
jeff |
no. |
| 16:29 |
* paxed |
wonders whose idea that was... |
| 16:30 |
phasefx |
paxed: with your win client, it defaulted to Finnish; everything looks correct except the patron editor. When I view source, I see lang='fi_fi' xml:lang='fi_fi', and some obviously Finnish stuff like the title and some text in the hidden oils login template |
| 16:31 |
berick |
s/2 different languages/1 slightly different language for every ils implementation/ |
| 16:32 |
csharp |
"standards"-- |
| 16:32 |
csharp |
karma "standards" |
| 16:32 |
csharp |
@karma "standards" |
| 16:32 |
pinesol_green |
csharp: Karma for "standards" has been increased 0 times and decreased 2 times for a total karma of -2. |
| 16:32 |
paxed |
phasefx: yes, i see the same thing. (and the idl labels are in english for you too...?) |
| 16:33 |
dbwells |
jeff: That's exactly the problem we hit back in April, fwiw. Still waiting for an "update" on that... |
| 16:33 |
paxed |
so maybe it is just the windows client. |
| 16:33 |
jeff |
if ($conf{broker_uses_patron_barcode_for_uniqueuserid}) :P |
| 16:34 |
jeff |
dbwells: in your environment, are you able to use patron barcodes, and it was just a few places in the ncip responder cgi that were confused with regard to barcode vs id? |
| 16:35 |
jeff |
dbwells: or are you hoping/trying to use database ids as the central broker identifier, either because your barcodes are not able to be sent to the state library system, or because you'd rather not deal with the long involved merge process when someone gets a new barcode? |
| 16:35 |
paxed |
phasefx: interesting though that FF shows the page in Finnish, but the login texts in english, and the page lang/xml:lang = en_us ... |
| 16:37 |
dbwells |
jeff: right, the second. We didn't want users getting a new account whenever they got a new barcode. If they would just "believe" the UniqueUserId we sent them, all would be fine. |
| 16:38 |
* tsbere |
has at least one vendor he wants to use unique IDs from evergreen but can never seem to get anyone willing to talk to him about it - They assume that barcodes never change or something. |
| 16:39 |
dbwells |
jeff: I think we even tested sending back nothing *but* the ID in the the lookupuser response, but it didn't matter; whatever the patron typed in to get looked up, that was their UniqueId for the rest of the interaction (based on my 5 months old memories, at least). |
| 16:41 |
paxed |
berick: |
| 16:41 |
paxed |
Sep 10 23:41:23 egdev apache2[31157]: [debug] EGWeb.pm(77): [client 62.240.71.4] egweb: messages locale = fi_fi |
| 16:41 |
paxed |
Sep 10 23:41:24 egdev apache2[31166]: [debug] IDL2js.pm(79): [client 62.240.71.4] Invalid IDL2js locale ''; using en-US |
| 16:42 |
jeff |
dbwells: yeah. i'm testing a little more, but i've seen nothing that would contradict that. |
| 16:43 |
berick |
paxed: which test does that match? |
| 16:43 |
berick |
which client/locale/etc, i mean |
| 16:44 |
paxed |
berick: windows client, finnish. and shows english, of course. |
| 16:44 |
berick |
paxed: hm, did you remove my working patch? |
| 16:47 |
phasefx |
if I change the URL for the frame containing the patron editor to /IDL2js, I get a decidedly English set of javascript back |
| 16:48 |
berick |
phasefx: change it to /IDL2js?locale=fi-FI |
| 16:48 |
paxed |
berick: ok, i added your change back now. |
| 16:48 |
berick |
paxed: cool, thanks |
| 16:48 |
phasefx |
berick: that gives me Finnish |
| 16:48 |
berick |
phasefx: try now w/ the patch applied |
| 16:48 |
berick |
the reg page, i mean |
| 16:50 |
phasefx |
Finnish througha nd through |
| 16:50 |
phasefx |
well, except for the User Settings |
| 16:52 |
phasefx |
and the word Example: next to some of the date widgets |
| 16:52 |
phasefx |
and a few other strings |
| 16:52 |
berick |
phasefx: buttons along top/right are english, too, i bet? |
| 16:52 |
berick |
save, save clone |
| 16:52 |
phasefx |
yeah, IDL is Finnish, Save, Generate Password, etc. are not |
| 16:52 |
paxed |
berick: everything on the patron reg page has been translated |
| 16:52 |
berick |
paxed: indeed. i can see it all in FF |
| 16:52 |
berick |
so, how does EGWeb.pm have the right locale, but serves up the wrong template? |
| 16:52 |
berick |
that's confuzzling |
| 16:52 |
phasefx |
no proxies or anything? |
| 16:57 |
phasefx |
paxed: berick: the columns in patron search are powered by the IDL. In the column picker, they're not getting translated. The Finnish you do see there is being prefixed by the javascript |
| 16:58 |
phasefx |
for Mailing Address and Billing Address |
| 16:59 |
berick |
yeah, i'm thinking my idl patch is just a partial crutch. at a different level, something's borked at a slightly different level |
| 16:59 |
berick |
typing-- |
| 17:00 |
paxed |
well, it's midnight. |
| 17:00 |
paxed |
i'm off to bed... |
| 17:00 |
berick |
paxed: i'll pull up my windows vm tomorrow, see if I can't give it a poke |
| 17:02 |
paxed |
berick: thanks. i'm sure i can survive few more days ... |
| 17:03 |
|
jeff joined #evergreen |
| 17:03 |
|
jeff joined #evergreen |
| 17:04 |
phasefx |
there are horrible kludges available if things get dire |
| 17:07 |
|
mllewellyn left #evergreen |
| 17:15 |
|
mmorgan left #evergreen |
| 17:16 |
csharp |
@eightball by "kludges" does phasefx mean "windows"? |
| 17:16 |
pinesol_green |
csharp: Maybe... |
| 17:39 |
|
dboyle joined #evergreen |
| 17:47 |
|
stevenyvr2 joined #evergreen |
| 17:53 |
|
ldwhalen joined #evergreen |
| 18:09 |
|
akilsdonk_ joined #evergreen |
| 18:16 |
|
ldwhalen left #evergreen |
| 18:17 |
|
ldwhalen_ joined #evergreen |
| 18:19 |
|
ldwhalen_ joined #evergreen |
| 19:02 |
|
smyers__ joined #evergreen |
| 20:08 |
|
stevenyvr2 left #evergreen |
| 21:04 |
|
kmlussier joined #evergreen |
| 21:28 |
bshum |
@roulette |
| 21:28 |
pinesol_green |
bshum: *click* |
| 21:28 |
kmlussier |
bshum: Good evening! |
| 23:06 |
|
stevenyvr2 joined #evergreen |
| 23:06 |
|
stevenyvr2 left #evergreen |
