| Time |
Nick |
Message |
| 06:02 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 07:12 |
|
rjackson_isl joined #evergreen |
| 08:38 |
|
mmorgan joined #evergreen |
| 08:43 |
|
Dyrcona joined #evergreen |
| 08:46 |
|
mantis joined #evergreen |
| 09:04 |
|
rfrasur joined #evergreen |
| 09:17 |
|
dguarrac joined #evergreen |
| 09:25 |
|
stephengwills joined #evergreen |
| 09:45 |
|
mmorgan left #evergreen |
| 09:48 |
stephengwills |
where is wsid and ws_ou supposed to be getting created for a patron? |
| 09:51 |
Dyrcona |
stephengwills: I don't they are supposed to be created for a patron. |
| 09:52 |
stephengwills |
well account.pm wants them…. that is why my 3.7 TPAC is blowing up |
| 09:53 |
stephengwills |
I’m going to wrap it in an if defines and move on. |
| 10:03 |
JBoyer |
stephengwills, I'm not sure those lines are causing the ISE; they would just drop some (annoying) warnings in the logs; I would suspect that the retrieve_actor_user call is failing to return a user object (hard to say why though) and the return ... unless at line 54 is what's complaining. |
| 10:04 |
JBoyer |
Commenting that out will stop it from returning an ISE at that point, but I expect more things would break / be broken in that case. |
| 10:07 |
JBoyer |
git blame shows most of that function (including the wsid and ws_ou lines) at 2013, so something else is up. |
| 10:12 |
mantis |
A library reported that their receipt printer coughs up a notification each time a receipt is printed. "Epson TM-T90 not installed". They don't have Hatch installed on their computers. To me it seems more like a hardware problem than an Evergreen problem. |
| 10:13 |
JBoyer |
mantis, sounds like a "open the control panel and see what's complaining" type of problem, yeah. Epson's printers are great, their installers are ... less so. |
| 10:14 |
mantis |
JBoyer: great idea, we'll see what's under the hood there |
| 10:14 |
mantis |
JBoyer++ |
| 10:14 |
mantis |
Personally, I think the Star printers are the worst |
| 10:32 |
|
abowling joined #evergreen |
| 10:46 |
stephengwills |
so it is an error, not a warning. [perl:error] [pid 1302] [client 74.78.7.223:58467] egweb: Context Loader error: Can't call method "ws_ou" on an undefined value at /usr/local/share/perl/5.26.1/OpenILS/WWW/EGCatLoader/Account.pm line 50. I wrapped the wsid in a if defined and now the ws_ou is barfing…I agree this is because retrieve_actor_user isn’t defining wsid or ws_ou for a patron, which, I guess makes sense. I’ll wrap |
| 10:46 |
stephengwills |
ws_ou as well and see what else gets bent out of shape |
| 10:47 |
stephengwills |
but the correct fix is proably for retrieve actor user to hand those back in the usr object. |
| 10:48 |
stephengwills |
actually i’m gonna if define else make em so |
| 10:55 |
berick |
stephengwills: what JBoyer said. the lack of a wsid or ws_ou is not the issue. one of the user retrieval calls, possibly the one a few lines up, is not returning a value. |
| 10:58 |
berick |
you should see calls in the logs for the open-ils.cstore.direct.actor.user.retrieve API which are not returning values. |
| 10:58 |
berick |
the logs will show what ID it's trying to fetch |
| 11:25 |
|
collum joined #evergreen |
| 11:43 |
jeff |
stephengwills: When you have an expression like "$usr->ws_ou" that's throwing a message like "Can't call method "ws_ou" on an undefined value", it is $usr that is undefined. |
| 11:50 |
|
jihpringle joined #evergreen |
| 11:56 |
stephengwills |
so, on line 30 fo Account.pm where it makes a backup of the user object there may not be one yet? |
| 11:56 |
stephengwills |
s/fo/of/ |
| 12:03 |
jeff |
seems likely. |
| 12:03 |
jeff |
$self->ctx->{user} may be undefined for unknown reasons at the point where that line defiens $usr: my $usr = $self->ctx->{user}; |
| 12:05 |
jeff |
$ctx->{user} should be getting set in OpenILS::WWW::EGCatLoader::load_common |
| 12:05 |
jeff |
I think that attempting a fresh login and consulting logs would be a good next step. Is this happening to all users on a test system? |
| 12:11 |
stephengwills |
i’ll tst with multiple users |
| 12:13 |
jeff |
how did the problem come to your attention? |
| 12:13 |
jeff |
ah, I see the "I just upgraded from 3.2.2 to 3.7.1" earlier in channel. |
| 12:14 |
stephengwills |
nod. sorry phone call here |
| 12:26 |
stephengwills |
well that is wierder. it doesn’t crash for my regular patron. |
| 13:03 |
|
mmorgan joined #evergreen |
| 13:23 |
|
collum joined #evergreen |
| 13:24 |
|
nfBurton joined #evergreen |
| 13:28 |
jeff |
stephengwills: if you find out what's different between the user it affects and the user it doesn't, I'd be curious to hear. |
| 13:28 |
stephengwills |
i’m working on it….I have three users now two crash, one doesn’t |
| 13:53 |
stephengwills |
the two that are causing the crash appear to be expired patron who were, never the less able to log onto the opac but crash looking at myprefs. |
| 13:53 |
stephengwills |
renewing on now |
| 14:00 |
|
nfBurton joined #evergreen |
| 14:08 |
|
alynn26 joined #evergreen |
| 14:17 |
stephengwills |
nope, no joy |
| 14:30 |
jeff |
i'd recommend reviewing logs for more clues. |
| 14:31 |
jeff |
keeping in mind that the last error is often not as helpful as the errors/warnings leading up to it. |
| 14:32 |
jeff |
having a user that doesn't cause the error is handy, because you can compare logs for the two users, and rule out certain things, like a warning that happens with both. |
| 14:47 |
|
jihpringle joined #evergreen |
| 14:48 |
stephengwills |
nod. so far the usr objects didn’t reveal anything useful |
| 14:49 |
stephengwills |
but I’m still flogging it… it’ll get up and run yet! |
| 15:06 |
stephengwills |
column aoa.latitude does not exist? |
| 15:10 |
stephengwills |
:q! |
| 15:10 |
jihpringle |
stephengwills: I'd guess it's related to the new Sort Holdings by Geographical Proximity feature in 3.7 (simply based on it fact that it references latitude) |
| 15:11 |
stephengwills |
i think i’ve missed some db updates here. I have a few missing colum errors and maybe that is why it can’t flesh out some users |
| 15:18 |
jeff |
if your IDL and your database schema are not in agreement, all kinds of fun things can happen. |
| 15:18 |
Dyrcona |
Yeah, that will do it. |
| 15:22 |
jeff |
actor.usr difference between 3.2 and 3.7 would include the addition of the "guardian" column. |
| 15:23 |
jeff |
also the addition of the user privacy waiver, but I'm not sure if that would trip you up in this way on OPAC login. |
| 15:36 |
stephengwills |
is there a tool that lets me check my database against my IDL? |
| 15:36 |
stephengwills |
that would be a horrible thing to have to do my hand |
| 15:36 |
rhamby |
not that I'm aware of but I can imagine such a thing pretty easily, use an xml library to crawl through the nodes and use dbi to check information schema |
| 15:37 |
stephengwills |
nod. |
| 15:51 |
stephengwills |
my apologies for useing so much bandwidth, my 3.6.2-3.7.0-upgrade-db.sql changes got rolled back to a missing String::KeyboardDistance module |
| 16:03 |
Dyrcona |
stephengwills: Yeah, there are some new prerequisites for 3.7. |
| 17:19 |
|
mmorgan left #evergreen |
| 18:00 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
