| Time |
Nick |
Message |
| 06:02 |
pinesol_green |
News from qatests: Test Success <http://testing.evergreen-ils.org/~live> |
| 08:06 |
|
jvwoolf joined #evergreen |
| 08:10 |
|
jvwoolf1 joined #evergreen |
| 08:16 |
|
remingtron joined #evergreen |
| 08:31 |
|
collum joined #evergreen |
| 08:42 |
rhamby |
csharp++ : for being the mailing list maestro |
| 08:49 |
|
kmlussier joined #evergreen |
| 08:49 |
|
dbwells joined #evergreen |
| 08:49 |
* csharp |
taps baton on music stand |
| 08:58 |
|
Dyrcona joined #evergreen |
| 09:00 |
csharp |
ok, back to working on bug 1724321... |
| 09:00 |
pinesol_green |
Launchpad bug 1724321 in Evergreen "Web Client -- existence of record with duplicate TCN does not prevent import" [High,New] https://launchpad.net/bugs/1724321 |
| 09:01 |
|
bos20k joined #evergreen |
| 09:01 |
csharp |
in Open-ILS/web/js/ui/default/staff/cat/z3950/app.js, it looks like I could theoretically call $scope.import() within $scope.spawn_editor, but I'm not sure that would work the way I want it to |
| 09:03 |
csharp |
something along the lines of '$scope.save = $scope.import()', but I'm pretty sure $scope in that context means "within the array within this function" |
| 09:04 |
csharp |
the point is, the code in open-ils.cat.biblio.record.xml.import needs to be run when clicking "Import" within the marc editor in this context |
| 09:05 |
csharp |
probably better to break the import code off into its own function, then have existing $scope.import and $scope.spawn_editor use that |
| 09:06 |
kmlussier |
csharp++ |
| 09:06 |
csharp |
I know conceptually what I want to do, but my JS/angularjs foo is not well-developed yet |
| 09:10 |
|
yboston joined #evergreen |
| 09:51 |
|
jvwoolf joined #evergreen |
| 10:08 |
|
Christineb joined #evergreen |
| 10:28 |
|
Jillianne joined #evergreen |
| 10:56 |
|
jvwoolf joined #evergreen |
| 11:12 |
|
collum_ joined #evergreen |
| 11:39 |
Bmagic |
@coffee |
| 11:39 |
* pinesol_green |
brews and pours a cup of NOBLE Coffee, and sends it sliding down the bar to Bmagic |
| 11:40 |
bshum |
Oooh the good stuff |
| 11:40 |
kmlussier |
Ooh! I can attest to the fact that NOBLE makes good coffee. |
| 11:41 |
berick |
to date, the only coffee to win the NOBLE prize in coffee |
| 11:41 |
berick |
(wakka wakka) |
| 11:41 |
Dyrcona |
:) |
| 11:50 |
|
_adb joined #evergreen |
| 12:04 |
Bmagic |
haha |
| 12:06 |
Bmagic |
berick: can more than one org unit have the same san? |
| 12:07 |
Bmagic |
I have the same SAN for the same org unit but on two different addresses |
| 12:12 |
Bmagic |
I am trouble shooting "create_acq_invoice_from_edi(..., <acq.edi_message #XXXX>): unable to determine buyer (org unit) in invoice; buyer_san=; buyer_acct=XXXXXX" |
| 12:13 |
kmlussier |
Bmagic: We have org units with providers with the same SAN |
| 12:14 |
Bmagic |
kmlussier: actor.org_address ? |
| 12:14 |
kmlussier |
Bmagic: Oh, sorry. I was thinking of SANs for providers. Ignore me. |
| 12:17 |
berick |
Bmagic: i don't think that would be a problem. the invoice code grabs the first addr with the san. |
| 12:17 |
Bmagic |
thanks! |
| 12:17 |
berick |
since they're both pointing to the same org, end result is the same regardless of which address is randomly returned from the search |
| 12:18 |
Dyrcona |
Bmagic: This looks like you have an invoice with no SAN in it. I could be wrong, but that's what it looks like from the error. |
| 12:18 |
Bmagic |
I think the issue is the NAD+BY |
| 12:18 |
Dyrcona |
Yeah, that. :) |
| 12:18 |
|
jihpringle joined #evergreen |
| 12:19 |
Bmagic |
I have hundreds of examples where the NAD+BY from the vendor is referencing their own SAN in the NAD+BY clause in the invoice EDI |
| 12:19 |
|
khuckins_ joined #evergreen |
| 12:19 |
* Dyrcona |
isn't sure of the exact field. |
| 12:22 |
* berick |
is reminded to update another edi bug |
| 12:22 |
csharp |
gmcharlt: (or others) if I have a showstopper bug (like bug 1724321, which threatens our ability to move catalogers off the xul client), should I target that to a particular series/release to get the most attention on it? |
| 12:22 |
pinesol_green |
Launchpad bug 1724321 in Evergreen "Web Client -- existence of record with duplicate TCN does not prevent import" [High,New] https://launchpad.net/bugs/1724321 |
| 12:25 |
gmcharlt |
csharp: I suggest (in addition to high priority) a tag such as webstaffblocker |
| 12:25 |
csharp |
gmcharlt: cool - thanks |
| 12:25 |
gmcharlt |
(used judiciously, of course) |
| 12:27 |
csharp |
gmcharlt: yes, definitely |
| 12:28 |
csharp |
I don't think we have anything else that's been describe to me in dire terms, so that's probably the only one |
| 12:28 |
gmcharlt |
cool |
| 12:49 |
jeff |
fun thought experiment of the afternoon: i'm not sure that there's anything in a given action.circulation row with checkin_time set that indicates that the circulation is the last in its chain. |
| 12:51 |
Bmagic |
berick: something new has come to light, a different problem, and it sounds like it's an FTP issue: ERR :24851:EDI.pm:179:] (S)FTP get(FTPSERVER//FOLDER1/FOLDER2/FOLDER1/FOLDER2\OUT141427438653.INT) failed |
| 12:51 |
berick |
jeff: stop_fines ? |
| 12:51 |
berick |
renew vs checkin |
| 12:53 |
Bmagic |
is there a significance to the folder path being repeated in the logs like that? Technically, that path doesn't exist as writtin |
| 12:53 |
csharp |
'where c.parent_circ is not null and not exists (select 1 from action.circulation p where p.parent_circ = c.id)' |
| 12:53 |
berick |
csharp: i read jeff's comment as "you can only look at the circ in question" |
| 12:54 |
jeff |
berick: stop_fines is only set to RENEW if not already set, so a circ ending in a renewal can have MAXFINES as a stop_fines reason. |
| 12:54 |
csharp |
berick: ah - good point :-) |
| 12:54 |
berick |
csharp: but yes that's the right answer :) |
| 12:54 |
jeff |
and yes, that was my constraint on the exercise (which has practical implications for a report i've just been asked for) :-) |
| 12:54 |
berick |
jeff: right... |
| 12:55 |
csharp |
Bmagic: just ruling out possibilities - are you passing "FTP_PASSIVE=1" to the pusher script? |
| 12:55 |
Bmagic |
yes |
| 12:55 |
csharp |
k |
| 12:55 |
Bmagic |
it's fetcher in this case |
| 12:55 |
csharp |
ah |
| 12:56 |
berick |
Bmagic: never seen it repeat. what's the configured path (or anonymized fascimile thereof)? |
| 12:57 |
Bmagic |
"/FOLDER1/FOLDER2/*.INT" |
| 12:57 |
berick |
Bmagic: is /FOLDER1/FOLDER2/ the base/home login directory by any chance? |
| 12:58 |
berick |
if you just ftp by hand |
| 12:58 |
Bmagic |
FTPing by hand, we are placed into FOLDER1 |
| 12:58 |
jeff |
csharp: the logic in your snippet above isn't suitable for my needs (yours appears to limit to only circs that are renewals by requiring a parent circ), but that's the direction i'm going in terms of making this work. |
| 12:59 |
berick |
Bmagic: then you probably want the path to be FOLDER2/*.INT |
| 12:59 |
Bmagic |
the configuration needs to be absolute path right? Not relative to where we are after initial connection? |
| 12:59 |
berick |
relative |
| 12:59 |
Bmagic |
oh boy, then if that is true, how in the world has this been working until now? |
| 12:59 |
Bmagic |
maybe they changed our home folder sometime in the recent past |
| 12:59 |
berick |
abs path may work in some cases. i've only ever used relative |
| 13:00 |
Bmagic |
alright, let me try relative |
| 13:00 |
berick |
that could certainly mess things up, re: path changing |
| 13:00 |
Bmagic |
perhaps absolute for pusher and relative for fetcher |
| 13:00 |
Bmagic |
path vs. in_dir |
| 13:00 |
csharp |
jeff: ah - I was led astray by inferring from "last in its chain" that the chain had more than one link ;-) |
| 13:00 |
* berick |
uses relative for both fwiw |
| 13:01 |
csharp |
we use relative paths everywhere too |
| 13:01 |
Bmagic |
berick: when using relative path do you start with "/" ? |
| 13:01 |
berick |
no |
| 13:01 |
csharp |
and we don't let end users near that UI if we can help it :-) |
| 13:34 |
csharp |
ugh - our offline block list has grown to 77MB, only including lost, barred, and penalized cards |
| 13:35 |
csharp |
it's large enough now that it's not downloading in FF or Chrome |
| 13:35 |
csharp |
or maybe we're not giving it enough time |
| 13:36 |
csharp |
in any case, looking for ways to 1) limit the number of barcodes generated, perhaps based on last edit of the user or 2) look into compressing the file for download and uncompressing on the other end (not sure about the feasibility of that) |
| 13:38 |
csharp |
zip compresses it to 23M - still not awesome |
| 13:53 |
csharp |
huh - actually, looks like we *are* including expired cards, though we don't mean to :-/ |
| 13:57 |
jeff |
do you generally use offline mode when a library is offline, or when evergreen servers are offline? |
| 13:57 |
csharp |
well, it helps if you call the script with the right parameters - apparently the syntax for those changed along the way from "L" to "L=lost" (for example) |
| 13:57 |
csharp |
jeff: both scenarios, but unplanned outages are rare on both ends nowadays |
| 13:58 |
* csharp |
knocks on wood |
| 13:59 |
berick |
csharp: apache can compress it for you in a way browsers understand. mod_deflate |
| 13:59 |
jeff |
we've mused on this before, but if we used something like a bloom filter plus a small network service to handle "maybe" checks, it would probably reduce the size even more, but it sounds like you may have already solved it by excluding expired cards. |
| 14:00 |
csharp |
berick: don't we still disable mod_deflate by default? |
| 14:02 |
csharp |
yeah, we do |
| 14:02 |
csharp |
it's in DEB_APACHE_DISMODS in Makefile.install |
| 14:03 |
csharp |
though eg_vhost.conf appears to anticipate it being there |
| 14:04 |
csharp |
t'would be handy to have enabled in this instance though |
| 14:05 |
berick |
yeah, doesn't mean you can't use it. i enable it locally by default |
| 14:07 |
csharp |
good |
| 14:07 |
csharp |
probably worth re-(re-?)visiting whether anyone thinks it's still necessary to disable |
| 14:08 |
csharp |
good one-off hackaway item |
| 14:11 |
berick |
csharp: and of course https://www.nginx.com/resources/admin-guide/compression-and-decompression/ |
| 14:11 |
berick |
may be easier to dial that in, since the stock EG configs have gzip stuff that will just magically start happening as soon as you turn it on |
| 14:12 |
|
rlefaive joined #evergreen |
| 14:15 |
miker |
also, no point in doing compression in apache if you've got nginx or pound or something in front of of apache for LB |
| 14:24 |
csharp |
berick++ # nginx docs |
| 14:24 |
csharp |
that looks like the way to go |
| 14:25 |
csharp |
(still prolly worth discussing mod_deflate at the hackaway tho) |
| 14:45 |
|
StomproJosh joined #evergreen |
| 15:28 |
csharp |
ok - even after filtering out expired cards and enabling gzip in nginx, the page keeps timing out with "Page Unresponsive" dialogs in Chrome when I try to download the offline blocked list |
| 15:28 |
csharp |
can others confirm that they are able to download from their sites? |
| 15:28 |
* csharp |
has to run to an appointment soon so may disappear |
| 15:31 |
|
rlefaive joined #evergreen |
| 17:34 |
|
jvwoolf left #evergreen |
| 18:00 |
pinesol_green |
News from qatests: Test Success <http://testing.evergreen-ils.org/~live> |
| 18:04 |
|
sandbergja joined #evergreen |
| 19:57 |
|
rlefaive joined #evergreen |
| 21:24 |
|
rlefaive joined #evergreen |
