| Time |
Nick |
Message |
| 00:01 |
|
stevenyvr2 joined #evergreen |
| 00:01 |
|
stevenyvr2 left #evergreen |
| 01:09 |
|
adbowling-isl joined #evergreen |
| 01:10 |
|
zerick joined #evergreen |
| 01:19 |
|
adbowling-isl joined #evergreen |
| 07:16 |
|
Dyrcona joined #evergreen |
| 07:18 |
|
Dyrcona joined #evergreen |
| 07:49 |
|
collum joined #evergreen |
| 08:02 |
|
jboyer-isl joined #evergreen |
| 08:12 |
|
kmlussier joined #evergreen |
| 08:35 |
|
mrpeters joined #evergreen |
| 08:35 |
|
ericar joined #evergreen |
| 08:42 |
|
krvmga joined #evergreen |
| 08:44 |
krvmga |
in our catalog bark.cwmars.org our search format filters show up in a different order than either their sort position or ID. does anyone know why? |
| 08:46 |
kmlussier |
krvmga: Did you verify that you have the fix from https://bugs.launchpad.net/evergreen/+bug/1116549? |
| 08:46 |
pinesol_green |
Launchpad bug 1116549 in Evergreen 2.3 "Search filter group sorts alphanumerically by sort position instead of numerically" (affected: 1, heat: 6) [Undecided,Fix released] |
| 08:50 |
krvmga |
yes, i verified that i applied the fix. |
| 08:53 |
kmlussier |
krvmga: I don't know then. I know NOBLE hasn't encountered the same problem with their search filter groups. |
| 08:54 |
kmlussier |
Sorry I can't be more help. :( |
| 08:54 |
paxed |
we haven't seen that since that fix went in |
| 08:55 |
krvmga |
i made a picture showing the search format filters as they appear in the staff client and as they appear in the basic search dropdown. would anyone like to see it? maybe you'll see something i can't. |
| 08:58 |
krvmga |
brb |
| 08:58 |
kmlussier |
krvmga: Here's a thought. You're have filter_group_selector.tt2 in your custom directory. But I'm guessing you haven't made any local customizations to that file. Have you considered deleting that custom file and just using the one that is in the main templates directory? |
| 08:58 |
kmlussier |
Maybe there was some other change made to that file that you aren't seeing because it wasn't applied to the custom file. |
| 08:58 |
|
krvmga left #evergreen |
| 08:59 |
|
krvmga_ joined #evergreen |
| 09:00 |
csharp |
krvmga_: please share your image if you can |
| 09:01 |
krvmga_ |
just for information, i don't have a custom coded_value_selector.tt2 file. i use the one that's in the templates/opac/parts directory |
| 09:04 |
krvmga_ |
http://www.randompractice.com/images/format_filters.png |
| 09:05 |
krvmga_ |
randompractice.com is just a play area that i set up for myself |
| 09:05 |
krvmga_ |
where i can build and destroy things without having to worry about affecting anyone |
| 09:06 |
|
kivilahtio joined #evergreen |
| 09:12 |
krvmga_ |
okay, i double-checked and i thought i had removed any custom format filter files from our templates director (because the fix had been made in the main templates directory) but i hadn't |
| 09:12 |
krvmga_ |
problem solved |
| 09:12 |
krvmga_ |
sorry for the bother |
| 09:13 |
kmlussier |
krvmga_: No bother. Happy to see it solved! :) |
| 09:20 |
kivilahtio |
Hi! Do you have any experience in running Evergreen only via https and from a non-standard port, like :444. |
| 09:21 |
Dyrcona |
kivilahtio: Staff client and/or tpac? |
| 09:21 |
kivilahtio |
WE have some trouble running a staff_client -only version of evergreen in port :444. Here: 62.148.106.91:444 |
| 09:21 |
kivilahtio |
Dyrcona: only staff_client. |
| 09:22 |
Dyrcona |
staff client should be doing only https anyway. |
| 09:22 |
jeff |
it uses a mix. |
| 09:22 |
kivilahtio |
We are running the default Evergreen installation with default apache2 configurations. Only the Evergreen configurations server is accessible to port :444 only |
| 09:22 |
Dyrcona |
Why am I not surprised. |
| 09:22 |
* jeff |
considers re-visiting a branch that addresses both that and kivilahtio's needs |
| 09:22 |
kivilahtio |
I am looking into the issue and I realized that data_cache in the staff_client gives a worng server parameter |
| 09:23 |
kivilahtio |
D_DATA_RETRIEVE :> Retrieving server : http://10.1.63.89:444 |
| 09:23 |
kivilahtio |
should be |
| 09:23 |
kivilahtio |
D_DATA_RETRIEVE :> Retrieving server : https://10.1.63.89:444 |
| 09:23 |
kivilahtio |
so the staff_client data_caching component assumes http, even if https is required |
| 09:24 |
kivilahtio |
you are free to try that url with the latest evergreen master staff client |
| 09:25 |
kivilahtio |
my guestion is, where does the staff_client get these values? |
| 09:27 |
kivilahtio |
this assumption is pretty strange considering I have always thought staff_client is strictly https |
| 09:27 |
kivilahtio |
which surprise surprise it is not |
| 09:27 |
Dyrcona |
kivilahtio: I'd call it a bug. |
| 09:28 |
kivilahtio |
Dyrcona: Also this xulrunner and it's javascript components are a bit complex |
| 09:28 |
Dyrcona |
Only "a bit?" :) |
| 09:28 |
kivilahtio |
I mean was there really no easier way to achieve these things? |
| 09:28 |
kivilahtio |
well maybe back then when Eg first started there wasn't |
| 09:28 |
Dyrcona |
tsbere mentioned replacing it with Java so we could have a core that would work on PC and Android. |
| 09:29 |
Dyrcona |
And tsbere is not a fan of Java. |
| 09:29 |
kivilahtio |
Even Perl would be more understandable |
| 09:29 |
kivilahtio |
but JAva works for me :) |
| 09:30 |
phasefx |
data.server is set in main.js |
| 09:30 |
phasefx |
on the chrome/ side |
| 09:30 |
kivilahtio |
Why not use Template Toolkit for the new staff client? |
| 09:30 |
kivilahtio |
I mean defeinetely we don't need any more varied technologies added to this evergreen mix |
| 09:30 |
Dyrcona |
Well, someone (not me) needs to come up with a plan for a new staff client and get people involved and working on it. |
| 09:31 |
phasefx |
almost any direction is viable |
| 09:31 |
eeevil |
grabbing 0824 |
| 09:31 |
kivilahtio |
Dyrcona: I hear you. I would love to take up the challenge, but unfortunately we are a bit tied up here |
| 09:31 |
Dyrcona |
kivilahtio: I say not me because I'm too busy, but not too busy for IRC, apparently. ;) |
| 09:32 |
* Dyrcona |
agrees with phasefx. |
| 09:32 |
kivilahtio |
phasefx: thanks |
| 09:33 |
phasefx |
kivilahtio: so line 398 is the culprit |
| 09:33 |
kivilahtio |
Dyrcona: phasefx: Quite frankly I feel like this xulrunner issue wouldnt be so bad if I just sat down and read a good book about xulrunner |
| 09:33 |
phasefx |
I think the thought at the time that only sensitive data need go through SSL |
| 09:33 |
Dyrcona |
kivilahtio: Many components of the staff client would need serious modification to work with current XulRunner releases. |
| 09:34 |
Dyrcona |
A couple of the components we use have been removed in recent versions. |
| 09:34 |
Dyrcona |
phasefx and dbs know more of the details, I'm sure. |
| 09:34 |
phasefx |
kivilahtio: xulrunner promised to be something others could build on, but in practice, its developers just focus on the needs of Firefox |
| 09:34 |
kivilahtio |
phasefx: are you sure this is the culprit? I am getting some https urls as well, but I'll try |
| 09:34 |
phasefx |
kivilahtio: that's the culprit for the log output you pasted |
| 09:35 |
phasefx |
not saying it will magically encrypt everything, it won't |
| 09:35 |
kivilahtio |
ok ok |
| 09:35 |
phasefx |
in constants.js, there is a security flag next to each API call that wants to be encrypted |
| 09:35 |
phasefx |
you could hack network.js to ignore that |
| 09:36 |
* paxed |
suggests kivilahtio files a bug ;) |
| 09:37 |
kivilahtio |
* kivilahtio is too curious |
| 09:37 |
pinesol_green |
[evergreen|Bill Erickson] LP1209291 vandelay item import defaults - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=0b8e167> |
| 09:37 |
pinesol_green |
[evergreen|Bill Erickson] LP1209291 vandelay item defaults release notes - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=db143bf> |
| 09:37 |
pinesol_green |
[evergreen|Mike Rylander] Stamping upgrade script for Vandelay item import defaults - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=2f2fccb> |
| 09:38 |
paxed |
kivilahtio: then you'll be able to submit a fix for it too :P |
| 09:38 |
phasefx |
if we do end up encrypting everything, just as the xul files themselves, I'm not sure if caching will still work. I could imagine it pulling down, say, checkout.xul on every load of that interface |
| 09:38 |
phasefx |
s/just/such/ |
| 09:39 |
phasefx |
client would be very slow if that happens |
| 09:39 |
kivilahtio |
So there is no way to make this work even if I opened, say port :82 for Evergreen= |
| 09:39 |
kivilahtio |
? |
| 09:39 |
Dyrcona |
On a new staff client: I think everyone agrees that we need one. However, no one seems to have the zots to do it right now. This is why I made the $50 000 comment in the list email the other day. |
| 09:40 |
kivilahtio |
Staff client needs to be able to access port :80? |
| 09:40 |
phasefx |
kivilahtio: personally, the tact I would take is to use ssh tunelling to map your ports |
| 09:40 |
phasefx |
set your hostname to localhost, but your tunnel to whatever |
| 09:40 |
Dyrcona |
ssh_tunneling++ |
| 09:41 |
kivilahtio |
staff client expects https for 443 and http for 80 |
| 09:41 |
kivilahtio |
can you be more specific? |
| 09:41 |
kivilahtio |
well |
| 09:42 |
kivilahtio |
I would like to stress out that ouor port :80 is already taken for another development version of eg |
| 09:43 |
kivilahtio |
What I am trying to achieve is to have a development EG in :80 & :443, then a configuration baseline eg in :444 |
| 09:43 |
Dyrcona |
kivilahtio: You'd want to tunnel localhost ports 80 and 443 to ports on your server. You could send 80 to 82, for instance. |
| 09:43 |
Dyrcona |
you then tell Evergreen it server is localhost. |
| 09:44 |
Dyrcona |
If you want to do this on a LAN, you setup the ssh tunnel on one computer on the LAN, and then tell Evergreen staff clients to use that computer as the server. |
| 09:44 |
kivilahtio |
i cant get to localhost via :80 |
| 09:44 |
Dyrcona |
you can if you run a ssh tunnel that listens on localhost 80 and tunnels to your actual server. |
| 09:45 |
kivilahtio |
ah on the staff client machine! |
| 09:45 |
kivilahtio |
now i get it |
| 09:45 |
Dyrcona |
Yeah. |
| 09:46 |
Dyrcona |
Here's something I do with OpenSSH on my laptop to access the file server from outside the firewall: |
| 09:46 |
kivilahtio |
so everybody who wants to access the configuration evergreen needs to define a tunnel besides their staff client |
| 09:46 |
kivilahtio |
well it would work for me ;) |
| 09:47 |
kivilahtio |
quite frankly I am quite disappointed about this as this totally demolishes the server architecture I had in mind |
| 09:47 |
Dyrcona |
You can configure the tunnel on a single PC on a LAN, and then everyone else uses that PC as their "server" in Evergreen. |
| 09:47 |
kivilahtio |
all out port :80's are taken |
| 09:47 |
Dyrcona |
We do this with SIP. |
| 09:47 |
Dyrcona |
Ports don't have to be the same. |
| 09:47 |
Dyrcona |
You can redirect :80 on one computer to any port on another as long as the destination port is open. |
| 09:48 |
Dyrcona |
Fer instance, I could "fake" having Apache on my workstation running on 8080 with something like this (assumin OpenSSH): |
| 09:49 |
phasefx |
I wonder if you can you use virtual hosting to manage multiple Evergreen instances with one apache. Probably not; I'd expect there is stuff loaded outside of the vhost stuff |
| 09:49 |
Dyrcona |
ssh -N -f -L 8080:webserver.address:80 |
| 09:49 |
kivilahtio |
Dyrcona: I know, but we have 3 public IP's and all of them already have :80 taken |
| 09:49 |
kivilahtio |
or du you mean I could share port :80 with several apps? |
| 09:49 |
kivilahtio |
I think I could do a redirecting apache2 virtual host definition |
| 09:50 |
phasefx |
kivilahtio: I think Dyrcona is saying your server would have an apache listening on 8080 |
| 09:50 |
Dyrcona |
phasefx++ |
| 09:50 |
kivilahtio |
say 10.1.63.89/egconfig -> egconfig:80 |
| 09:50 |
Dyrcona |
You host Evergreen for remote sites, right? |
| 09:51 |
phasefx |
kivilahtio: you can have localhost 80 tunnel to server 8080, and localhost 443 tunnel to server 8443 |
| 09:51 |
kivilahtio |
yes |
| 09:51 |
Dyrcona |
At the remote site setup the tunnel to send 80 to egserver:8080 or whatever. |
| 09:51 |
kivilahtio |
phasefx: but then we need to configure tunneling for each staff_client installation |
| 09:51 |
Dyrcona |
Just once per site. |
| 09:52 |
kivilahtio |
but the issue is that staff_client doesnt work on other ports than 80 |
| 09:52 |
Dyrcona |
The real issue is we need a new staff client, if you want to get down to it. |
| 09:52 |
kivilahtio |
if i try to connect with staff client to 10.1.63.89:444, it looks for the http and https service at the same port |
| 09:52 |
phasefx |
kivilahtio: from the staff client perspective, it'll be using ports 80/443 as normal |
| 09:52 |
Dyrcona |
I'm trying to help you with the immediate problem. |
| 09:52 |
kivilahtio |
ok ok |
| 09:52 |
kivilahtio |
I think I understand |
| 09:52 |
kivilahtio |
lets skip this topic |
| 09:53 |
kivilahtio |
:) |
| 09:53 |
kivilahtio |
I feel I am wasting your itme |
| 09:53 |
|
tspindler joined #evergreen |
| 09:53 |
Dyrcona |
ok. If you're sure you understand. |
| 09:53 |
kivilahtio |
:) |
| 09:54 |
Dyrcona |
I just want to add that we use ssh tunneling for sip at our libraries. |
| 09:54 |
kivilahtio |
Dyrcona: any idea how much time programming a new staff client would take? |
| 09:54 |
kivilahtio |
I feel it would be something like 1-2 years= |
| 09:54 |
kivilahtio |
maybe 1 year |
| 09:54 |
Dyrcona |
They set up the tunnel on 1 computer and the sip clients (usually 2) use that computer as their sip server. |
| 09:55 |
Dyrcona |
I usually think in terms of hours. |
| 09:55 |
kivilahtio |
2400h ? ;) |
| 09:55 |
Dyrcona |
4 000 to 8 000 on the outside. |
| 09:55 |
|
tspindler left #evergreen |
| 09:55 |
Dyrcona |
Maybe less than 4 000. |
| 09:56 |
kivilahtio |
If we did it with Ruby it would be 3 weeks |
| 09:56 |
kivilahtio |
;) |
| 09:56 |
Dyrcona |
But, I've not really sat down and tried to figure it out. |
| 09:56 |
Dyrcona |
heh. |
| 09:56 |
kivilahtio |
just kidding |
| 09:56 |
Dyrcona |
:) |
| 09:56 |
kivilahtio |
Grails is the way to go |
| 09:57 |
kivilahtio |
O_o |
| 09:57 |
Dyrcona |
And it would be 80 lines of Lisp that no one would understand. :) |
| 09:57 |
kivilahtio |
FRIDAY! |
| 09:57 |
kivilahtio |
so true |
| 09:57 |
kivilahtio |
maybe god could create that ? |
| 09:57 |
phasefx |
for some things, there are alternatives now. You can do simple circ through a web interface |
| 09:58 |
phasefx |
paying bills or cataloging, not so much |
| 09:58 |
Dyrcona |
I'm all for whatever as long as it mostly works. |
| 09:59 |
phasefx |
and some of the interfaces in the staff client can run outside of the client, if not in a normal browser, then at least in firefox |
| 09:59 |
kivilahtio |
then you chose the right ILS! |
| 09:59 |
Dyrcona |
ha! |
| 09:59 |
kivilahtio |
what I am curious of is the offline capabilities of the new staff client |
| 09:59 |
phasefx |
Dyrcona: let's do a skunk works and polish your text based interface |
| 10:00 |
kivilahtio |
phasefx: I think a text based interface would be awesome |
| 10:00 |
kivilahtio |
nice green background, black test |
| 10:00 |
kivilahtio |
add some flicker just for nostalgy's sake |
| 10:00 |
Dyrcona |
phasefx: go for it. i'm a bit tied up ATM. |
| 10:00 |
Dyrcona |
curses! |
| 10:00 |
* phasefx |
would spice it up with color and glyphs for drawing borders, etc. |
| 10:00 |
kivilahtio |
but raelly a text-based interface! Fast, furious, kickass effective |
| 10:01 |
Dyrcona |
ANSI terminal codes. |
| 10:01 |
kivilahtio |
<3 |
| 10:01 |
phasefx |
kivilahtio: and limited printing and no offline mode or weird peripheral integration |
| 10:01 |
Dyrcona |
Make it a MUD while we're at it and cataloging could be a question with rewards in gold and XP. |
| 10:01 |
kivilahtio |
we could go dumpster diving for new staff clients! |
| 10:01 |
phasefx |
but that stuff could be separate helper apps |
| 10:01 |
kivilahtio |
machines |
| 10:02 |
Dyrcona |
phasefx: groff and lp for printing! |
| 10:02 |
kivilahtio |
I mean our librarians are always asking for quick keys. They can't use a mouse, so... |
| 10:03 |
phasefx |
my understanding is that there are some terminal codes that could kick off local printing |
| 10:03 |
kivilahtio |
win-win |
| 10:03 |
kivilahtio |
+ no cross browser compatibility issues |
| 10:03 |
phasefx |
but I have no experience with that at all |
| 10:04 |
phasefx |
kivilahtio: instead we can worry about character encoding issues :) |
| 10:04 |
phasefx |
make sure you set putty to use UTF-8 |
| 10:04 |
Dyrcona |
kivilahtio: Most of ours can't use a computer, never mind a mouse. |
| 10:04 |
atheos |
I know it's really late, but I just now got the green light from my library in regards to Hack-A-Way. I see that that Eventbrite shows the signup as closed. Is it going to be possible for me to signup this late? |
| 10:05 |
Dyrcona |
atheos: only dbwells can answer that question. |
| 10:05 |
|
zxiiro_ joined #evergreen |
| 10:05 |
Dyrcona |
hopefully, he'll notice he was mentioned and come answer it. |
| 10:05 |
atheos |
Righto, Thank You. |
| 10:06 |
* dbwells |
like-magic appears |
| 10:07 |
kivilahtio |
ok I tested phasefx's hint |
| 10:07 |
* krvmga_ |
returns to room. Had to take a quick cold shower after hearing screen flicker and ANSI terminal codes. |
| 10:07 |
kivilahtio |
and staff_cleint logs in |
| 10:07 |
* Dyrcona |
builds a new VM. |
| 10:08 |
dbwells |
atheos: Basically, I am not really sure. |
| 10:08 |
kivilahtio |
phasefx: So how do we know should we use https or http? |
| 10:08 |
dbwells |
atheos: The easiest answer is yes, it is way too late. However, that doesn't mean we can't make it happen. |
| 10:10 |
dbwells |
atheos: The most serious obstacle would be finding space in the hotel. |
| 10:10 |
|
zxiiro joined #evergreen |
| 10:11 |
dbwells |
atheos: If you are 100% committed to coming, I can start looking into making it happen. |
| 10:13 |
atheos |
yes, I can call and see about available hotels, running to grab the library CC right now |
| 10:14 |
|
zerick joined #evergreen |
| 10:14 |
kivilahtio |
dbwells: atheos: I think finding a hotel is the least of your problems. I always use booking.com and there are always plenty of hotels to pick from |
| 10:15 |
kivilahtio |
dbwells: And you could always just find a couch to crash on? |
| 10:15 |
Dyrcona |
kivilahtio: I think dbwells is referring to the onsite hotel at Calvin College where the event is being held. |
| 10:16 |
dbwells |
atheos: Last I heard, we had one or two more rooms kept in reserve for the on-campus hotel. Of course this was around two weeks ago. |
| 10:16 |
atheos |
right, I'll try there first. |
| 10:16 |
Dyrcona |
atheos: might be better if you let dbwells check. I caused a little confusion by not going through him for the room. |
| 10:16 |
dbs |
kivilahtio: re SIP tunneling, http://coffeecode.net/archives/219-Setting-up-secure-self-check-connections-using-SIP-tunneled-through-SSH.html might be of use |
| 10:17 |
kivilahtio |
dbs: Thanks! I'll archive that |
| 10:17 |
kivilahtio |
looking to SIP soon |
| 10:17 |
berick |
hmm, dbwells while you're in there, can you confirm that I have a room? i know I responded to your room-related queries.. |
| 10:18 |
atheos |
you could use OpenVPN to tunnel through to your SIP server too, which does a great job of maintaining secure connectivity. |
| 10:19 |
kivilahtio |
thanks atheos |
| 10:19 |
kivilahtio |
Alas, I must flee |
| 10:19 |
kivilahtio |
for the time is nigh |
| 10:20 |
kivilahtio |
need to go clean our flat for my girl to come home |
| 10:20 |
kivilahtio |
was nice cahtting with you, as always |
| 10:29 |
|
mcooper joined #evergreen |
| 10:30 |
dbwells |
berick: yep, all set. |
| 10:34 |
berick |
dbwells: great, thanks |
| 10:48 |
phasefx |
any known issues with pub sorting? I created three records with variations on 260 3|_ $a foo $b bar $c 1900, and Date: Newest to Oldest isn't working like I expect |
| 10:51 |
eeevil |
phasefx: pub date sorting is based on 008/date1 |
| 10:51 |
eeevil |
by default |
| 10:52 |
eeevil |
you can configure the pubdate record attr to pull from 260$c if you want to, but generally date1 is more likely to be populated in the wild |
| 10:52 |
phasefx |
ah ha |
| 10:52 |
phasefx |
eeevil++ |
| 10:52 |
kmlussier |
eeevil: Is the publication year filter from the advanced search page also based on the 008? |
| 10:53 |
eeevil |
kmlussier: yep |
| 10:53 |
kmlussier |
eeevil: Thanks! Good to know. |
| 10:53 |
phasefx |
I think I'm still seeing a counter example, but will double-check |
| 10:54 |
eeevil |
if you find one, please do point me at it |
| 11:01 |
|
dboyle joined #evergreen |
| 11:10 |
phasefx |
so for reference, this is bug 1174860 that I hit upon |
| 11:10 |
pinesol_green |
Launchpad bug 1174860 in Evergreen "tpac: search filter groups don't play nicely with facets" (affected: 3, heat: 20) [Undecided,Triaged] https://launchpad.net/bugs/1174860 |
| 11:17 |
|
mrpeters joined #evergreen |
| 11:31 |
Dyrcona |
dbwells++ #organizing the hackaway details |
| 11:32 |
|
mllewellyn joined #evergreen |
| 11:47 |
|
jdouma joined #evergreen |
| 12:04 |
|
acoomes joined #evergreen |
| 12:15 |
mrpeters |
anyone have experience with using WAL archives to roll back transactions that occured in say, the past hour? |
| 12:16 |
mrpeters |
we have a customer who we discovered wasn't doing backups, and found that there was an errant query run and we're hoping we can use the wal archives to roll back in time a bit to before the errant query |
| 12:24 |
Dyrcona |
mrpeters: you should also ask in a postgres channel. you might get more answers there, if you haven't already. |
| 12:25 |
mrpeters |
i may, if it gets to that....i know PITR will be my friend here, just not finding any good guides yet |
| 12:25 |
eeevil |
mrpeters: you cannot do that with only the WAL. you must have a database snapshot from before the point in time you want to restore to, and all the WAL between that snapshot and the point in time you want |
| 12:25 |
mrpeters |
eeevil: thats what i was afraid to hear... |
| 12:25 |
mrpeters |
how you run a production evergreen with no db backups is beyond me...but here we are |
| 12:26 |
eeevil |
mrpeters: http://www.postgresql.org/docs/9.1/static/continuous-archiving.html for reference |
| 12:26 |
mrpeters |
yep, been reading |
| 12:26 |
Dyrcona |
i've seen a site running a production environment with what they thought were backups, but couldn't be used to recover anything. |
| 12:26 |
* mrpeters |
prays they ran a report at some point with all of their users and barcodes |
| 12:31 |
|
misilot joined #evergreen |
| 12:34 |
* dbs |
wrote http://en.flossmanuals.net/evergreen-in-action/care-and-feeding-of-evergreen/ to try and avoid situations like this :/ |
| 12:36 |
Dyrcona |
Ah, but one must first find it, then read it, then understand what has been read, and then act upon the new information. |
| 12:36 |
|
smyers_ joined #evergreen |
| 12:37 |
Dyrcona |
It so much easier to put coconuts on your head and wave palm fronds at the end of the dilapidated runway. |
| 12:37 |
dbs |
Dyrcona: yeah, that's why we targeted it at new EG admins |
| 12:47 |
mrpeters |
dbs++ |
| 12:48 |
Dyrcona |
I'm not knocking the manual. |
| 12:48 |
* Dyrcona |
is more snarky than usual lately.... |
| 12:56 |
* mrpeters |
wishes he knew who configured this server so he could /slap them |
| 12:57 |
* mrpeters |
smacks self for assuming they had backups |
| 13:04 |
Dyrcona |
hmm.. Having trouble with recent checkout of OpenSRF master. |
| 13:07 |
Dyrcona |
SettingsParser fails to parse the bootstrap config, even when told exactly where it is. |
| 13:10 |
* Dyrcona |
smacks himself with a red herring. |
| 13:10 |
Dyrcona |
config issue on my end, nm. |
| 13:11 |
mrpeters |
are there any tables, like circ or billing that would have kept a link to a users actor.card value? |
| 13:11 |
mrpeters |
problem = 3,000 users with a null actor.usr.card and no entries in actor.card.usr that match actor.usr.id |
| 13:11 |
Dyrcona |
nope. only actor.usr. You could check the auditor tables though, there might be information there. |
| 13:11 |
mrpeters |
ooh |
| 13:12 |
mrpeters |
damn no audit table for actor.card |
| 13:12 |
Dyrcona |
it's small consolation, but you can configure it for the future. |
| 13:13 |
mrpeters |
indeed |
| 13:13 |
mrpeters |
real backups are coming |
| 13:14 |
mrpeters |
sigh -- SELECT count(id) FROM actor.usr WHERE card IS NULL AND id IN (SELECT id FROM auditor.actor_usr_history WHERE card IS NOT NULL); results in nothin |
| 13:14 |
mrpeters |
which seems shocking, but iguess maybe the users never got edited |
| 13:16 |
jeff |
unless you intentionally disabled triggers when running the query that did the damage, those updates should have triggered audit entries to be created. it might not help at all, since actor.usr.card is going to most likely point to an actor.card.id that no longer exists, and you're going to want the value of actor.card.barcode. |
| 13:17 |
jeff |
disclaimer: i'm not fully focused on your issue, so i could be off. :-) |
| 13:17 |
eeevil |
mrpeters: any chance their usrname fields are the barcodes? |
| 13:17 |
mrpeters |
a few |
| 13:18 |
jeff |
depending on your log level and log retention, given enough determination you could recoer some barcodes that way. |
| 13:19 |
jeff |
really unfortunate to be in a situation where you're forced to consider that kind of thing. |
| 13:26 |
* Dyrcona |
chuckles @ --kill-with-fire |
| 13:48 |
|
yboston joined #evergreen |
| 13:53 |
paxed |
http://i.imgur.com/SDDpwfw.gif |
| 14:25 |
|
stevenyvr2 joined #evergreen |
| 14:43 |
|
dconnor joined #evergreen |
| 14:52 |
|
dconnor joined #evergreen |
| 14:55 |
jboyer-isl |
Here's a (not very) fun exercise. Does anyone know the simplest way to generate -half- of a cross join on ou ids? We're going to start using the prox adjustment table after an upgrade and rather than come up with 24,806 rows, I'm hoping to only do 12,403 and then mirror the ous. |
| 14:57 |
* tsbere |
hasn't dug into that process much |
| 14:59 |
jboyer-isl |
Oh, oops. I hadn't thought it all the way through either. Those numbers still include adjustments to locations in the same system, which isn't necessary... I should have a lot less rows to worry about. Just coming up with the correct count is proving difficult. Hmm. |
| 14:59 |
tsbere |
jboyer-isl: For note, my understanding is that you can set the values at a parent ou and the children inherit it |
| 14:59 |
tsbere |
which you obviously just realized |
| 14:59 |
jboyer-isl |
Oh, that would be awesome. |
| 15:00 |
jboyer-isl |
I didn't realize that exactly, but it's a much bigger savings than what I was thinking. |
| 15:09 |
|
kmlussier joined #evergreen |
| 15:12 |
|
krvmga__ joined #evergreen |
| 15:13 |
|
dconnor joined #evergreen |
| 15:13 |
krvmga__ |
if a bib record has multiple 007 fields will search look at them all? this has come up in conjunction with libraries who bought DVD/Blu-Ray combo packs and did NOT split them. |
| 15:14 |
krvmga__ |
what i thought would happen is that, if i searched on DVD, the combo pack would show up AND if i did a separate search on Blu-Ray the combo pack with also show up |
| 15:14 |
krvmga__ |
that doesn't seem to be happening, though |
| 15:14 |
krvmga__ |
what seems to happen is that search takes the first 007 that it finds and ignores the others |
| 15:14 |
krvmga__ |
i just wanted to find out if this was actually true |
| 15:15 |
Dyrcona |
that's exactly how it works. |
| 15:15 |
jboyer-isl |
@marc 007 |
| 15:15 |
pinesol_green |
jboyer-isl: This field contains special information about the physical characteristics in a coded form. The information may represent the whole item or parts of an item such as accompanying material. (Repeatable) [] |
| 15:15 |
krvmga__ |
so...is this a bug or a feature? :) |
| 15:16 |
krvmga__ |
jboyer-isl++ |
| 15:16 |
krvmga__ |
thx for doing that |
| 15:16 |
jboyer-isl |
I didn't actually expect it to be repeatable. I guess that nudges it more towards the bug side of the line. |
| 15:16 |
Dyrcona |
I'll leave that up to eeevil to decide. He's talked about making it work with multiple fields before. |
| 15:17 |
krvmga__ |
i'm just wondering if it's worth putting on launchpad then |
| 15:17 |
Dyrcona |
Right now, we call it SVF, single value fields. The goal would be MVF, multi-value fields. |
| 15:17 |
kmlussier |
krvmga_Yes, the multiple fields Dyrcona refers to is the possible MVF development we've been talking about at MassLNC. |
| 15:17 |
krvmga__ |
sweet |
| 15:17 |
krvmga__ |
thx |
| 15:17 |
Dyrcona |
Doesn't hurt to put it on Launchpad. It will likely end up there anywya. |
| 15:17 |
kmlussier |
krvmga_: But it still isn't certain that we're moving forward with it. It's the icon development. |
| 15:18 |
Dyrcona |
I'd say it is useful even without the icon bit. |
| 15:19 |
krvmga__ |
i'll try to describe it for launchpad |
| 15:20 |
* eeevil |
agrees that it's useful by itself. though the described issue was always a known limitation of SVF ... the trade-off was between getting 99.9% coverage of bibs in the wild, or not get any for another release or more |
| 15:21 |
* Dyrcona |
agrees something is better than nothing. |
| 15:25 |
* krvmga_ |
also agrees. |
| 15:25 |
krvmga_ |
https://bugs.launchpad.net/evergreen/+bug/1219056 |
| 15:25 |
pinesol_green |
Launchpad bug 1219056 in Evergreen "OPAC - search doesn't handle multiple MARC 007 fields; only takes the first one" (affected: 1, heat: 6) [Undecided,New] |
| 15:25 |
eeevil |
the reason I didn't start with MVF is it's a hard problem to solve well (and an easy problem to solve badly) ... fwiw |
| 15:28 |
|
dconnor joined #evergreen |
| 15:32 |
|
dboyle joined #evergreen |
| 15:47 |
|
dconnor joined #evergreen |
| 16:02 |
|
moodaepo_nb joined #evergreen |
| 16:08 |
Dyrcona |
@hate MARC |
| 16:08 |
pinesol_green |
Dyrcona: But Dyrcona already hates MARC! |
| 16:08 |
Dyrcona |
Well, I hate it some more! |
| 16:11 |
jboyer-isl |
@hate perl |
| 16:11 |
pinesol_green |
jboyer-isl: The operation succeeded. jboyer-isl hates perl. |
| 16:11 |
jboyer-isl |
So damn much. |
| 16:12 |
|
acoomes joined #evergreen |
| 16:12 |
Dyrcona |
Well, it isn't MARC so much as the crap that proprietary ILS vendors allow their systems to spew out in the guise of MARC. |
| 16:13 |
Dyrcona |
Though MARC as a data storage format, plain sucks. |
| 16:13 |
Dyrcona |
It was fine for transmitting data to print catalog cards, but for what we do now, fuggedaboutit. |
| 16:21 |
|
mrpeters left #evergreen |
| 16:34 |
|
smyers_ joined #evergreen |
| 16:41 |
|
dconnor joined #evergreen |
| 16:54 |
|
dconnor joined #evergreen |
| 16:57 |
|
dconnor joined #evergreen |
