| Time |
Nick |
Message |
| 00:55 |
|
phasefx_ joined #evergreen |
| 00:55 |
|
Callender joined #evergreen |
| 01:04 |
|
sbrylander joined #evergreen |
| 01:05 |
|
berick joined #evergreen |
| 01:06 |
|
phasefx__ joined #evergreen |
| 01:46 |
|
pinesol_green` joined #evergreen |
| 01:46 |
|
brahmina joined #evergreen |
| 02:10 |
|
vlewis joined #evergreen |
| 07:10 |
kmlussier |
@weather |
| 07:10 |
pinesol_green` |
kmlussier: Seekonk, MA :: Clear :: 39F/4C | Wind Chill: 35F/2C | Friday: Sunshine and clouds mixed. High around 55F. Winds W at 15 to 25 mph. Friday Night: Some clouds this evening will give way to mainly clear skies overnight. Low 28F. Winds NW at 10 to 15 mph. |
| 07:45 |
|
rjackson_isl joined #evergreen |
| 07:54 |
|
JBoyer joined #evergreen |
| 07:56 |
bshum |
Hmm, just noticed that the README still says "Evergreen 2.8 has been tested with..." for both 2.9 and 2.10. Probably a matter of updating that part in master/branching of new series |
| 07:57 |
bshum |
That's better, stupid ` |
| 08:00 |
|
remingtron_ joined #evergreen |
| 08:25 |
|
collum joined #evergreen |
| 08:38 |
|
mmorgan joined #evergreen |
| 08:40 |
|
maryj joined #evergreen |
| 08:46 |
|
ethomsen1 joined #evergreen |
| 08:49 |
|
Dyrcona joined #evergreen |
| 08:50 |
|
krvmga joined #evergreen |
| 09:36 |
Dyrcona |
Are there any good examples of queries using xpath in the database to extract fields from bre.marc? |
| 09:37 |
jeff |
if you don't care about the fields potentially being normalized, i go with metabib.real_full_rec |
| 09:37 |
miker |
Dyrcona: what are you looking for? there are plenty in the code... |
| 09:38 |
jeff |
but since you're asking, you probably care about the values being normalized (i.e., don't want the normalization) |
| 09:38 |
Dyrcona |
I just wanted to check that a subfield I added was added correctly. |
| 09:38 |
Dyrcona |
I was trying just xpath() and got {} |
| 09:39 |
Dyrcona |
The I tried oils_xpath and I get a bunch of spaces. |
| 09:39 |
Dyrcona |
I have verified by spot checking a couple of the records in the client. |
| 09:40 |
miker |
Dyrcona: select oils_xpath_string('//*[@tag="245"]/*[@code="a"]', marc) from biblio.record_entry order by id desc limit 10; |
| 09:40 |
Dyrcona |
Hmm. Works with oils_xpath if I use * instead of tag names. |
| 09:40 |
jeff |
Dyrcona: https://gist.github.com/jeff/45d799d50b9f87ad83b5 |
| 09:40 |
miker |
stuff like that? |
| 09:41 |
jeff |
er, almost identical to what miker said. :-) |
| 09:41 |
* dbwells |
deletes his version of the same thing |
| 09:41 |
miker |
Dyrcona: you have to supply the namespace or use *, or use [local-name()="datafield"] and the like |
| 09:41 |
jeff |
dbwells: pretty sure my version was partially based on an example you provided long long ago. :-) |
| 09:41 |
Dyrcona |
I was using the namespace with just xpath, but not oils_xpath. |
| 09:45 |
jeff |
I am reminded that I would really love to move away from temporary bibs for NCIP requests, and use (now holdable) precats. |
| 09:46 |
miker |
Dyrcona: select xpath('//m:datafield[@tag="245"]/m:subfield[@code="a"]', marc::xml, ARRAY[ARRAY['m','http://www.loc.gov/MARC21/slim']]) from biblio.record_entry order by id desc limit 10; -- with element names |
| 09:47 |
Dyrcona |
Um, y'know....It helps to get the correct tag names. :) |
| 09:47 |
miker |
;) |
| 09:48 |
gmcharlt |
:) |
| 09:48 |
|
yboston joined #evergreen |
| 09:48 |
Dyrcona |
jeff: The staff client doesn't handle now holdable precats the way I'd like. |
| 09:49 |
Dyrcona |
I was doing basically what miker posted, 'cept I had "field" not "datafield." |
| 09:49 |
Dyrcona |
'Cause y'know, I had just used MARC::Record->field() |
| 09:51 |
miker |
I wish the perl module had always separated those ... the structure is different, and the "tag < 010" rule is just a USMARC thing |
| 09:52 |
|
mllewellyn joined #evergreen |
| 09:54 |
jeff |
Dyrcona: since it's NCIP, mostly i don't care about the staff client, but... in what way would you like them to behave that they do not currently? |
| 09:55 |
Dyrcona |
jeff: LP 1485077 |
| 09:55 |
pinesol_green |
Launchpad bug 1485077 in Evergreen "Precat Checkout always ITEM_NOT_CATALOGED even when copy exists" [Undecided,New] https://launchpad.net/bugs/1485077 - Assigned to Jason Stephenson (jstephenson) |
| 09:56 |
Dyrcona |
I should unassign myself....I'm not actually working on that. |
| 09:58 |
Dyrcona |
The issue as I saw it, is that even the copies and holds, etc., are placed by the ILL software, staff still have to check it out to the patron in Evergreen. |
| 09:58 |
jeff |
oh, in your workflow the final checkout to the patron takes place not in the ILL software, but in the staff client? Interesting. |
| 09:58 |
Dyrcona |
Well, yeah... It goes on hold, the patron gets a notice, and staff check it out when the patron shows up. |
| 09:59 |
jeff |
In our workflow, all checkouts must be performed in the ILL software. |
| 09:59 |
jeff |
(which is in many ways frustrating/annoying, but we get no choice in the matter) |
| 10:00 |
Dyrcona |
Yep. |
| 10:00 |
Dyrcona |
I think it's "dangerous" to just automagically check things out to patrons. |
| 10:00 |
jeff |
We then get notice of the checkout via NCIP and check the item out to the patron in Evergreen automatically based on that NCIP message, etc. |
| 10:00 |
Dyrcona |
We've had a few ILLs not get picked up. I had to modify NCIPServer to handle that case. |
| 10:01 |
jeff |
Yeah, in our environment the checkout doesn't happen until the patron comes to pick up their item. |
| 10:01 |
jeff |
But I remember this bug report now that you've reminded me of it. |
| 10:02 |
Dyrcona |
I thought I'd try changing the behavior, but no time. |
| 10:03 |
Dyrcona |
So, NCIPServer has the ability to do precats, but with a note in the documentation not to use it. :) |
| 10:03 |
Dyrcona |
As for the workflow, that's pretty much how it works with this vendor. |
| 10:05 |
Dyrcona |
I suppose I *could* check it out when they send the AcceptItem message, but that's at the time it actually goes on hold. |
| 10:20 |
jeff |
yeah, i think checking it out to the patron any time before the patron has the item in hand is probably a bad idea. |
| 10:21 |
jeff |
one library that we work with participates in the same statewide system, but without NCIP. since they have a limited number of staff who are trained on the ILL client, and that staff member is only in one or two days a week, their practice is to check the item out to the patron when it arrives. |
| 10:21 |
jeff |
then the patron just comes in and picks up the book, without any further action in the ILL software. |
| 10:21 |
jeff |
in their case, there are no checkouts (NCIP or manual) in the ILS, just the ILL software. |
| 10:29 |
|
mmorgan1 joined #evergreen |
| 10:46 |
|
Christineb joined #evergreen |
| 10:46 |
|
rjackson_isl_ joined #evergreen |
| 10:51 |
csharp |
acq-- |
| 10:52 |
gmcharlt |
demo.evergreencatalog.com is now running 2.10.0 |
| 10:52 |
jeff |
gmcharlt++ |
| 11:05 |
|
brahmina joined #evergreen |
| 11:08 |
Bmagic |
gmcharlt++ # 2.10 yay! |
| 11:40 |
|
bmills joined #evergreen |
| 11:57 |
Bmagic |
added code to bug 1552409 |
| 11:57 |
pinesol_green |
Launchpad bug 1552409 in Evergreen "Invent a page in the OPAC for redirection to referring URL based authentication on external services such as EBSCO" [Undecided,New] https://launchpad.net/bugs/1552409 |
| 12:01 |
|
jihpringle joined #evergreen |
| 12:02 |
* tsbere |
comments on that bug along the lines of "wait, doesn't that not work?" |
| 12:02 |
tsbere |
Bmagic: ^^ |
| 12:02 |
* tsbere |
goes to lunch |
| 12:02 |
Bmagic |
tsbere: yes, that was one of the things we addressed with EBSCO in this case |
| 12:03 |
Bmagic |
it turns out if it's https -> https, it works |
| 12:03 |
Bmagic |
Chrome keeps the referring url headers during a 301 but FF does not |
| 12:03 |
Bmagic |
so, we used JS to load the final URL from the opac, which works in FF. The headers are intact and it's working! The electronic resources are accessable! |
| 12:07 |
|
mmorgan joined #evergreen |
| 12:11 |
csharp |
does anyone here use Google Books preview from TPAC? |
| 12:11 |
dbs |
csharp: I can't remember if we still have it turned on or not |
| 12:12 |
dbs |
I was getting super frustrated because it kept insisting on making HTTP calls even though we were invoking it with HTTPS |
| 12:12 |
csharp |
ah |
| 12:12 |
dbs |
Nope, disabled atm |
| 12:13 |
csharp |
we just got a request from a library staff to add some sort of sample (ala Amazon Kindle) feature, and I was just trying to see if Google Books preview was still in action anywhere |
| 12:13 |
* dbs |
tries turning it on to see if they've fixed that yet |
| 12:14 |
mmorgan |
We have just turned on the google books preview in our live system http://catalog.noblenet.org |
| 12:16 |
|
ethomsen joined #evergreen |
| 12:16 |
dbs |
Yeah, I think it works fine if you're using HTTP |
| 12:16 |
* dbs |
just got "api_8eb3a7c…__fr.js?callback=google.loader.callbacks.books:63 Mixed Content: The page at 'https://laurentian.concat.ca/eg/opac/record/729926?locg=105#gbp_extra' was loaded over HTTPS, but requested an insecure script 'http://books.google.ca/books?id=ELieXMxR1h4C&printsec=frontcover&dq=isbn%3A…25&hl=fr&cd=1&source=gbs_api&jscmd=run&callback=google.books.__callback._0'. This request has |
| 12:16 |
* dbs |
been blocked; the content must be served over HTTPS." |
| 12:17 |
mmorgan |
...after incorporating some changes to the way it opens the preview to get around the http/https thing |
| 12:17 |
csharp |
mmorgan: do you have a sample record that shows it? my test searches aren't yielding results with Google Preview |
| 12:17 |
csharp |
dbs: gotcha |
| 12:17 |
mmorgan |
looking... |
| 12:18 |
dbs |
http://catalog.noblenet.org/eg/opac/record/2279981 |
| 12:18 |
dbs |
Preview at right, opens in a new tab. That would avoid HTTP/HTTPS issues, yep. |
| 12:19 |
csharp |
dbs: perfect |
| 12:19 |
mmorgan |
dbs: Thanks! |
| 12:19 |
mmorgan |
Good choice!! |
| 12:20 |
* csharp |
has that book on his office shelf |
| 12:20 |
mmorgan |
There's a small issue in the staff client where it opens a blank xul window, but we're considering that minor. |
| 12:20 |
mmorgan |
It does then open the preview in the browser. |
| 12:20 |
bshum |
dbs: Oh, never noticed the http/https issue, figures |
| 12:22 |
csharp |
mmorgan: when stuff like that doesn't work in the xul client, we just say "Don't do that in the xul client." :-) |
| 12:22 |
StomproJosh |
How likely is it that 2.10.1 will be released on April 20th, the first day of the conference? |
| 12:23 |
csharp |
we had a similar issue with NoveList Select when we first implemented that |
| 12:23 |
dbs |
it's a two-line patch to make the links open in a new tab, fwiw |
| 12:23 |
mmorgan |
We're expecting it to work just fine in web client. |
| 12:23 |
csharp |
dbs++ mmorgan++ # thanks for your help |
| 12:23 |
ethomsen |
Our libraries liked the idea of just having a button that links to the Google Books page rather than embedding the book in the Evergreen page; some had concerns about having the Buy option display on an Evergreen page |
| 12:24 |
csharp |
ethomsen: yeah, I can see that being an issue |
| 12:24 |
csharp |
"Buy for my library" might be okay :-P |
| 12:41 |
ethomsen |
Best selling point for Google Books Preview in catalog -- see if the book has the specific info you want: https://goo.gl/Trb8D6 |
| 12:48 |
* tsbere |
returns from lunch and continues his launchpad back and forth with Bmagic |
| 13:11 |
JBoyer |
I apologize for the lazyweb search, but is there a way in the receipt editor to transform a string to a vertical line? (not rotation. I want to change "word" into "w<br/>o<br/>r<br/>d" |
| 13:16 |
miker |
JBoyer: there is, with css. http://www.css3maker.com/text-rotation.html |
| 13:16 |
JBoyer |
That works in the xulrunner client? |
| 13:16 |
miker |
ah... well. no |
| 13:17 |
miker |
:) |
| 13:17 |
miker |
well, maybe, with the -moz- prefix |
| 13:18 |
JBoyer |
I have rotated blocks of text before to do holds slips that stick out off the shelf, but I'm specifically not looking for rotation. :) (see that short example) |
| 13:18 |
miker |
JBoyer: arg... I read that as not wanting to do that |
| 13:19 |
miker |
element.innerHTML = ... ? |
| 13:19 |
* miker |
ducks |
| 13:19 |
JBoyer |
No problem. |
| 13:19 |
JBoyer |
If it were for my library I'd have taken care of it with a JS include, I can't tell put "Now have your JS dev write a function to transform this text, put it on your site, and put this in your LSE" ;) |
| 13:20 |
JBoyer |
I can't put that *in a ticket* Is how one would english that. |
| 13:21 |
miker |
how about: make the container 1 em wide and give it "word-wrap: break-word" in css |
| 13:21 |
miker |
er, .5em |
| 13:21 |
miker |
http://www.w3schools.com/cssref/tryit.asp?filename=trycss3_js_word-wrap |
| 13:22 |
miker |
change the width in the style to .5em, click see result, then click try it |
| 13:22 |
JBoyer |
That's kind of how I'm leaning given what I've seen online. I was hoping that one of the receipt macros that I can never remember could do it. |
| 13:23 |
JBoyer |
In the interest of helping the lib out I'll just do that, once we're all web, all the time, we'll look at other options. |
| 13:23 |
JBoyer |
Thanks |
| 13:23 |
JBoyer |
miker++ |
| 13:23 |
miker |
np |
| 14:06 |
kmlussier |
Good afternoon #evergreen. How is everyone today? |
| 14:07 |
* berick |
is swell |
| 14:09 |
kmlussier |
berick: That's a word that really should make a comeback IMO |
| 14:11 |
berick |
yeah, I only hear it used sarcastically these days. (not that I meant it that way). |
| 14:11 |
kmlussier |
Since it looks like we don't have a developer meeting for a few more weeks, I wanted to bring up in here the recent list discussion we've had re: submitting a grant request to help fund an Evergreen developer bootcamp. |
| 14:12 |
kmlussier |
There seems to be some interest/enthusiasm in seeing it happen, but I'm wondering if there is any interest from some of the developers in becoming instructors/mentors. |
| 14:13 |
* kmlussier |
sometimes wishes we could see each other so that she could stare people down. ;) |
| 14:16 |
berick |
kmlussier: do you have an idea of what how long the commitment will be, how often the bootcamps will happen, etc.? |
| 14:18 |
kmlussier |
My vision (which doesn't have to be the final vision) is that it would require two in-person meetups. One shorter ones among the instructors/leaders to possibly engage in mentorship training, discussing ideas to lower barriers, fleshing out content, etc. |
| 14:18 |
kmlussier |
And then a longer time period (a week tops?) for the actual bootcamp. |
| 14:19 |
kmlussier |
For now, I just need to know if there is general interest / willingness to help out if we were to get the grant. Not a firm commitment. |
| 14:20 |
kmlussier |
I would like to work with a few others to really flesh out the details and then maybe present it to the group. I think it would make a great discussion at the hackfest. |
| 14:20 |
kmlussier |
Unfortunately, I can't wait that long before deciding to whether to submit the idea to the granting foundation. |
| 14:22 |
berick |
i'm certainly interested, moreso if it's not an extended engagement |
| 14:22 |
kmlussier |
Looks like grant winners are announced at the end of June, so nothing could start until then. |
| 14:24 |
kmlussier |
berick: I think the extended bit comes in with keeping connectings after the bootcamp. We might want to query folks as to where they need further learning, and maybe have a monthly IRC chat to focus on that area. But I wouldn't expect everyone involved to be available at each chat. |
| 14:25 |
kmlussier |
Also, if it increases the amount of new people who are poking around on a regular basis, the questions posed in here will increase. |
| 14:27 |
* berick |
nods |
| 14:27 |
gmcharlt |
kmlussier: do you envision the grant compensating the mentors/instructors for their time? the commitment called upon to do the boot camp in particular isn't just the week, but time spent preparing for it |
| 14:28 |
gmcharlt |
i.e., it's one thing to encourage folks to be generous of their time answering questions in IRC and the mailing lists and so forth |
| 14:28 |
kmlussier |
gmcharlt: It could. My idea was to have the grant compensate for travel expenses. But if there's a thought that compensation for time is needed, we could work it in. |
| 14:29 |
kmlussier |
My question is who are we compensating? Since we all work for organizations that use Evergreen, does the money go to the organization or to the individual person? |
| 14:30 |
kmlussier |
Who, I suppose, would then need to take time off to help out rather than working in their 'official' capacity. |
| 14:30 |
|
tspindler joined #evergreen |
| 14:30 |
jeff |
each individual's situation is likely to differ. |
| 14:30 |
kmlussier |
hmmmm...I'm now thinking back to the old KLCS community partners grant. |
| 14:31 |
kmlussier |
That might be a good model to follow. The organization was the partner, the participants reported the amount of time they worked on Evergreen, and the grant reimbursed the organization. |
| 14:32 |
gmcharlt |
and cases where the individual is also the organization, as it were, follow the same model |
| 14:32 |
JBoyer |
That sounds more workable than having the individual take time off and also get paid for it. The pay might be appreciated, but you still took a bunch of time off work. |
| 14:32 |
gmcharlt |
possibly with some "Hi Self, shall I reimburse myself for self activities" silliness |
| 14:34 |
kmlussier |
OK, so I'll rephrase my question from earlier. If we work something into the grant that reimburses for the time instructors/mentors put into the bootcamp, would there be interest from some of your to help out? I don't want to make it sound like you're only in it for the money, but I do understand there are other factors that come into play. |
| 14:35 |
kmlussier |
I just don't want to submit what I think is a great idea to a granting organization and then withdraw it because the community is too stretched to make it work well. |
| 14:39 |
JBoyer |
I would like to, but I'm not certain at this point if I could do it justice. |
| 14:40 |
csharp |
adding new developers slows down a project, period: https://en.wikipedia.org/wiki/Brooks%E2%80%99_law but I think the payoff is worth it for the long-term |
| 14:40 |
JBoyer |
Being some months out probably helps, I suppose. |
| 14:44 |
Dyrcona |
csharp: I'm not sure Brooks' Law holds under all circumstances... Gilb's Law on the other hand.... ;) |
| 14:45 |
Dyrcona |
At this point, I think Evergreen would definitely benefit from more developers. |
| 14:47 |
|
bmills joined #evergreen |
| 14:48 |
csharp |
my point was supposed to be that if we want the project to grow, we need developers to take time out of current development to help train n00bs :-) |
| 14:48 |
csharp |
sounds like everyone gets that, though :-) |
| 14:48 |
kmlussier |
I think a lot also depends on what we decide to cover. A good first step, probably before the conference, might be to do a small survey among newer code contributors on things they wish they new when they got started. |
| 14:48 |
Dyrcona |
csharp: Point well taken, and in the long run, I think that will be time well spent. |
| 14:48 |
kmlussier |
csharp: I got it. |
| 14:52 |
|
bmills joined #evergreen |
| 15:25 |
|
ethomsen left #evergreen |
| 15:37 |
jeff |
heh. kinda' hard to search for "anything but english" using the advanced search form -- pretty easy to get "Request URI is too long" :-) |
| 15:37 |
tsbere |
jeff: I usually cheat for that. |
| 15:38 |
tsbere |
jeff: enter "-item_lang(eng)" into a search box and all |
| 15:38 |
* jeff |
nods |
| 15:39 |
jeff |
is demo.evergreencatalog.com running with a set of bibs other than concerto? |
| 15:43 |
JBoyer |
@hate z39.50 |
| 15:43 |
pinesol_green |
JBoyer: The operation succeeded. JBoyer hates z39.50. |
| 15:45 |
jeff |
woo. closer to figuring out mods encoding issues. |
| 15:53 |
gmcharlt |
jeff: IIRC yes; it has concerto bibs, but also rather more |
| 15:53 |
csharp |
@who wants rather more? |
| 15:53 |
pinesol_green |
Christineb_away wants rather more. |
| 15:55 |
Dyrcona |
@who wants less? |
| 15:55 |
pinesol_green |
rashma wants less. |
| 15:55 |
Dyrcona |
@eightball is less more? |
| 15:55 |
pinesol_green |
Dyrcona: No. |
| 15:59 |
jeff |
huh. i thought one of our sample datasets had a bib for "Harry Potter Y El Cáliz De Fuego |
| 15:59 |
jeff |
", but apparently not. |
| 15:59 |
gmcharlt |
jeff: feel free to add it :) |
| 16:00 |
jeff |
quite possible that i was confused by the fact that we have it in production. :-) |
| 16:00 |
gmcharlt |
(to demo.ev.., I mean) |
| 16:00 |
* jeff |
nods |
| 16:01 |
|
jlitrell joined #evergreen |
| 16:02 |
jeff |
bre.id 7071 :-) |
| 16:05 |
|
tspindler left #evergreen |
| 16:05 |
rhamby |
Am I the only one who has ever wondered what kind of classification system Hogwarts uses? |
| 16:05 |
rhamby |
Their library I mean. |
| 16:06 |
mmorgan |
I wouldn't imagine it matters. I suspect the books would move around by themselves anyway. |
| 16:06 |
kmlussier |
rhamby: Yes, I think you are the only one. |
| 16:06 |
tsbere |
rhamby: Based on a little evidence a friend of mine came up with a description of "We know this is on subject X, and it looks good in this place on the shelf" ;) |
| 16:06 |
rhamby |
kmlussier: fair enough |
| 16:06 |
kmlussier |
@sortinghat rhamby |
| 16:06 |
pinesol_green |
Hmm... rhamby... Let me see now... RAVENCLAW! |
| 16:06 |
kmlussier |
Seems like a good place for you rhamby! |
| 16:07 |
kmlussier |
I would like to think they organize their books based on color. |
| 16:07 |
rhamby |
they must have some degree of sorting though or they wouldn't have things in an actual restricted section. |
| 16:07 |
tsbere |
kmlussier: So if you dislike a book's shelving location you slap a color changing charm on it? ;) |
| 16:08 |
kmlussier |
tsbere: Yes, exactly |
| 16:08 |
tsbere |
might be easier to charm the shelves to color the books when they are placed |
| 16:09 |
jeff |
revisiting this again... supercat mods record retrieval has proper encoding. all other mods variants (mods{3|32|33}) appear to produce double-encoded output. |
| 16:09 |
jeff |
it looks like the only reason mods works is becuase its xsl:output lacks an encoding. |
| 16:09 |
jeff |
the others specify UTF-8 and then we're presumably encoding it to UTF-8, or failing to decode it from UTF-8. |
| 16:12 |
jeff |
a similar but (i think) unrelated issue exists with SIP2 data based on record_to_mvr, at least under certain circumstances. |
| 16:17 |
jeff |
(i suspect OpenILS::SIP::clean_text there) |
| 16:18 |
jeff |
you can see the difference comparing the following two: |
| 16:18 |
jeff |
good title: https://demo.evergreencatalog.com/opac/extras/supercat/retrieve/mods/record/7071 |
| 16:18 |
jeff |
mangled title: https://demo.evergreencatalog.com/opac/extras/supercat/retrieve/mods32/record/7071 |
| 17:08 |
rhamby |
out |
| 17:23 |
|
artunit_away_ joined #evergreen |
| 17:27 |
|
dbs joined #evergreen |
| 17:27 |
|
dluch joined #evergreen |
| 17:27 |
|
hopkinsju joined #evergreen |
| 17:27 |
|
Bmagic joined #evergreen |
| 17:34 |
dbs |
jeff: more of the Encode.pm fallout I would guess |
| 18:17 |
jeff |
dbs: haven't ruled out that possibliity yet. :-) |
| 18:18 |
jeff |
there's some other fun with SIP2 messages having their length calculated incorrectly for POSIX::write and how some SIP2 clients are tolerant of that and some are not. |
| 18:19 |
jeff |
also, some clients that are tolerant of the length issue are not tolerant of "SIP server went away without replying to our last message (because Encode.pm crashed the SIP2 process)" |
| 18:44 |
jeff |
i looked at a bunch of this a while back, but seem to have made poor notes or stored my notes in a poor location. |
| 18:55 |
|
vlewis joined #evergreen |
| 19:26 |
|
bmills joined #evergreen |
