| Time |
Nick |
Message |
| 04:30 |
pinesol_green |
News from qatests: Test Success <http://testing.evergreen-ils.org/~live> |
| 06:40 |
|
rlefaive joined #evergreen |
| 07:29 |
|
agoben joined #evergreen |
| 08:48 |
|
bos20k joined #evergreen |
| 08:50 |
|
mmorgan joined #evergreen |
| 08:50 |
|
Dyrcona joined #evergreen |
| 09:00 |
|
kmlussier joined #evergreen |
| 09:09 |
|
rlefaive joined #evergreen |
| 09:20 |
|
JBoyer joined #evergreen |
| 09:25 |
|
maryj joined #evergreen |
| 09:35 |
kmlussier |
Good morning #evergreen |
| 09:35 |
kmlussier |
@coffee [someone] |
| 09:35 |
* pinesol_green |
brews and pours a cup of Decaf Organic Caffe Volcan, and sends it sliding down the bar to eady |
| 09:36 |
kmlussier |
@tea [someone] |
| 09:36 |
* pinesol_green |
brews and pours a pot of Wild Snow Sprout Tea, and sends it sliding down the bar to eady (http://ratetea.com/tea/wild-tea-qi/wild-snow-sprout-tea/6447/) |
| 09:36 |
csharp |
@praise [coffee] |
| 09:36 |
* pinesol_green |
brews and pours a cup of El Salvador Pacamara Finca Los Alpes The Bank, and sends it sliding down the bar to csharp is the hardest working person in #evergreen. |
| 09:36 |
kmlussier |
pinesol_green is playing favorites today. |
| 09:36 |
pinesol_green |
kmlussier: Yeah, well, you know, that's just like uh, your opinion, man. |
| 09:36 |
pinesol_green |
kmlussier: I am only a bot, please don't think I'm intelligent :) |
| 09:37 |
* dbs |
needs to intercept a caffeinated coffee |
| 09:37 |
dbs |
@coffee me |
| 09:37 |
* pinesol_green |
brews and pours a cup of Guatemala El Injerto Pacamara, and sends it sliding down the bar to me |
| 09:37 |
dbs |
DAMN YOU |
| 09:37 |
JBoyer |
dbs++ |
| 09:38 |
kmlussier |
@coffee dbs |
| 09:38 |
* pinesol_green |
brews and pours a cup of Organic Colombia, and sends it sliding down the bar to dbs |
| 09:39 |
pinesol_green |
[evergreen|Remington Steed] Docs: Fix docs build warnings - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=cf182c5> |
| 09:40 |
kmlussier |
I haven't looked at bug 1681943 yet, but I'm tempted to think of it as a bug fix since the My Lists interface is so broken on mobile devices as it stands now. |
| 09:40 |
* csharp |
grabs dbs's decaf before it gets too cold |
| 09:40 |
pinesol_green |
Launchpad bug 1681943 in Evergreen "Improve Responsive Design in My Lists" [Wishlist,Confirmed] https://launchpad.net/bugs/1681943 |
| 09:43 |
dbs |
kmlussier++ # much obliged |
| 09:46 |
* kmlussier |
now wanders off to get a for real cup of coffee. |
| 09:48 |
dbs |
also I would concur that would be a bug fix; there appear to be no string changes |
| 09:50 |
Dyrcona |
I'm OK with the consensus on whether or not it is a bug fix. |
| 09:53 |
Dyrcona |
Y'know, should we reiterate that it would be nice if non-devs would add the whole "I consent to signing off on this branch...." if they've looked at it and think it works? |
| 09:53 |
Dyrcona |
I've seen two comments this week that basically say, "looks good to me." |
| 09:54 |
Dyrcona |
It would be better if they'd just signoff on the branch. |
| 09:55 |
Dyrcona |
But, I don't think either of the people in question have a key in the repo. (Well, I actually know that they don't.) |
| 09:55 |
Dyrcona |
Oh, wait, one of them does. :) |
| 09:57 |
pinesol_green |
[evergreen|Kyle Huckins] LP#1621178 Copy Status Picker - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=d57d767> |
| 09:57 |
pinesol_green |
[evergreen|Kathy Lussier] LP#1621178: Tweaks to item status column picker option - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=3945de1> |
| 09:57 |
pinesol_green |
[evergreen|Galen Charlton] LP#1621178: fix fleshing of copy status for available items - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=5957d7c> |
| 10:03 |
kmlussier |
Dyrcona: When I've seen that in the past, I've added a comment to the LP bug in question asking the person to do the whole "I consent to signing off..." thing. |
| 10:03 |
* kmlussier |
doesn't recall anything coming through recently on that. |
| 10:05 |
kmlussier |
Looking at dbs' comment on bug 1681943, I think we also need to do a lot of education around the rtl stylesheet. I know I've missed it in a couple of changes I've made. |
| 10:05 |
pinesol_green |
Launchpad bug 1681943 in Evergreen "Improve Responsive Design in My Lists" [Wishlist,Confirmed] https://launchpad.net/bugs/1681943 |
| 10:05 |
kmlussier |
Maybe it would be worth an email to the dev list? |
| 10:09 |
Dyrcona |
Yeah. I was thinking of commenting on the bug, but in a meeting now. |
| 10:20 |
|
jvwoolf joined #evergreen |
| 10:27 |
* dbs |
has a > 50% merge of the RTL stylesheet into the regular stylesheet done |
| 10:27 |
dbs |
which would help *a lot* |
| 10:27 |
dbs |
bug 1681009 |
| 10:27 |
pinesol_green |
Launchpad bug 1681009 in Evergreen "RTL stylesheets should be merged with LTR stylesheets" [Undecided,New] https://launchpad.net/bugs/1681009 |
| 10:27 |
dbs |
*not* spreadsheet :) |
| 10:28 |
kmlussier |
:D |
| 10:29 |
bshum |
I was just thinking that we should merge dbs' fix first for combining RTL and LTR and then work on adding work for new changes in other branches |
| 10:29 |
gmcharlt |
@coffee dbs |
| 10:29 |
* pinesol_green |
brews and pours a cup of Kenya Ndiara, and sends it sliding down the bar to dbs |
| 10:29 |
bshum |
So that we don't end up doing extra work twice |
| 10:31 |
kmlussier |
Oh, I see what we're doing there. |
| 10:32 |
kmlussier |
dbs: Would you like help in moving those styles over? |
| 10:45 |
dbs |
kmlussier: given that I havenèt touched that branch since the conference, absolutely |
| 10:46 |
dbs |
Make that a standing invite to anyone to contribute patches to branches :) |
| 10:54 |
gmcharlt |
berick: thinking aloud re bug 1685356 - it occurred to me that maybe we should split egOrgSelector into two directives |
| 10:54 |
pinesol_green |
Launchpad bug 1685356 in Evergreen "egOrgSelector's notion of a change event is eccentric " [Undecided,New] https://launchpad.net/bugs/1685356 |
| 10:55 |
gmcharlt |
with egOrgSelector itself having its purpose narrowed to being a widget that you use to select an OU and have something happen (e.g., change a grid's filter or the like) |
| 10:56 |
gmcharlt |
and create a new egOrgInput that is meant to be used in forms and is based on select rather than uib-dropdown |
| 11:05 |
berick |
gmcharlt: so if we have a select-based org picker that works well with forms, any reason not to just use that everywhere? |
| 11:05 |
berick |
trying to understand what each provides |
| 11:05 |
gmcharlt |
berick: in the short term, an easier transition |
| 11:06 |
berick |
OK so goal may be to deprecate egOrgSelector, but time will tell? |
| 11:06 |
gmcharlt |
in the long run, there may not be much reason to have two, beyond a possible use case of wanting to have an OU selector in the form that manipulates the form display without marking the form as dirty |
| 11:07 |
gmcharlt |
^^ I'm stretching here, of course |
| 11:09 |
berick |
right |
| 11:09 |
berick |
could also be a toggle/setting |
| 11:09 |
berick |
having something play nice w/ forms without a lot of hackery seems like a reasonable goal |
| 11:11 |
gmcharlt |
OK; I'll run with it under the name egOrgInput; once I've got something working, we can evaluate where it fits overall |
| 11:14 |
* berick |
nods |
| 11:17 |
|
rlefaive_ joined #evergreen |
| 11:36 |
|
Christineb joined #evergreen |
| 11:43 |
kmlussier |
jeff: I don't know if I'm misremembering, but did you mention that you're working on the bill referral feature for 3.0? |
| 11:49 |
rhamby |
gmcharlt++ : for calling for updated templates |
| 11:49 |
gmcharlt |
tw |
| 12:06 |
|
mmorgan1 joined #evergreen |
| 12:06 |
Dyrcona |
Anyone know how the webstaff client http proxy setups work with a load balancer? |
| 12:08 |
|
jihpringle joined #evergreen |
| 12:15 |
|
yboston joined #evergreen |
| 12:27 |
jeff |
kmlussier: yes, and i'm hoping to find time to reply on-list to the at least one inquiry that i saw earlier this week. |
| 12:27 |
bshum |
Dyrcona: Maybe it's worth looking at using nginx as a load balancer: http://nginx.org/en/docs/http/load_balancing.html |
| 12:27 |
kmlussier |
jeff: heh, that's why I was asking. I was just going through my vacation email and was wondering if you had seen it. :) |
| 12:28 |
bshum |
Dyrcona: Or maybe looking at how ldirectord (assuming that's what you're using) can/can't do the proxying for us for websockets too. |
| 12:36 |
|
mmorgan joined #evergreen |
| 12:37 |
bshum |
Course now the more I google, the more I think I should learn about HAProxy and how that works |
| 12:38 |
* bshum |
imagines a crazy world where existing LB pointed at an nginx instance that could point at multiple apache backend servers to do its thing |
| 12:39 |
bshum |
Or cutting out LB and using nginx direct. Or using haproxy to do everything (SIP?) |
| 12:40 |
Dyrcona |
bshum: I'll look into it. |
| 12:40 |
Dyrcona |
"It" meaning those options. |
| 12:42 |
dbs |
bshum: yeah, I think getting nginx to serve up the static assets would help reduce the load on Apache; and judicious caching could help too (mayyyybe) |
| 12:43 |
bshum |
dbs: Right, but the stock nginx example would only serve a single apache instance backend, right? So if we had an LB in front of it sending you between multiple nginx and individual apache, I'm not sure what that would do to session switching between multiple nginx instances |
| 12:44 |
bshum |
Hence, my thinking to cut out the middle man there and use just one nginx instance talking to mulitple backend apaches as its own load balancer |
| 12:44 |
bshum |
But I'm just speculating |
| 12:45 |
bshum |
By one nginx, I of course mean try setting it up as a highly available setup with more than one nginx node... |
| 12:45 |
bshum |
but anyways :D |
| 12:47 |
dbs |
bshum: no i mean that too |
| 12:47 |
dbs |
Just using nginx directly, with HTTP/2, talking to multiple apaches if necessary, but also taking care of serving up static content |
| 12:49 |
Bmagic |
Is there a wishlist item for making newly cataloged items include in related hold_copy_maps without the hold targeter? |
| 12:50 |
bshum |
Bmagic: You mean without the flag that kicks off in checkin modifier for doing a retarget? |
| 12:50 |
mmorgan |
Bmagic: There is one in my head if nowhere else. |
| 12:51 |
Bmagic |
bshum: The checkin modifier does this? |
| 12:51 |
mmorgan |
The one in my head would apply to any item that gets saved into a holdable state. |
| 12:51 |
Bmagic |
I thought the hold targeter was the only* thing that put copies in the hold_copy_map |
| 12:52 |
bshum |
Bmagic: My recollection was that if you used the checkin modifier for retarget local holds or whatnot on new copies it would make an attempt to do so on first-checkin of a newly cataloged item |
| 12:52 |
bshum |
Of course that had a tendency to break on items with too many holds to target through due to timeout |
| 12:52 |
bshum |
Though I wonder how that'll behave with the new v2 hold targeter taking over someday |
| 12:52 |
mmorgan |
Bmagic: You can perform manual retargets in the client, with the checkin modifier, and from holds themselves. |
| 12:53 |
Dyrcona |
bshum: If you checkin the new item with retarget local holds, it does it immediately. |
| 12:53 |
Bmagic |
even if the copy doesn't exist in action.hold_copy_map ? |
| 12:53 |
Dyrcona |
Retargeting builds the hold_copy_map. |
| 12:53 |
kmlussier |
Yeah, our people use the holds targeter, but the wishlist item Bmagic asked about is something people ask me about all the time. They prefer not to use the checkin modifier. |
| 12:54 |
kmlussier |
s/use the holds targeter/use the checkin modifier |
| 12:54 |
Bmagic |
well, I didn't realize the checkin modifier solves the issue. I will pass that along |
| 12:54 |
Dyrcona |
I understand what Bmagic is asking, but it would require develop, possibly tied to as status or copy location change. |
| 12:54 |
Dyrcona |
You have to check new items in most ILSes to capture/trigger holds. It's a common workflow. |
| 12:55 |
Dyrcona |
Not saying it's good, necessarily. |
| 12:55 |
Dyrcona |
It would be nice if I could type complete words. |
| 12:55 |
mmorgan |
Bmagic: The checkin modifier works for local holds only, BTW. |
| 12:55 |
Dyrcona |
I think my brain works faster than my fingers. :) |
| 12:55 |
kmlussier |
Dyrcona: I think the desire for not using the checkin modifier is partially due to the fact that checkin for popular items time out. Also, it only captures local holds. |
| 12:56 |
Dyrcona |
mmorgan Bmagic: Right, because checking all holds would definitely take too long. |
| 12:56 |
kmlussier |
I'm hoping berick's holds work will address the timeout issue. |
| 12:56 |
bshum |
That'd be cool |
| 12:56 |
bshum |
Addressing the timeout issue I mean |
| 12:56 |
Dyrcona |
kmlussier: I think checkin does too much as it is, but I haven't thought about how to break it all up. |
| 12:56 |
Bmagic |
I see, there are some issues with it but at least there is something |
| 12:57 |
Dyrcona |
Timeout depends on a lot of variables, very few of which are controlled. |
| 12:57 |
kmlussier |
I don't think that workflow is documented well either. The same question came up during the Wait Wait Do Tell Me Session. A lot of people seem to be unaware of what its purpose is. |
| 12:57 |
kmlussier |
By its, I mean the checkin modifier. |
| 12:58 |
Dyrcona |
Yes, even at MVLC staff didn't know what it was for. :) |
| 12:58 |
kmlussier |
We should share a weekly tip to the general list. |
| 12:59 |
* mmorgan |
would love it if the checkin modifier was no longer necessary. |
| 12:59 |
Bmagic |
mmorgan: wishlist item on LP doesn't exist? |
| 13:00 |
kmlussier |
Bmagic: I don't remember seeing one. |
| 13:01 |
Dyrcona |
One fix is to allow on order copies to fill holds, but that isn't what you want either. |
| 13:01 |
* mmorgan |
was thinking about opening one, but doesn't think it actually happened :-( |
| 13:01 |
Dyrcona |
You want a status that can target a hold but not fill it, i.e. end up in hold_copy_map, but not in target_copy. |
| 13:01 |
mmorgan |
Another observation about the checkin modifiers, two checkins are required if the item is in a non-holdable status. |
| 13:02 |
Dyrcona |
Maybe the first use of target was a bad choice of words. :) |
| 13:07 |
|
Newziky joined #evergreen |
| 13:07 |
|
Newziky left #evergreen |
| 13:12 |
jeffdavis |
Dyrcona: re HTTP proxy, right now we've got a load balancer that passes requests to one of several servers running EG + Apache. All I'm planning to do initially is run nginx separately on each of the servers in front of Apache; I don't expect this to make any difference from the LB's point of view. |
| 13:13 |
Dyrcona |
jeffdavis: I thought that's how I'd set it up, too, but then I wondered if there would be a better way. |
| 13:14 |
bshum |
We might want to read more into nginx ssl session caching options. First link I grabbed to read, for example: https://auxbuss.com/blog/posts/2011_06_28_ssl_session_caching_on_nginx/ |
| 13:14 |
Dyrcona |
We're about to replace our load balancer hardware, so I'm looking at options. |
| 13:15 |
bshum |
jeffdavis: I only mention ssl session cache for the resumption options so that you don't have to spend too much time renegotiating from scratch every time you skip around app servers |
| 13:15 |
bshum |
We did something similar for apache for multi-server to speed up deliver of things |
| 13:15 |
bshum |
So if you put multiple nginx in front, same deal. Maybe. |
| 13:16 |
* bshum |
is still reading up on the subject area to learn more |
| 13:16 |
mmorgan |
Bmagic: lp 1686463 |
| 13:16 |
pinesol_green |
Launchpad bug 1686463 in Evergreen "Wishlist: Background targeting of holds when items are edited into a holdable state" [Undecided,New] https://launchpad.net/bugs/1686463 |
| 13:17 |
Bmagic |
mmorgan++ |
| 13:18 |
mmorgan |
It has been brewing in my head for a while, but I'm encouraged now with the improvements to the hold targeter. |
| 13:37 |
* dbs |
has also been contemplating pulling the OpenLibrary AC lookups and OpenURL lookups out of TPAC server code and back into JavaScript to improve catalogue performance |
| 13:37 |
dbs |
I did a lot of experimentation with nginx in front of apache for the PWA explorations |
| 13:50 |
jeffdavis |
bshum: in our setup, certs are installed on the load balancer and so the external client does the handshake with the LB rather than with individual app servers. I assume HAProxy does something like session caching for the internal connections between the LB and our app servers ... but I think I'll ask about that. :) |
| 13:51 |
dbs |
Even ripping out as much Dojo as I can and cutting out the OpenLibrary AC lookups, our server seems to have trouble serving up the core HTML in under 2 seconds, which isn't great :/ |
| 13:51 |
dbs |
even webby takes > 1 second to get the HTML out: https://www.webpagetest.org/result/170426_MH_1EYA/ |
| 13:52 |
dbs |
so if we can cut that down _and_ get rid of Dojo _and_ use HTTP/2 to multiplex requests _and_ use cache-busting to allow lots of caching we can make some good progress |
| 13:52 |
bshum |
jeffdavis: Neat, I wondered about that approach too. |
| 13:53 |
dbs |
6 seconds to interactive on a Fibre connection, yikes. |
| 13:54 |
dbs |
err, s/Fibre/cable/ but still |
| 13:56 |
dbs |
meanwhile our production server has something *crazy* happening on repeat views that makes the HTML take forever to get out: https://www.webpagetest.org/result/170426_WZ_1F2T/ |
| 14:00 |
|
jvwoolf joined #evergreen |
| 14:03 |
jeffdavis |
dbs: I wonder if there is useful overlap between JS-based added content lookups and ebook API. |
| 14:04 |
dbs |
jeffdavis: I was thinking it might roll into the existing Novelist etc JS-based lookups |
| 14:22 |
* kmlussier |
keeps reaching for a cup of tea that doesn't exist. |
| 14:22 |
berick |
@tea [eightball will kathy get tea?] |
| 14:22 |
* pinesol_green |
brews and pours a pot of Masala Chai, and sends it sliding down the bar to NO! (http://ratetea.com/tea/rishi/masala-chai/4495/) |
| 14:23 |
* berick |
chuckles |
| 14:25 |
kmlussier |
lol |
| 14:25 |
Dyrcona |
heh |
| 14:26 |
Dyrcona |
@tea kmlussier |
| 14:26 |
* pinesol_green |
brews and pours a pot of BH02: Holy Basil Purple Leaf, and sends it sliding down the bar to kmlussier (http://ratetea.com/tea/upton/bh02-holy-basil-purple-leaf/1937/) |
| 14:26 |
Dyrcona |
"Holy Basil Purple Leaf" sounds like Fawlty Towers meets Cheech and Chong. :) |
| 14:40 |
pinesol_green |
[evergreen|Jeff Davis] LP#1681466: Text for web client opt-in dialog - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=9ff3c37> |
| 14:40 |
pinesol_green |
[evergreen|Galen Charlton] LP#1681466: move text to body of confirm dialog and add title - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=3de90c3> |
| 14:44 |
berick |
well now I can't remember.. 'needsrepatch' means we should remove 'pullrequest' right? |
| 14:45 |
Dyrcona |
Probably. |
| 14:45 |
Dyrcona |
Yes. |
| 14:45 |
gmcharlt |
hmm... yeah, it should |
| 14:45 |
* Dyrcona |
changed his mind. It does mean that. |
| 14:45 |
kmlussier |
berick: Yes. I used to keep the pullrequest on, but then people would keep testing it on Bug Squashing Day. |
| 14:45 |
berick |
ok, good, thanks everyone. I dun good. |
| 14:45 |
gmcharlt |
berick++ # granular bugging |
| 14:45 |
* kmlussier |
pats berick on the back. |
| 14:46 |
berick |
heh. i needed that |
| 14:47 |
Dyrcona |
:) |
| 14:57 |
jeffdavis |
Strange, I seem not to get emails from Launchpad on bugs I'm subscribed to, *except* for private security bugs. |
| 14:58 |
Dyrcona |
jeffdavis: There was trouble with Lp sending emails yesterday, but I've not noticed anything today. |
| 15:04 |
jeffdavis |
It's been that way for a while for me. I even tried changing the email address on my account to no effect. Hopefully just some trivial local issue that I'll figure out now that I've mentioned it here. ;) |
| 15:13 |
|
maryj joined #evergreen |
| 15:21 |
|
mllewellyn joined #evergreen |
| 15:32 |
bshum |
jeffdavis: Well as a bug wrangler member, i would have expected you to get emails about all bugs traffic |
| 15:32 |
bshum |
Do you have any local filtering going on? |
| 15:34 |
bshum |
jeffdavis: When you say that you're subscribed to, are you manually adding yourself to bugs? |
| 15:36 |
bshum |
jeffdavis: Maybe double check what your settings say for subscriptions for LP Evergreen: https://bugs.launchpad.net/evergreen/+subscriptions |
| 15:36 |
bshum |
And make sure your subscription as a bug wrangler is set to receive an email when any change is made or a comment is added |
| 15:36 |
bshum |
Just to verify that you're setup to receive correctly |
| 16:00 |
jeffdavis |
I think that's it - I have Bug Wranglers mail disabled, and I bet that means that instead of just sending me mail for bugs I've manually subscribed to, it's not sending mail at all (except for the Security team). |
| 16:01 |
bshum |
jeffdavis: Sounds like that might be the problem, yup |
| 16:01 |
bshum |
Maybe |
| 16:04 |
|
mmorgan1 joined #evergreen |
| 16:15 |
|
Jillianne joined #evergreen |
| 16:30 |
pinesol_green |
News from qatests: Test Success <http://testing.evergreen-ils.org/~live> |
| 17:12 |
|
jvwoolf left #evergreen |
| 17:26 |
|
jihpringle joined #evergreen |
| 18:28 |
|
alynn26 joined #evergreen |
| 19:07 |
|
Dyrcona joined #evergreen |
| 22:20 |
|
genpaku joined #evergreen |
