Log of the #duraspace-ff channel on chat.freenode.net

Using timezone: Eastern Standard Time
* eddies leaves01:00
* eddies joins01:06
* eddies leaves
* eddies joins
* eddies leaves01:23
* eddies joins01:32
* eddies leaves
* eddies joins
* eddies leaves02:48
* eddies joins03:48
* eddies leaves
* eddies joins
* fasseg joins05:38
Hola guys!
Do you knowof some more installation instructions for RDFDatabank than the two files called "Install_VM..." and "install_debian"?05:39
these are not very helpful for a plain install on a diff linux
hmm ok ill try to parse the dabina install package in my mind then ;)05:50
python servers for downloading dependencies are awfully slow it seems05:59
hmm but install seems fairly straightforward from what i see in the debian install scripts06:11
would it make sense to write up a small installation instruction for a plain install if I'm succesful?06:12
Wargggh: ERROR: Failure while downloading apache-solr-4.0.0.tgz06:15
they kills me
quite the version hell right now...06:29
<eddies>fasseg: sounds like par for the course. definitely would be good to document the process06:49
<fasseg>especially since it's not that trivial if you've never worked with python, django and the like06:50
http://knowyourmeme.com/photos/234719-i-have-no-idea-what-im-doing06:51
it seems i got the webapp installed though, just the db and solr connection are still broken06:52
this file helps a lot: https://github.com/dataflow/RDFDatabank/blob/master/debian/postinst06:53
* eddies leaves07:19
<fasseg>Finally got it up and running....08:25
Gonna create a Wiki site on how to install plain
* eddies joins08:30
* eddies leaves
* eddies joins
* JasonDGI joins09:20
<fasseg>Ok here's a wki site about m installation experience: https://wiki.duraspace.org/display/FF/Task%3A+Installation+of+RDFDatabank+on+linux09:27
oops wrong category though
Ahh I don't have premission to move wiki items, so if anyone feels like moving these pages in a proper place: be my guest:09:28
https://wiki.duraspace.org/display/FF/Required+API+Endpoints+-+eSciDoc
<cbeer>eddies++09:46
fasseg: thanks for writing that up. i never did figure out paster10:23
one note (not that it matters now..), we should be using pip (and the distributed requirements.txt, which is broken currently) to install all those python libraries
rather than easy_installing them ourselves.
<fasseg>err ok...
<- python virgin here10:24
so whatever you think can be done better just put it on the site ;)
Im just guessing what paster is, never used it10:25
<cbeer>yeah. once we find out what repository we SHOULD be working against, i'll send a pull request
fasseg: ah, so you just accepted their default paths, configurations, etc?
<fasseg>nah actually i put it in "/opt/databank"
but i sed'ed over the py files ;)
<cbeer>got it. i got to that point, and gave up.10:26
<fasseg>but as i understand it paster should put the right values in the *.py files based on the ini files so it should work to put it in different directories10:27
<cbeer>that's kinda nice.. kinda scary.
<fasseg>well it had it frustrating moments :)
indeed
brb for the call10:28
<cbeer>ah, i see where you got the easy_install stuff from10:29
ah, and that's where they're hiding the unreleased RecordSilo eggs!
fasseg: for eSciDoc, are you using ingest with a FOXML payload, or just programmatically creating objects?10:30
(and, same for export. your goal is to get FOXML out?)10:31
<fasseg>FOXML
I don't know for export I'll have to check10:32
but ingest is FOXML
<cbeer>thanks. i might take your list and try to break out some of confused api endpoints (ingest w/ FOXLM vs just creating the object; export with all its different flavors; etc)10:33
so we actually know what we're trying to support for an MVP
<fasseg>sure go ahead10:34
It's still a work in progress though, since the eSciDoc didn't respond to my inquiry yet10:35
*eSciDoc guys
so I might have missed something...10:36
btw: committer call?
only Eddie, Kai and myself there ;)10:37
<JasonDGI>did the call move, or are we still having it now?10:58
<fasseg>Committer call did move to 10.30, but Scrum meeting is now i think..10:59
<cbeer>JasonDGI: just getting on now.
ok, in.11:00
<fasseg>Has this been booked by the way, since I need to know for the ESTA application: http://www.vrbo.com/71811?11:03
* barmintor joins11:08
<cbeer>eddies: ok, looks like the secretKey and accessKey are sufficient for glacier access.. maybe nothing else, but that's fine.11:09
<fasseg>dropped and line is offline again :/11:10
<JasonDGI>same here
<fasseg>can someone bridge me in the call pls?
I also have a mumble server up and running ;)11:11
<JasonDGI>i have a mumble client
<fasseg>https://wiki.duraspace.org/display/FF/Task%3A+Installation+of+RDFDatabank+on+linux11:15
https://wiki.duraspace.org/display/FF/Required+API+Endpoints+-+eSciDoc
https://github.com/escidoc11:21
<barmintor>fasseg: just question about whether there is some kind of object-mapping for Fedora objects in eSciDoc11:23
<cbeer>barmintor: for active-fedora, i think we should leave it in projecthydra.. whatever ruby-level databank client should go into futures. agree?
or, ideally, we won't need to change active-fedora at all.. at least in any backwards-incompatible way11:24
<fasseg>should i create a Story for such a research task?11:25
<barmintor>cbeer: I would put the databank library into futures. I don't know that active-fedora needs to change, but hyhead might
<cbeer>barmintor: databank + solr4 should be in thecurrent iteration, right?11:36
<barmintor>cbeer: yes
<cbeer>fasseg: https://github.com/futures/ff-modeshape-jmeter11:37
fasseg: https://www.pivotaltracker.com/story/show/42293255 is updated.11:40
<fasseg>ok cool now I really know what to do ;)11:42
was about to ask ;)11:43
<cbeer>CONFIGURATION CHANGED11:46
* eddies is a crazy product owner.11:48
<JasonDGI>+1
<fasseg>Chris can you give me your skype name?11:49
and ben...?11:50
I'm "frank.asseg"
* fasseg leaves11:54
<barmintor>back in a minute, grabbing a sandwich11:56
* fasseg joins11:57
<eddies>frank, i'm creating a skype group w/ everyone11:58
<fasseg>FedoraMadness.jmx <- I like that!
<eddies>actually, i think i might not have chris
<fasseg>k eddie thx
but im off now gf came home...see you tomorrow11:59
<eddies>cbeer: send me your skype handle when you have a sec
* fasseg leaves
<cbeer>eddies: c.a.beer12:11
barmintor: this was new-to-me (last updated the end of last month): http://hessbrewing.blogspot.com/2009/11/nanobreweries-in-usa.html12:16
<barmintor>cbeer: cool. B&T is one of the 3.5 breweries that are running out of Queens right now12:19
<cbeer>eddies: not sure whether you wanted me to finish https://www.pivotaltracker.com/story/show/42231937 or https://www.pivotaltracker.com/story/show/42231339 after adding the hybrid use case object13:21
i just harvested a whole IA video object (minus a large MOV)13:22
<eddies>cbeer: i didn't see 42231339 earlier. so i'd say 42231937 was actually finished with your sample image object. and let's call the latter ticket tentatively finished if it's a reasonable approximation of a hybrid use case object.13:28
<cbeer>k13:31
so, ready for you to deliver.13:32
man, we need zoia in here.. i really want to @later something to frank. guess i should just stop being lazy and send an email13:44
jmeter--13:45
i don't want to write java to look through directories13:46
<eddies>cbeer & barmintor: anything that needs to be taken care of for using columbia's jenkins or is that indeed all sorted out?13:52
<barmintor>afaik, osrted out
<cbeer>eddies: i think the only outstanding question is if we can add shell accounts for people.. but we don't really need them, imo.
<eddies>ok. thanks13:53
<barmintor>cbeer: do you have a solr.xml file for solr4 that doesn't use cores?14:25
<cbeer>just use cores, but set a default core
i can find that configuration for you in a sec.
struggling with jmeter.14:26
<barmintor>don't bother, I've got it
<cbeer>k. there's no reason not to use cores.14:35
cores are your friends.
barmintor: do you understand loops in jmeter?14:36
<barmintor>cbeer: the only reason I was trying not to was to minimize the changes to databank at this stage
cbeer: no, don't know loops in jmeter
<cbeer>yeah, but it's not worth trying to not do multicore.
<barmintor>agreed14:37
<cbeer>ok, thrashing on https://www.pivotaltracker.com/story/show/42293231.. giving up.15:01
and increasing its point value. clearly it's not something jmeter wants to make easy.15:02
guess it's back to the glacier mines for me.
i guess that means figuring out ajs6f's relation-based hints.15:03
<barmintor>If those hints are too obtuse, we should change the akubra api to make it easier. It's not a 1.0 API yet anyway.15:10
* JasonDGI leaves15:27
* eddies leaves18:01
<cbeer>jmeter--18:20
* eddies joins21:04
* eddies leaves
* eddies joins
* eddies leaves21:15
* eddies joins21:57
* eddies leaves
* eddies joins
* eddies leaves22:05