Time
S
Nick
Message
00:35 nanozz joined #dataverse
03:35 nanoz joined #dataverse
08:10 juancorr joined #dataverse
08:20 jri joined #dataverse
10:29 nanoz joined #dataverse
11:41 pdurbin joined #dataverse
13:28 poikilotherm joined #dataverse
13:54 isullivan joined #dataverse
14:24 pdurbin
poikilotherm: hi. Still unblocked?
14:27 poikilotherm
pdurbin: morning :-)
14:28 poikilotherm
I have a plan and will execute it... :-D Also it takes a while today. The day feels pretty clumsy so far.
14:29 pdurbin
plans are good :)
14:31 poikilotherm
I'm opting for the maven copies deps to the lib directory, so the war plugin simply strips out all jars and that kind of makes them "provided" scope like
14:32 poikilotherm
This can happen within a separate layer as far as I can see and thus the layer with the war itself is pretty small
14:33 poikilotherm
The WAR without deps is just 3.5MB
14:33 poikilotherm
Instead of errr... 184MB?
14:33 * poikilotherm
looks up github releases
14:34 poikilotherm
146MB
14:35 poikilotherm
As this will happen inside a separate maven profile, the other deployment approaches will not be broken, which I think is pretty important
14:40 poikilotherm
Oh I forgot. I am a bit blocked by the workaround for the javamail issue with the AWS SDK introduced in https://github.com/IQSS/dataverse/issues/3921 . As I strip out ALL jars, this plugin will crash the build and the bad jar will be within the deployment. Did you ever try to fix that by selecting the individual modules as described in 3921?
14:40 pdurbin
poikilotherm: wow, only 3.5 MB . That's great!
14:42 pdurbin
poikilotherm: I personally didn't try and the people who worked on this code aren't usually in IRC . You might want to ask in a new issue or on the dataverse-dev mailing list.
14:43 poikilotherm
Ok opening a GH issue seems to be the best way IMHO
14:43 pdurbin
sure, please go for it
14:48 cdsp-rmo joined #dataverse
14:57 poikilotherm
For cross references: done. https://github.com/IQSS/dataverse/issues/5274
15:02 pdurbin
poikilotherm: thanks, I haven't followed any of this, the workaround or whatever. Are you saying you want to remove the "remove-javamail-providers-from-exploded" stuff? Maybe you could add a diff to the issue showing the change you'd like to make (but that you know breaks stuff).
15:10 pameyer joined #dataverse
15:12 poikilotherm
Done :-)
15:12 poikilotherm
Good idea BTW
15:14 pdurbin
ooo, I like the use of "diff" as a language in the fenced code block
15:15 pameyer47 joined #dataverse
15:15 pdurbin
poikilotherm: by the way, bsilverstein was a intern for us for two summers but he hasn't been very involved in the code lately
15:15 poikilotherm
Sorry to hear that...
15:16 poikilotherm
Maybe that's the reason why this hasn't been adressed yet?
15:16 pameyer47
I'm not 100% sure that the email address his github notifications go to is one he's still got access to
15:16 pdurbin
Well, we aren't paying him anymore. He's busy with his new job. It's fine.
15:16 pdurbin
he works for pameyer47 now. sort of :)
15:16 pameyer47
yeah - he's keeping busy :)
15:17 pameyer47
not really - but we're in the same room some of the time
15:17 pdurbin
close enough
15:18 pdurbin
pameyer47: please tell him I'd like more Slack emojis. Maybe Davis with lion ears or something. He can pick.
15:19 isullivan left #dataverse
15:23 pameyer47
pdurbin: I'll pass that along
15:24 poikilotherm
Is there anybody else I should trigger with this? Or should I "just do it"?
15:24 pdurbin
poikilotherm: what's the risk of making a pull request? We'll see some failures in Travis?
15:25 poikilotherm
@ferrys seemd to be involved in #3921
15:25 poikilotherm
I would have to dig in the code if anything else from the Amazon SDK is used besides S3
15:25 pdurbin
She's another intern who grew up and got a job. :)
15:26 poikilotherm
The more you say, the more I think "just go and hack on it..." ;-)
15:26 pdurbin
seems pretty low risk
15:26 pdurbin
and maybe we'll learn something from the error in travis
15:27 poikilotherm
Uhm I can't just remove it - I will have to fix it. Things won't fail unless someone want's to send a mail... :-D
15:27 poikilotherm
https://github.com/IQSS/dataverse/issues/3921#issuecomment-320367631
15:28 pameyer47
I don't think the discussions about "how to integration test sending emails" ever turned into anything concrete
15:28 poikilotherm
Maybe matthew-a-dunlap should look into this, too?
15:28 pdurbin
he'd definitely be a good candidate to take a look
15:29 * poikilotherm
takes up the gun and starts firing at bugs... :-D
15:30 * pdurbin
notices how quickly the ammunition is disappearing
15:30 poikilotherm
Maybe I better take the spade and dig for them...
15:31 pdurbin
Would a 4th solution be to upgrade to a newer version of the SDK? That comment is over a year old.
15:31 poikilotherm
That would be a good thing anyway. I remember some comment in the pom.xml about the AWS version...
15:31 * poikilotherm
goes looking for it
15:32 poikilotherm
Ah here we go:
15:32 poikilotherm
<!-- Due to our old version of glassfish and its version of jackson,
15:32 poikilotherm
we are pointing to an aws sdk with unique references to needed libraries.
15:32 poikilotherm
If we update glassfish, we should return to the normal sdk as well. -->
15:32 poikilotherm
I love circle deps.
15:33 poikilotherm
I think I got a sword in my inventory to hit that knot.
15:33 pdurbin
Heh. Well, we do want to get off of old Glassfish, right?
15:34 pdurbin
Glassfish 4.1 came out 2014-09-09.
15:34 poikilotherm
I heard rumors about "a new fish in town". Parayara or something like that.
15:36 pdurbin
yeah, I have the t-shirt
15:36 poikilotherm
Ok, looks like I can simply switch to the s3 sdk module and thats it. quick grep for amazon only spit out S3 related stuff.
15:43 poikilotherm
pdurbin do you think it is alright to just do this inside my branch for 4172? Hopefully this is just a simple commit.
15:44 poikilotherm
(I would reference the new issue in the commit message)
15:47 nanoz joined #dataverse
15:50 poikilotherm
Gotta go - pdurbin please respond here, will read up the logs later today. Cu all tomorrow
16:07 pdurbin
Sure, in the same branch is fine with me. It's a small change. We can always pull it out later if we need to.
17:23 isullivan joined #dataverse
17:46 donsizemore joined #dataverse
17:47 bsilverstein joined #dataverse
17:48 bsilverstein
pdurbin: hello, hope all is well! word on the github street is I wrote some code w/ a fatty AWS dependency a while ago
17:48 pdurbin
bsilverstein: Marty!
17:49 pdurbin
Why are things so heavy in the future? Is there a problem with the Earth's gravitational pull?
17:49 pameyer47
bsilverstein: and there I was, saying you probably didn't see those notifications ;)
17:50 bsilverstein
pdurbin: there's that word again!
17:50 bsilverstein
pameyer47: hello from the window side of the office if you're at 1295 today
17:52 bsilverstein
how was this year's IQSS halloween party? do those still happen?
17:52 pdurbin
Davis won, as usual.
17:54 bsilverstein
fantastic news! what was his getup this year?
17:55 pdurbin
same as last year, I think
17:57 pdurbin
here you go: https://i.imgur.com/uXPcMy6.jpg
18:00 bsilverstein
phenomenal
18:43 barryr joined #dataverse
18:55 Jim__ joined #dataverse
20:34 pameyer joined #dataverse
21:16 pameyer
https://commons.apache.org/proper/commons-io/javadocs/api-release/org/apache/commons/io/IOUtils.html#closeQuietly-java .io .Closeable...-
21:17 pameyer
deprecation w\out replacement :(
22:38 pdurbin
Huh. Sure enough. "As of 2.6 removed without replacement. Please use the try-with-resources statement or handle suppressed exceptions manually."
22:38 pdurbin
at least they said "please"
22:45 pameyer
yeah - and silently ignoring all exceptions is kinda the same thing as ignoring the return values from functions
23:08 nanoz joined #dataverse
