Friday, 2015-11-06

*** bgeorges <bgeorges!~bgeorges@> has joined #wildfly-swarm00:00
*** bgeorges <bgeorges!~bgeorges@> has quit IRC (Ping timeout: 252 seconds)00:07
*** jcrossley3 is now known as jcrossley3-away00:59
*** bgeorges <bgeorges!~bgeorges@> has joined #wildfly-swarm01:03
*** helio-frota <helio-frota!> has quit IRC (Quit: Leaving)01:05
*** rbenevides|Afk <rbenevides|Afk!~rbenevid@redhat/jboss/RafaelBenevides> has quit IRC (Quit: Quit)01:07
*** bgeorges <bgeorges!~bgeorges@> has quit IRC (Ping timeout: 240 seconds)01:10
*** jclingan <jclingan!> has joined #wildfly-swarm01:33
*** bgeorges <bgeorges!~bgeorges@> has joined #wildfly-swarm02:07
*** bgeorges <bgeorges!~bgeorges@> has quit IRC (Remote host closed the connection)02:08
*** bgeorges <bgeorges!~bgeorges@> has joined #wildfly-swarm02:08
*** enebo <enebo!> has quit IRC (Quit: enebo)02:45
*** jerr <jerr!> has quit IRC (Quit: jerr)03:10
*** jclingan <jclingan!> has quit IRC (Quit: jclingan)04:51
*** jclingan <jclingan!> has joined #wildfly-swarm04:53
*** jclingan <jclingan!> has quit IRC (Quit: jclingan)05:12
*** hbraun <hbraun!~hbraun@pdpc/supporter/professional/hbraun> has joined #wildfly-swarm07:54
*** jnetherton <jnetherton!> has joined #wildfly-swarm08:43
*** sblanc|bbl is now known as sblanc10:04
*** hbraun is now known as hbraun_afk10:43
*** hbraun_afk is now known as hbraun_away10:44
*** hbraun_away is now known as hbraun10:51
*** sblanc is now known as sblanc|lunch11:44
*** sblanc|lunch is now known as sblanc12:17
*** johnament <johnament!ad3dbe50@gateway/web/freenode/ip.> has joined #wildfly-swarm12:27
*** johnament <johnament!ad3dbe50@gateway/web/freenode/ip.> has quit IRC (Quit: Page closed)12:40
*** bobmcw <bobmcw!~bobmcw@redhat/jboss/bobmcw> has joined #wildfly-swarm13:00
*** ChanServ sets mode: +o bobmcw13:00
*** bbrowning_away is now known as bbrowning13:12
*** helio-frota <helio-frota!> has joined #wildfly-swarm13:22
*** helio-frota <helio-frota!> has quit IRC (Client Quit)13:25
*** hbraun is now known as hbraun_away13:25
*** tcrawley-away is now known as tcrawley13:26
sblancHi all, I would like to check if no one has started on "adding fraction to pom" for the swarm forge addon ? I started to play with that this night (yeah jetlag is terrible) and if you are okay I can continue on it13:29
sblancquick teaser
sblancit's using the list produced by wildfly-swarm-fraction-list // thx bobmcw for that ;)13:30
*** jcrossley3-away is now known as jcrossley313:31
*** lance|afk is now known as lanceball13:33
lanceballsblanc: as far as I know, nobody has done any work in that area yet - but bobmcw probably has the definitive answer13:35
sblanclanceball: thx13:38
*** kenfinnigan <kenfinnigan!~ken@redhat/jboss/kenfinnigan> has joined #wildfly-swarm13:57
*** bgeorges is now known as bgeorges_afk14:16
*** bgeorges_afk <bgeorges_afk!~bgeorges@> has quit IRC (Quit: Leaving...)14:17
*** os890 <os890!> has quit IRC (Ping timeout: 250 seconds)14:41
*** tcrawley is now known as tcrawley-away14:43
bobmcwsblanc: I've done nothing other than the fraction-list.txt14:53
*** tcrawley-away is now known as tcrawley14:54
bobmcwgastaldi did the first round of the add-on, but beyond, I think you're The Man :)14:54
bobmcwsblanc: there'd been discussion of making the fraction-list enumerate dependencies also14:54
bobmcwdunno if that'd help your UI, but if you pick, say, ribbon, that implies -clustering (or now -jgroups)14:55
bobmcwlikewise, a lot of things imply -logging14:55
bobmcwor -security14:55
bobmcwor -io14:55
sblancbobmcw: Okay, make sense, I got the the addon working now (it adds the fraction dep to the pom) , if you add the deps of a fraction in the list, will be really easy to add this as well in the pom15:01
bobmcwreally, it'd be removing stuff from the POM, since maven will drag it transitively15:02
bobmcwthe UI would be able to make that clear15:02
bobmcw"You don't have to pick -security, because -ejb already brings it in"15:02
sblancahhh sorry , this way ok , make sense15:03
bobmcwand -security doesn't have to added to he user's pom.xml, since -ejb transitives it15:03
sblancwe can make it nice on the UI, you pick -ejb , and -security automatically get "selected" and grayed out15:05
*** tcrawley is now known as tcrawley-away15:06
*** tcrawley-away is now known as tcrawley15:08
*** tcrawley is now known as tcrawley-away15:13
*** tcrawley-away is now known as tcrawley15:14
*** jcrossle_ <jcrossle_!> has joined #wildfly-swarm15:15
*** jcrossley3 <jcrossley3!> has quit IRC (Ping timeout: 252 seconds)15:16
*** os890 <os890!> has joined #wildfly-swarm15:19
*** jcrossle_ is now known as jcrossley315:21
lanceballbobmcw: do you WANT gh-pages?15:24
lanceballor do you want something else15:24
bobmcwI want something easy that CI can build and push when we commit15:24
bobmcwwith a staging/production(master) branch like we did torquebox.org15:25
lanceballbobmcw: sure, but I mean push where?15:25
bobmcwif we need to host at contegix, okay, else gh-pages works for me also15:25
lanceballok - so staging/production means no gh-pages15:25
bobmcwactually, I'm okay w/o staging really15:25
bobmcwsince if we author in adoc, we can "stage" it on GH itself15:25
bobmcwand only publish to gh-pages when we push to production branch or such15:25
bobmcwconsidering we really only use staging for proof-reading blogs etc15:26
bbrowninglanceball: well you could do staging / production w/ gh-pages I think?15:26
bbrowningdifferent branches can use different hosts?15:26
bbrowningor it would require separate repos?15:26
lanceballI think it would require separate repos15:26
bobmcwso, I'm okay with `git push origin master:production` -> CI -> gh-pages15:26
*** enebo <enebo!> has joined #wildfly-swarm15:26
bobmcwand `git push origin master` just committing, and making readable on browsery15:27
lanceballbobmcw: fwiw, I really kind of like metalsmith15:27
bbrowninglanceball: huh - I would have guessed separate branch w/ different CNAME in each15:27
lanceballoh, mayhap that's possible15:27
* lanceball shrugs15:27
bobmcwlanceball: bonus points for setting up CI15:28
bbrowningit would take some discipline to keep the CNAME separate on each though15:28
bbrowningbecause the repos wouldn't be identical15:28
bbrowningunless you .gitignore CNAME perhaps15:28
lanceballI need to write a little script to build and shuffle onto the gh-pages branch15:28
bobmcwalso, do note that we'll need to do the GoldenGate header crap-ola at some point15:32
bobmcwand google analytics etc15:32
bobmcwbut we can wait until pete/ploski whinge15:32
lanceballyeah - what I have now was just getting a metalsmith thing going15:32
lanceballmetalsmith is pretty neat15:33
lanceballguess I said that already15:33
bobmcwyah, the GG is just a heads-up, ignore until it becomes un-ignorable :)15:33
jbossbotTitle: TinyGrab - Simple. Screenshot. Sharing.15:34
bobmcwthat header, and all the google analytics and other tracking bits marketing likes15:34
lanceballright - just didn't make the acronym connection15:43
*** rbenevides <rbenevides!~rbenevid@redhat/jboss/RafaelBenevides> has joined #wildfly-swarm16:07
*** jnetherton <jnetherton!> has quit IRC (Quit: Leaving)16:35
*** jcrossle_ <jcrossle_!> has joined #wildfly-swarm17:00
bbrowningbobmcw: so classpathresourcemanager is downright dangerous, at least as we're using it now17:02
bbrowningbut I'm not sure of any better solution for serving resources from inside a jar file17:02
*** jcrossley3 <jcrossley3!> has quit IRC (Ping timeout: 252 seconds)17:04
*** jcrossle_ is now known as jcrossley317:21
*** sblanc is now known as sblanc|out17:31
*** hbraun_away <hbraun_away!~hbraun@pdpc/supporter/professional/hbraun> has quit IRC (Quit: afk)18:04
sblanc|outbobmcw: ;) have a nice weekend18:07
jbossbotgit pull req [wildfly-swarm-addon] (open) sebastienblanc add the add fraction(s) command
*** hbraun <hbraun!~hbraun@pdpc/supporter/professional/hbraun> has joined #wildfly-swarm18:10
hbraunbobmcw congrats btw18:10
hbraunbbrowning maybe should with stuart in the wildfly hipchat room. there should be an easy easy way to filter what can be accessed18:11
hbrauns/should/we should talk18:12
bbrowninghbraun: yeah it's fairly easy to filter things18:12
bbrowningbut I'm not sure our general approach is right to begin with for static stuff in jars18:12
bbrowningif this was a war, all the static content would be in src/main/webapp instead of src/main/resources?18:13
hbraunyou mean static stuff in jars, or the way we serve it?18:13
bbrowninghbraun: the way we figure out what to treat as static content and what to not when run from a jar18:13
hbraunah, i see18:13
hbraunbbrowning do we retain the source directory structures as they are in maven layouts?18:15
hbraunprobably not18:15
hbraunit get's flattened into a package structure18:16
bbrowningthe booker example app is my only concrete example I'm really looking at, and static content like src/main/resources/index.html gets put in the jar as index.html at the top level18:16
bbrowningbasically how you'd expect18:16
hbraunyep, the regular maven approach18:17
bbrowningI think I'd like to propose that src/main/resources not be used for static content but instead src/main/webapp like you'd do for a jar18:17
bbrowningerr for a war18:17
bbrowningso even when not creating a war the swarm maven plugin would take stuff from src/main/webapp and put it somewhere known inside the jar so we make sure we only serve static content from there18:17
bbrowninginstead of potentially opening up the entire jar contents to be served accidentally18:17
hbraunsounds reasonable18:19
bbrowninglanceball: you did the ribbon-webapp stuff - does ^^^ sound reasonable to you?18:19
bbrowningor do we think the ability to serve static files from src/main/resources is important?18:20
lanceballI think webapp is totes reasonable18:20
kenfinniganbbrowning: sounds reasonable18:20
*** hbraun <hbraun!~hbraun@pdpc/supporter/professional/hbraun> has quit IRC (Quit: afk)18:55
*** lanceball is now known as lance|afk18:56
*** hbraun <hbraun!~hbraun@pdpc/supporter/professional/hbraun> has joined #wildfly-swarm19:11
hbraunkenfinnigan with regard to
jbossbotgit issue [wildfly-swarm] (open) hwellmann Wrapping standalone.xml
kenfinniganhbraun: yeah19:18
hbraunit's interesting to realize how different our perspectives are19:18
hbraunhowever I see your point19:19
hbrauni never thought about bypassing fractions completly19:19
kenfinniganhbraun: what is being proposed is quite likely needed to go further, but don't want us getting confused that this first step may actually be useful to a developer19:20
hbraunmaybe that's because I don't believe we'd be able to fully derive the fu9nctional runtime scope from the xml19:20
kenfinniganI could be wrong, but that was my impression of the original request, and certainly something that came out of last week19:20
hbraunah, interesting19:20
kenfinniganbeing able to take an existing standalone.xml, and app, and have Swarm "just work"19:20
kenfinniganhbraun: Mark sees it as an important migration path for existing WF/EAP users19:21
hbraunlet's assume we have support for basically anything that's in stock wildfly19:21
kenfinniganbeing able to take what they have, without necessarily grokking Swarm, and have it work19:21
hbraunand use pass it a regular standalone.xml, what would be the difference to regular wildfly ?19:22
kenfinniganhbraun: possibly nothing, other than an uber jar19:22
kenfinniganbut there are people that "slim" WF/EAP via standalone.xml to get only what they need19:22
kenfinniganin that case, it would certainly be different19:23
*** os890_ <os890_!> has joined #wildfly-swarm19:23
hbraunok, so if the point in swarm to deconstruct wildfly, to "just enough app server", how would you to that from xml?19:23
kenfinniganhbraun: that's certainly one of the benefits to Swarm, but not the only one19:24
kenfinniganthe use cases with standalone.xml may not be part of a "just enough app server" use case19:24
*** os890 <os890!> has quit IRC (Ping timeout: 250 seconds)19:24
kenfinniganbut more a, "how do I run on Swarm" use case19:24
*** os890_ is now known as os89019:25
hbrauni mean i know how you do it from xml, but i think very few people significantly "slim' down WF19:25
bbrowningor maybe someone just prefers their config as xml vs java code19:25
kenfinniganbbrowning: there's always that possibility19:25
bbrowningespecially if that xml config is outside of the jar potentially19:26
hbraunbbrowning yes, i thikn we've agreed on that one19:26
bbrowningie use 1 jar but change some config stuff per environment19:26
kenfinniganbbrowning, hbraun: That's certainly a valid situation, dev's write the app but Ops "owns" the standalone.xml19:26
hbraunyeah i agree19:26
hbraunfor config19:26
hbraunthe part i am not sure about is what the appropriate vehicle to slim down the runtime would be19:27
hbraunfractions seems to serve that purpose well19:27
hbraunnot sure if server.xml does it too19:27
bbrowningyeah in a standalone.xml the only thing you can really do is remove extension and subsystem elements19:28
bbrowningwe'd have to map those back to fractions to figure out which can be left out19:28
hbraunand hope it doesn't break19:28
hbraunthat's a major downside in WF for years19:28
kenfinniganhbraun, bbrowning: you can certainly slim standalone.xml, but you need to know what you're doing or you get in a mess19:28
bbrowningbut say standalone.xml vs standalone-full-ha.xml19:28
kenfinniganI would see this not so much as a "slimming" scenario and more a "support migration"19:29
hbraunthe relation between parts in not described19:29
bbrowningwe can certainly provide a slimmer jar for standalone.xml than if someone supplies a standalone-full-ha.xml19:29
kenfinnigantrue, there's lots in WF for HS only that we could ignore for regular standalone19:29
hbraunkenfinnigan the one thing I take out of this discussion for now, is that it's more complex then I thought19:30
kenfinniganhbraun: sorry19:30
hbraunno, that's a good thing19:30
*** sblanc|out <sblanc|out!~sebastien@redhat/jboss/sblanc> has quit IRC (Ping timeout: 240 seconds)19:31
hbraunit seems the current vehicles (fractions, maven deps, etc) are intricately related to the use cases19:32
hbraunkenfinnigan it's getting late over here, but i'll get back to you next week to better understand your point of view19:33
kenfinniganhbraun: no problem19:34
hbraunbye, have a nice weekend19:34
*** hbraun <hbraun!~hbraun@pdpc/supporter/professional/hbraun> has quit IRC (Quit: afk)19:35
*** os890 <os890!> has quit IRC (Quit: ChatZilla 0.9.92 [Firefox 42.0/20151029151421])19:41
*** enebo <enebo!> has quit IRC (Quit: enebo)20:11
*** lance|afk is now known as lanceball20:20
*** jcrossley3 is now known as jcrossley3-away20:28
*** rbenevides is now known as rbenevides|afk20:32
bbrowninghmm I have a hard time believing I took out almost all the static asset handling code and tests and examples passed20:43
bbrowningI guess that just means we don't really test static assets much20:44
kenfinniganbbrowning: correctamundo20:44
kenfinniganimproving tests is on my list20:44
kenfinniganbut might be a while20:44
bbrowningsays pretty much every project everywhere :)20:44
tcrawleyI just pulled and build HEAD, and now I'm getting a duplicate service error when trying to start a swarmtool'ed war that uses messaging21:02
tcrawleymessaging-activemq is the duplicate service21:02
tcrawleyany idea what might be causing that, or how to debug?21:02
*** sblanc <sblanc!~sebastien@redhat/jboss/sblanc> has joined #wildfly-swarm21:02
tcrawleyit should just be using the default messaging fraction21:02
tcrawleydebug logging doesn't give any more info21:03
lanceballok - so site is functional currently
jbossbotTitle: Some Title | WildFly Swarm21:19
bbrowningtcrawley: looks like a simple bug21:19
bbrowningbasically .defaultServer() is getting called twice21:20
bbrowninghow do tests not pick this up and fail? no clue21:20
tcrawleyah, interesting21:20
bbrowningshould .defaultServer() be idempotent? I'd say yes21:21
tcrawleyI agree. I'll take a whack at it21:21
tcrawleyand see if I can get a test to fail first21:21
kenfinniganbbrowning: the test issue could be because i've found that a test can actually fail but JUnit reports success because of the arq container starting a new process for the uber jar that is built21:23
bbrowningkenfinnigan: yeah I've seen a couple incidents of stuff like that21:23
kenfinnigani've seen errors in the other process, but the junit process exits fine so it says test passes21:23
bbrowningwe're definitely not failing tests when things fail all the time21:23
kenfinniganfor sure, testing stuff will probably be top of my todo when i move over officially21:24
*** jcrossley3-away is now known as jcrossley321:28
*** kenfinnigan <kenfinnigan!~ken@redhat/jboss/kenfinnigan> has quit IRC (Quit: Leaving.)21:29
*** bbrowning is now known as bbrowning_away21:42
*** lanceball is now known as lance|afk21:47
*** tcrawley is now known as tcrawley-away22:02
*** bobmcw <bobmcw!~bobmcw@redhat/jboss/bobmcw> has quit IRC (Remote host closed the connection)22:46
*** rbenevides|afk is now known as rbenevides23:03
*** rbenevides <rbenevides!~rbenevid@redhat/jboss/RafaelBenevides> has quit IRC (Quit: Quit)23:46
*** jcrossley3 is now known as jcrossley3-away23:51

Generated by 2.9.2 by Marius Gedminas - find it at!