Differences

This shows you the differences between two versions of the page.

Link to this comparison view

irc:1441058400 [2017/05/27 13:44] (current)
Line 1: Line 1:
 +[00:04:37] <Sticky> make sure you call the async thing, otherwise the test will assume it blocks until completion
 +
 +[05:10:18] *** ChanServ sets mode: +o temporalfox
 +
 +[08:20:35] *** ChanServ sets mode: +o temporalfox
 +
 +[10:10:33] *** ChanServ sets mode: +o purplefox
 +
 +[10:15:03] <purplefox> pmlopes: cescoffier temporalfox: morning
 +
 +[10:15:13] <pmlopes> good morning
 +
 +[10:15:13] <temporalfox> hi
 +
 +[10:17:15] <cescoffier> good morning
 +
 +[10:17:38] <cescoffier> so, we have 3 talks accepted at devoxx, and one at softshake (geneva)
 +
 +[10:18:56] <Sticky> gratz, are the topics all vertx?
 +
 +[10:19:50] <cescoffier> Sticky: yes
 +
 +[10:19:51] <purplefox> cescoffier: awesome!
 +
 +[10:20:59] <purplefox> i don't know softshake
 +
 +[10:23:08] <cescoffier> purplefox: it's a small conference in Geneva (so very close), mostly financial / banking app developers (well, Geneva ;-))
 +
 +[10:24:42] <purplefox> cescoffier: so we have 1 lab + 1 tutorial + 1 conference talk now (at devoxx) ?
 +
 +[10:25:15] <cescoffier> purplefox: yes
 +
 +[10:25:22] <cescoffier> purplefox: no :-)
 +
 +[10:25:24] <purplefox> and you and paulo doing them all as joint presenters?
 +
 +[10:25:43] <cescoffier> purplefox: a lab (micro-services), a tools in action (vertx-web), a conference
 +
 +[10:26:03] <purplefox> is the lab the one where you use docker and openshift?
 +
 +[10:26:09] <cescoffier> purplefox, and yes, all of them are presented by pmlopes and me
 +
 +[10:26:17] <purplefox> cool
 +
 +[10:26:32] <purplefox> I am actually quite glad that I am having a break from conferences this year :)
 +
 +[10:26:33] <cescoffier> purplefox: yes, openshift, docker, hazelcast, and a web frontend
 +
 +[10:26:41] <purplefox> maybe next year i will do some again
 +
 +[10:27:33] <cescoffier> purplefox: the talks would be ready ;-)
 +
 +[10:27:48] <purplefox> he, that's a bonus :) so when are you writing them?
 +
 +[10:28:29] <cescoffier> starting next week (you know during our vacations)
 +
 +[10:28:41] <purplefox> I do have to give a private talk in November would be nice to reuse the conference one for that
 +
 +[10:28:51] <cescoffier> it will be ready
 +
 +[10:28:59] <purplefox> cool
 +
 +[10:29:07] <purplefox> are you going to user reveal.js for it?
 +
 +[10:29:13] <purplefox> s/user/use
 +
 +[10:30:22] <cescoffier> purplefox: we have to decide with pmlopes. It might be cool to have vert.x serve the slides ;-)
 +
 +[10:30:33] <purplefox> the thing with conferences is we could pretty much do all the conferences if we wanted to - I still get invited to a lot without submitting, and if we submitted we'd pretty much be guaranteed a talk as Vert.x is an interesting topic
 +
 +[10:30:46] <purplefox> so the question is really how much time we want to devote to them
 +
 +[10:31:14] <purplefox> i'm in two minds whether conferences are an effective form of marketing or not
 +
 +[10:31:32] <cescoffier> these conferences are more to advertise vert.x 3.x
 +
 +[10:32:18] <pmlopes> purplefox which ones do you consider more important to present?
 +
 +[10:33:03] <cescoffier> I would say the conference
 +
 +[10:33:20] <purplefox> the conference talk will get more audience
 +
 +[10:33:46] <purplefox> i've talked at devoxx about vert.x twice
 +
 +[10:33:57] <purplefox> both times room was full, around 500 people
 +
 +[10:34:21] <purplefox> devoxx is a good conf
 +
 +[10:34:49] <purplefox> but the thing with most conferences is you usually see the same old faces at every conference
 +
 +[10:35:01] <purplefox> who probably already know about vert.x anyway ;)
 +
 +[10:35:31] <purplefox> and 95% of devs never go to conferences because they are expensive
 +
 +[10:36:01] <purplefox> so confs only reach that 5% of devs who are probably alpha developers anyway
 +
 +[10:36:08] <purplefox> anyway just my 2c
 +
 +[10:36:35] <purplefox> they can be a fun holiday though and a chance to see a new place
 +
 +[10:38:40] <pmlopes> then the alternative is to start a PR team :) like other projects make a weekly mailing with articles, youtube videos, etc... that developers can subscribe so you can easily target the real users and not the alpha developers or managers that attend conferences
 +
 +[10:41:07] <pmlopes> for example once a week i get some email from docker, postgres and html5 with a couple of links of what is going on so i can keep update, but that is kind of a job on itself, to create content, curate it, and keep the flow so every week there is something interesting
 +
 +[10:55:54] <millrossjez> @purplefox, conferences are interesting, I go to conferences to learn new stuff so I might go see a vert.x talk if it was beyond a basic intro, but would swerve a complete introduction, but like you say confs cost money and most employers won't send their staff to them (and often the staff can't afford to send themselves)
 +
 +[10:56:20] <millrossjez> the main people you'll hit at conferences are people who are already interesting, and the self-employed
 +
 +[10:56:28] <millrossjez> the main people you'll hit at conferences are people who are already interested, and the self-employed
 +
 +[10:56:51] <millrossjez> that said, Josh Long always fills the room for his Spring Boot talks (which tends to be "what's new since last time I told you about it")
 +
 +[11:04:38] <purplefox> yeah, Josh is a great speaker
 +
 +[11:05:00] <purplefox> spring boot has a much bigger budget than us for marketing though
 +
 +[11:05:10] <purplefox> i mean, josh is pretty much exclusively on this these days
 +
 +[11:05:39] <purplefox> perhaps I am wrong, and conferences do help spread the word
 +
 +[11:05:58] <purplefox> i think vert.x could certainly do with more marketing in some way or other
 +
 +[11:06:26] <purplefox> trouble with the red hat marketing folks is there aren't many of them and they are spread thinly over many projects
 +
 +[11:06:55] <purplefox> i think pivotal has an advantage here as they have fewer projects so evangelist to project ratio is higher
 +
 +[11:35:20] <millrossjez> yeah
 +
 +[14:38:01] <vertx_user> I have two verticles, which need to communicate through eventbus. Right now, i am running both of them on my Mac, with "-cluster" , But they are not able to communicate with one another. How to proceed from here?
 +
 +[14:38:27] <vertx_user> any debugging tips, or suggestions?
 +
 +[14:47:29] <cescoffier> vertx_user: did you check if multicast is enabled ? It's disabled by default on mac
 +
 +[14:47:56] <cescoffier> vertx_user: check http://vertx.io/docs/vertx-hazelcast/#_trouble_shooting_clustering
 +
 +[14:50:53] <vertx_user> cescoffier: is there a way to check if multicast is enabled?
 +
 +[14:51:12] <vertx_user> I tried to ping at 224.0.0.1, which i thought was multicast address
 +
 +[14:51:39] <vertx_user> and i recieved reply from many different ips.
 +
 +[14:51:59] <vertx_user> Most of the systems on my lan are Mac.
 +
 +[14:52:12] <vertx_user> does it mean its working?
 +
 +[14:54:25] <cescoffier> you should try to enable the logging of hazelcast
 +
 +[14:54:34] <cescoffier> and see if the cluster is formed correctly
 +
 +[14:54:48] <cescoffier> maybe you have several network interface and it does not pick the right one
 +
 +[14:56:04] <vertx_user> i tried with cluster-host localhost
 +
 +[14:57:11] <vertx_user> that should ensure right interface i guess
 +
 +[15:00:44] <vertx_user> I just tried adding com.hazelcast.level=INFO to vertx-default-jul-logging.properties
 +
 +[15:01:08] <vertx_user> no difference
 +
 +[15:02:07] <vertx_user> after "Starting clustering..." , i get failure on eventbus.send
 +
 +[15:02:14] <vertx_user> nothing else is being logged
 +
 +[15:02:28] <vertx_user> does that mean cluster is not formed?
 +
 +[15:02:32] <vertx_user> what can i do about it?
 +
 +[15:02:55] <vertx_user> cescoffier: ^
 +
 +[15:04:52] <cescoffier> your cluster host is probably something like 192.168.1.x
 +
 +[15:05:33] <cescoffier> if enabled correctly, hazelcast is really verbose
 +
 +[15:19:50] <vertx_user> am enabling the logging correctly?
 +
 +[15:20:18] <vertx_user> I added com.hazelcast.level=INFO to a file named vertx-default-jul-logging.properties
 +
 +[15:20:26] <vertx_user> is that all?
 +
 +[15:20:51] *** ChanServ sets mode: +o temporalfox
 +
 +[15:20:55] <vertx_user> cause i don't see any detailed logging by hazelcast
 +
 +[15:43:08] <temporalfox> vertx_user hazelcast logging needs to be configured differently
 +
 +[15:46:27] <cescoffier> vertx_user: how are you launching your application, using the 'vertx' command line or in a fat jar
 +
 +[15:47:04] <cescoffier> in the vert.x command line, the easiest is to edit the logging properties that are in vertx_home/conf
 +
 +[15:47:56] <cescoffier> in a fat jar, create a logging.properties, copy the default logging file content, and make the modification you made, then add this file at the root of the fat jar
 +
 +[16:22:45] <melvinross> in a clustered event bus, does vertx deliver messages to vertices inside of the same vertx instance differently than it delivers messages to vertices inside of other vertx instances?
 +
 +[16:47:46] <Sticky> at the moment there is support for hamcrest matchers in any of vertx 3's testing is there?
 +
 +[16:48:11] <cescoffier> Sticky: nope
 +
 +[16:48:23] <cescoffier> (if you think to vertx-unit)
 +
 +[16:49:10] <cescoffier> you need to 'customize' your hamcrest usage to call context.fail (context it the test context)
 +
 +[16:49:30] <cescoffier> however, if you use plain junit, then no problem to use hamcrest, assertj...
 +
 +[16:49:41] <cescoffier> I genenerally use Awaitability in compbination to assertj
 +
 +[16:50:26] <Sticky> yeah, I am using the async stuff so cant use hamcrest out the box
 +
 +[17:33:24] <AlexLehm> Sticky, i think i have used hamcrest in some tests, let me check
 +
 +[18:03:29] <AlexLehm> ok, i wrote a assertThan method that calls context.fail
 +
 +[18:03:35] <AlexLehm> assertThat
 +
 +[18:03:59] <Sticky> yeah, I wrote the same method
 +
 +[22:28:38] *** ChanServ sets mode: +o purplefox
 +
 +[23:02:22] <jtruelove> are any of you guys going to JavaOne?