This version (2017/05/27 13:44) is a draft.
Approvals: 0/1

[09:58:16] * ChanServ sets mode: +o purplefox [10:09:13] <purplefox> cescoffier: pmlopes folks, sorry I overslept so can we delay the meeting? [10:09:37] <cescoffier> when would you be available ? [10:11:05] <purplefox> how about in an hour or so? [10:11:27] <cescoffier> ok for me [10:11:41] <pmlopes> me too [10:12:03] <cescoffier> 11:15 ? [11:06:40] <purplefox> that sounds good [11:06:53] <purplefox> pmlopes: cescoffier: we can do it now if you like [11:07:04] <cescoffier> Ok for me [11:16:42] <Narigo> cescoffier, nice work @ ExecutionContext and Future wrapper for Scala to Java port :) Not sure about the executeBlocking/isOnWorkerThread though - the driver itself is async, so shouldn't “runnable.run()” suffice in all cases? [12:16:55] <Narigo> cescoffier, I think I'm through the mysql-postgresql review… thanks for porting it! [15:17:01] <cescoffier[Away]> Narigo: I've made some changes. [15:18:19] <Narigo> cescoffier[Away], will have a look. Does volatile suffice? Maybe a test for races would be good… I've never implemented anything like that as it started as a service, not a client… [15:18:46] <cescoffier[Away]> (wow, I really need to understand how my IRC client is working, I'm NOT AWAY !!!) [15:19:50] <cescoffier> I think it should be enough, as it will always retrieve the actual value [15:20:42] <cescoffier> hum not [15:20:49] <cescoffier> there are co-conditions [15:30:12] <cescoffier> so, inTransaction is volatile, but inAutocommit is not. It is always used in combination to inTransaction, so a synchronized block is required here. [15:40:00] <cescoffier> pmlopes: are you around ? [15:40:05] <pmlopes> yes [15:40:19] <cescoffier> I'm back on the node examples [15:40:32] <cescoffier> I'm executing npm install and npm start [15:40:44] <cescoffier> But I get this: /Users/clement/Projects/vert.x/vertx-examples/web-examples/src/main/java/io/vertx/example/web/vertxbus/node/node_modules/vertx3-eventbus-client/vertxbus.js:27 [15:40:44] <cescoffier> factory(SockJS); [15:42:02] <pmlopes> let me try to open it [15:48:37] <pmlopes> ah i know what's wrong the npm package does does contain the latest js from from vertx-web/client [15:49:39] <pmlopes> we need to upload a new package to npm with the latest js from our git [15:50:32] <cescoffier> oh, I see [15:50:50] <cescoffier> it should wait the release no ? [15:53:44] <pmlopes> well i think so, otherwise we have a 3.1 artifact on npm and 3.0 on maven… [15:57:43] <cescoffier> I let you delete the example (or move it to a branch we can re-apply when the 3.1 will be alive) [16:01:50] <pmlopes> they all need to move to a 3.1 branch, since amd and commonjs also get the js file from npm [16:05:23] <cescoffier> hum, there are working right now [16:26:00] <pmlopes> yes the other ones do work because the loader fix was node specific the rest remained untouched [19:44:35] * ChanServ sets mode: +o purplefox