#aerogear: Aerogear.js Gang Meeting

Meeting started by lfryc at 12:43:50 UTC (full logs). 

Meeting summary

  1. Integration Tests (lfryc, 12:44:17)
    1. integration tests are unstable (see travis builds) (lfryc, 12:44:29)
    2. they also use several different approaches for managing and executing external dependencies (lfryc, 12:44:29)
    3. test code reusal (e.g. DataManager duplicates lot of code) (lfryc, 12:44:29)
    4. https://issues.jboss.org/browse/AGJS-211 (lfryc, 12:44:48)
    5. consider cross-browser testing (with Karma), Jasmine for BDD (lfryc, 12:44:59)
    6. AGREED: lfryc: yea, what i was thinking (lholmquist, 12:47:32)
    7. AGREED: (dbevenius, 12:47:41)
    8. ACTION: lfryc write-up on ML about pros/cons of QUnit and Jasmine (lfryc, 12:50:06)
    9. AGREED: (lholmquist, 12:51:05)
    10. AGREED: (dbevenius, 12:51:09)

  2. DataSync/Conflict Resolution (lfryc, 12:51:58)
    1. Dan works on creating JIRAs for Conflict Resolution (lfryc, 12:52:12)
    2. https://github.com/aerogear/aerogear.org/pull/393 (lfryc, 12:52:16)
    3. ACTION: lfryc lholmquist please review the PR (dbevenius, 12:52:57)
    4. continued work on Real Time DataSync in parallel (lfryc, 12:53:18)
    5. AGREED: dbevenius (lholmquist, 12:55:12)
    6. ACTION: lholmquist create JIRAs for next steps for the realtime sync js-client (lfryc, 12:56:03)
    7. ACTION: dbevenius create JIRAs for extracting and moving the DiffSync JS client/demo to aergear-js. (dbevenius, 12:56:04)
    8. I think it would be nice if it was separate. (dbevenius, 12:57:26)
    9. https://github.com/lholmquist/ag-js-ds-poc (lholmquist, 12:58:52)
    10. AGREED: (lholmquist, 12:59:53)
    11. AGREED: (dbevenius, 13:01:30)
    12. AGREED: (lfryc, 13:01:33)
    13. AGREED: (lholmquist, 13:01:38)
    14. ACTION: lfryc review Luke's syncer (conflict resolution) https://github.com/lholmquist/ag-js-ds-poc/blob/master/app.js (lfryc, 13:02:04)
    15. API: consider Data Binding solution - object that is watched for changes from outside and managed by API calls (save(), sync(), next()) - Object.observe, Proxy (lfryc, 13:02:45)
    16. AGREED: (dbevenius, 13:04:50)
    17. AGREED: (lholmquist, 13:04:53)
    18. AGREED: (lfryc, 13:04:55)

  3. ES6 modules (lfryc, 13:06:04)
    1. i think this can be done for 1.X also. Transpiling can give us AMD, Common.JS and Global versions (or use UMD) (lfryc, 13:06:12)
    2. http://esnext.github.io/es6-module-transpiler/ (lfryc, 13:06:22)
    3. Ember uses this one and is pretty nice (lfryc, 13:06:26)
    4. https://github.com/umdjs/umd (lfryc, 13:06:54)
    5. AGREED: (lholmquist, 13:09:08)
    6. ACTION: lholmquist check whether ES6 can be transpiled to AMD, CommonJS and offer global access at same time (lfryc, 13:11:43)
    7. Lukas was considering using Traceur for ES6 transpilation (lfryc, 13:12:32)
    8. https://github.com/google/traceur-compiler/wiki/LanguageFeatures#modules (lfryc, 13:12:41)
    9. Luke pointed out: traceurs modules still need the traceur loader api, so not sure i like this one (lfryc, 13:12:50)
    10. https://github.com/lfryc/traceur-playground (lfryc, 13:13:55)
    11. should we be creating separate repo's like the other libraries, are the package managers mature enough for this. (lfryc, 13:14:41)
    12. Lukas does not agree that more separate Git repositories are a way to go unless there is specific technical need (lfryc, 13:14:55)
    13. AGREED: , this is not a good idea (lholmquist, 13:15:05)
    14. AGREED: (dbevenius, 13:15:16)
    15. AGREED: (lfryc, 13:15:37)

  4. ES6 features (lfryc, 13:15:56)
    1. Promises - probably pretty soon we can remove the polyfill from being packaged with lib and just link to it (lfryc, 13:16:02)
    2. take a look at some of the other features that are available or easily shimmed (lfryc, 13:16:10)
    3. https://github.com/paulmillr/es6-shim/ (lfryc, 13:16:15)
    4. http://kangax.github.io/compat-table/es6/ (lfryc, 13:16:20)
    5. AGREED: that sounds like a plan (lholmquist, 13:17:23)
    6. AGREED: (dbevenius, 13:17:36)
    7. ACTION: lfryc create a wiki page listing ES6 features we will aim in next development (lfryc, 13:17:44)
    8. idea: we can explore and use ES6 for new features as needed, use transpilation to ES5 and revisit their use/shimmability/performance as we reach beta/candidates for 2.0 (lfryc, 13:18:00)
    9. AGREED: (lholmquist, 13:18:41)
    10. AGREED: (dbevenius, 13:18:46)
    11. concern: can we keep transpiled code as accessible to final user on ES5 as it will be for ES6 user? (lfryc, 13:19:27)

  5. custom builder (lfryc, 13:22:56)
    1. gulp might be a better fit for this on the server side (lfryc, 13:23:07)
    2. gulp might be a better fit for this on the server side (lholmquist, 13:23:35)
    3. this depends on the modules topic, but probably should either have an option for what you want, AMD, commonjs , Global or just package all three together (lholmquist, 13:23:35)

  6. possibility of removing Authentication Module from lib (lfryc, 13:26:25)
    1. http://aerogear-dev.1069024.n5.nabble.com/aerogear-dev-android-Removing-AGAuthenticationModule-tp9209.html (lfryc, 13:26:32)
    2. Auth doesn't do anyting "security wise", the browser handles all that anyway (lfryc, 13:26:39)
    3. still may be beneficial for automatic data sync, server key management (lfryc, 13:26:44)
    4. there is an idea to implement Keycloak and Mozilla's Personas adapters, do we want to keep this module around though? (lfryc, 13:26:53)
    5. ACTION: lfryc start a ML discussion about removing or keeping Auth with removed REST + mention Keycloak and Mozilla Personas (lfryc, 13:29:41)

  7. deprecating Pipelines (lfryc, 13:29:55)
    1. to be removed in 2.0 (lfryc, 13:30:01)
    2. need strategy how to fluently integrate with third-party libraries (jQuery.ajax, AngularJS $resource / Restangular, Ember Data?) (lfryc, 13:30:08)
    3. AGREED: (dbevenius, 13:30:34)
    4. AGREED: (lholmquist, 13:32:39)
    5. AGREED: (dbevenius, 13:32:54)
    6. AGREED: keep and support Pipelines in 1.x with no further development, deprecate in 2.0 (lfryc, 13:33:18)

  8. Bower for dependency management (lfryc, 13:33:30)
    1. https://github.com/aerogear/aerogear-js ? (lfryc, 13:35:03)
    2. https://github.com/lholmquist/aerogear-js-dist (lholmquist, 13:35:31)
    3. ACTION: lholmquist look to see what external deps for aerogear.js are now in bower (lholmquist, 13:37:31)
    4. ACTION: lholmquist think about using bower in cookbook examples (lholmquist, 13:39:59)

  9. Integration with third-party libraries (lfryc, 13:40:25)
    1. AngularJS, Ember (lfryc, 13:40:32)
    2. Polymer?, jQuery Mobile? (lfryc, 13:40:32)
    3. what about others? (lfryc, 13:40:32)

  10. Priorities for 2.0 (lfryc, 13:45:09)
    1. ES6 support: modularity, promises, ... (lfryc, 13:45:19)
    2. focus on integration with 3rd party frameworks (lfryc, 13:45:22)
    3. refactor build and tests (lfryc, 13:45:22)
    4. AGREED: (lholmquist, 13:46:37)
    5. AGREED: (dbevenius, 13:46:38)
    6. AGREED: ES6 support, data sync, refactor build and tests (lfryc, 13:47:01)
    7. AGREED: es6 modules should be sooner than later (lfryc, 13:47:35)

  11. When to start 2.0 (lfryc, 13:47:42)
    1. based on discussed usage of ES6 transpilation, I foresee a lot of breaking changes that will disallow simple merges from one branch to another (lfryc, 13:48:06)
    2. at some point we have to start limit 1.x changes and focus on 2.0 (lfryc, 13:48:06)
    3. ACTION: lfryc start a ML discussion about when and where to include Data Sync module (lfryc, 13:49:44)

  12. use of Agile / Scrum Boards in JIRA (lfryc, 13:50:28)
    1. Agile boads are awesome way to manage development, load and track progress from version to version (lfryc, 13:50:34)
    2. https://www.atlassian.com/software/jira/agile (lfryc, 13:50:39)
    3. AGREED: we will try JIRA Agile / Scrum board (lfryc, 13:52:55)

  13. Documentation & Cookbooks (lfryc, 13:53:48)
    1. iOS team did very good job in bringing cookbooks to Aerogear.org site (lfryc, 13:54:08)
    2. AGREED: (lholmquist, 13:54:35)
    3. AGREED: (dbevenius, 13:54:40)


Meeting ended at 13:56:25 UTC (full logs). 

Action items

  1. lfryc write-up on ML about pros/cons of QUnit and Jasmine
  2. lfryc lholmquist please review the PR
  3. lholmquist create JIRAs for next steps for the realtime sync js-client
  4. dbevenius create JIRAs for extracting and moving the DiffSync JS client/demo to aergear-js.
  5. lfryc review Luke's syncer (conflict resolution) https://github.com/lholmquist/ag-js-ds-poc/blob/master/app.js
  6. lholmquist check whether ES6 can be transpiled to AMD, CommonJS and offer global access at same time
  7. lfryc create a wiki page listing ES6 features we will aim in next development
  8. lfryc start a ML discussion about removing or keeping Auth with removed REST + mention Keycloak and Mozilla Personas
  9. lholmquist look to see what external deps for aerogear.js are now in bower
  10. lholmquist think about using bower in cookbook examples
  11. lfryc start a ML discussion about when and where to include Data Sync module


Action items, by person

  1. dbevenius
    1. dbevenius create JIRAs for extracting and moving the DiffSync JS client/demo to aergear-js.
  2. lfryc
    1. lfryc write-up on ML about pros/cons of QUnit and Jasmine
    2. lfryc lholmquist please review the PR
    3. lfryc review Luke's syncer (conflict resolution) https://github.com/lholmquist/ag-js-ds-poc/blob/master/app.js
    4. lfryc create a wiki page listing ES6 features we will aim in next development
    5. lfryc start a ML discussion about removing or keeping Auth with removed REST + mention Keycloak and Mozilla Personas
    6. lfryc start a ML discussion about when and where to include Data Sync module
  3. lholmquist
    1. lfryc lholmquist please review the PR
    2. lholmquist create JIRAs for next steps for the realtime sync js-client
    3. lfryc review Luke's syncer (conflict resolution) https://github.com/lholmquist/ag-js-ds-poc/blob/master/app.js
    4. lholmquist check whether ES6 can be transpiled to AMD, CommonJS and offer global access at same time
    5. lholmquist look to see what external deps for aerogear.js are now in bower
    6. lholmquist think about using bower in cookbook examples


People present (lines said)

  1. lfryc (277)
  2. lholmquist (103)
  3. dbevenius (36)
  4. jbossbot (8)
  5. jbott (6)
  6. aerobot (5)
  7. corinnekrych (2)


Generated by MeetBot 0.1.4.