Code And Learn at NodeFest 2016

[yosuke-furukawa]

Hi folks.
NodeFest is the largest Node.js conference at Japan.

I am an organizer of the conference, I would like to open code-and-learn-jp on NodeFest.
Japan has 1 collaborator (me) and 1 core team member (shigeki) and some great nodeschool staffs (martin, ledsun, watilde, tako-black, etc). We would like to support contribution to nodejs/core.

And I would like to list up the contribution point of Japanese node beginners.
I guess we will receive some PRs on 12th November.
We will follow English communication and some other commit techniques.

[yosuke-furukawa]

Contributing Point:

doc typo:

• doc/modules.md: line 132 lookups, it would be better looks up.
• doc/cluster.md: line 147, 499 has eg., it would be better e.g..
• doc/repl: line 6 has includable, it would be better includible. http://wikidiff.com/includable/includible
• doc/dns.md: line 263 e.g., => e.g.
• doc/http.md: line 553/856 e.g., => e.g.
• doc/tls.md: line 762 836 1026 e.g., => e.g.

doc improvement:

• doc/url.md does not have WHATWG-URL standard link and the description.

[Trott]

If you need more issues, one thing you can do is search through the test files for assert.equal(). In almost all cases, the test could be improved by using assert.strictEqual() instead. The one exception I can think of might be test-assert.js where use of assert.equa() may be necessary.

[yosuke-furukawa]

Contributing Point:

test:

• test/parallel/test-http-status-reason-invalid-chars.js: use port 0 instead of common.PORT,

const server = http.createServer((req, res) => {
  if (req.url === '/explicit') {
    explicit(req, res);
  } else {
    implicit(req, res);
  }
}).listen(common.PORT, common.mustCall(() => { // should use 0
  const url = `http://localhost:${common.PORT}`; // should use server.address().port
  let left = 2;
  const check = common.mustCall((res) => {
    left--;
    assert.notEqual(res.headers['content-type'], 'text/html');
    assert.notEqual(res.headers['content-type'], 'gotcha');
    if (left === 0) server.close();
  }, 2);
  http.get(`${url}/explicit`, check).end();
  http.get(`${url}/implicit`, check).end();
}));

[yosuke-furukawa]

@Trott thank you!!! sound so good. I am currently searching contribution point on code-and-learn on nodefest.

[Trott]

Here's another one: There are a number of instances of setTimeout() where the function is called without a duration (the second argument). Node.js will treat that the same as calling it with a duration of 1. In many of these cases, it might be better to use setImmediate().

Here are some of those cases. I have not reviewed them to verify that setImmediate() is better, but that might be an appropriate exercise for a first time committer.

• test/sequential/test-next-tick-error-spin.js line 43
• test/parallel/test-tls-hello-parser-failure.js line 30
• test/parallel/test-stream2-writable.js line 331

[yosuke-furukawa]

test: use port 0 instead of common.PORT

• test/parallel/test-cluster-net-send.js

  server.listen(common.PORT, function() { // use 0 port
    socket = net.connect(common.PORT, '127.0.0.1', socketConnected); // use server.address().port
  });

• test/parallel/test-tls-connect-address-family.js
• test/parallel/test-net-socket-destroy-twice.js

[Trott]

Japan has 1 collaborator (me) and 1 core team member (shigeki)

@ronkorving is a collaborator and lives in Tokyo.

[yosuke-furukawa]

@Trott ah, that's so good.
I will list these points.

[ronkorving]

@Trott @yosuke-furukawa Indeed :) I'm not too active in the Japanese Node community, but wouldn't mind trying harder :) My Japanese is not yet 100% what I would like it to be. In any case, I signed up for Sunday (visitor) and wouldn't mind helping with the organization next year either if that's appreciated. I'll see you on Sunday. 楽しみにしてます ^^

[jasnell]

@yosuke-furukawa I'm definitely looking forward to being there and will help out in whatever way that I can. I just want to confirm that the Code and Learn is scheduled for Saturday afternoon, correct? I want to take a little bit of time Saturday morning to sight see a little bit and I want to make sure I plan things so I won't be late to the Code and Learn :-)

[shigeki]

I'm going to join it to help participants and take care of what tests and docs can be good for the first contributions. But actually, the class is just one and half hour so I think it is too short for novices.

Instead of submitting a new PR to nodejs repo, I made an alternative repo of node to give such users the first experiences of contributions in https://github.com/shigeki/node_testpr/ . It can be a training for them if they submit a fake PR such as shigeki/node_testpr#1 .

Users can choice which repo they want to work on during the course.

[shigeki]

I've just submitted some issues of test failures to be reserved for this course in nodejs/node#9509, nodejs/node#9510 and nodejs/node#9511

[yosuke-furukawa]

@jasnell

I just want to confirm that the Code and Learn is scheduled for Saturday afternoon, correct?

Yes!!! Code And Learn will begin 15:45 on Saturday, and the end time is 17:15.

And I really would like you to attend NodeDiscussion on Saturday morning, almost 90 min (11:00-12:30). This discussion is unconference style. We would like to discuss about Node.js good point/bad point/wishlist. I would like share our opinions to node core committers.

If you don't have any time on Saturday morning, it is OK, but if you have time, please come!!!!!

[yosuke-furukawa]

@ronkorving Ya!!! please help us !!!! You can come our venue and speak to our receptionist as mentor. you can join :)

[yosuke-furukawa]

@shigeki awesome!!! I will wrap up these contribution point on this Thursday.

[Trott]

test/sequential/test-next-tick-error-spin.js line 43

You can remove that one from the list as I'm in the process of fixing it while refactoring a bunch of other things in the test. Sorry about that.

You can replace it, though with these tests which all have at least one instance of setTimeout() called without a second argument, and thus are candidates for possibly using setImmediate() instead:

• test/parallel/test-stream2-readable-wrap.js line 61
• test/parallel/test-stream2-readable-non-empty-end.js lines 19 and 34
• test/parallel/test-stream2-push.js line 116
• test/parallel/test-stream2-large-read-stall.js line 49
• test/parallel/test-stream-unshift-read-race.js line 44
• test/parallel/test-stream-unshift-empty-chunk.js line 16
• test/parallel/test-stream-transform-objectmode-falsey-value.js line 33

[ronkorving]

@yosuke-furukawa Sounds good, I'll try to make it! :) Basically just show up before 3:45?

[yosuke-furukawa]

@Trott thank you soooooooooooooooo much!!!!!!!!!!!!!!!!!!!!!!

[yosuke-furukawa]

@ronkorving Yes! we will start at 15:45 on dots Shibuya, tokyo. https://eventdots.jp/space

[Fishrock123]

https://coverage.nodejs.org/coverage-fb05e31466ac0bad/root/internal/process/stdio.js.html

1. A pseudo-tty test that manually fires 'SIGWINCH' and makes sure none of the properties set by _refreshSize have been changed.
2. Add the stderr.destroy/stderr.destroySoon error checking to the test which already checks that for stdout.

[yosuke-furukawa]

@Fishrock123
Thank you sooooooooo much !!!!! This coverage information is really helpful for us.

[yosuke-furukawa]

All contribution point:
docs:

• doc/modules.md: line 132 lookups, it would be better looks up.
• doc/cluster.md: line 147, 499 has eg., it would be better e.g..
• doc/repl: line 6 has includable, it would be better includible. http://wikidiff.com/includable/includible
• doc/dns.md: line 263 e.g., => e.g.
• doc/http.md: line 553/856 e.g., => e.g.
• doc/tls.md: line 762 836 1026 e.g., => e.g.

doc improvement:

• doc/url.md does not have WHATWG-URL standard link and the description.

test assertion improvement:

• assert.equal would be better to use assert.strictEqual

test failure with build option :

• test failures without openssl-cli node#9509
• test failures of v8 inspectors build without dependencies node#9510
• test-icu-transcode fails built without intl node#9511

test timing API improvement:
Use setImmediate instead of setTimeout().

• test/parallel/test-tls-hello-parser-failure.js line 30
• test/parallel/test-stream2-writable.js line 331
• test/parallel/test-stream2-readable-wrap.js line 61
• test/parallel/test-stream2-readable-non-empty-end.js lines 19 and 34
• test/parallel/test-stream2-push.js line 116
• test/parallel/test-stream2-large-read-stall.js line 49
• test/parallel/test-stream-unshift-read-race.js line 44
• test/parallel/test-stream-unshift-empty-chunk.js line 16
• test/parallel/test-stream-transform-objectmode-falsey-value.js line 33

test use port:0 instead of common.PORT :

• test/parallel/test-http-status-reason-invalid-chars.js: use port 0 instead of common.PORT,
• test/parallel/test-cluster-net-send.js
• test/parallel/test-tls-connect-address-family.js
• test/parallel/test-net-socket-destroy-twice.js

coverage improvement:

• A pseudo-tty test that manually fires 'SIGWINCH' and makes sure none of the properties set by _refreshSize have been changed.
• Add the stderr.destroy/stderr.destroySoon error checking to the test which already checks that for stdout.

https://coverage.nodejs.org/coverage-fb05e31466ac0bad/root/internal/process/stdio.js.html

[Trott]

assert.deepEqual would be better to use assert.deepStrictEqual

We have a lint rule for that so you shouldn't find any examples of deepEqual that haven't been explicitly excused as necessary exceptions (e.g., in test-assert.js).

[yosuke-furukawa]

@Trott OK, I updated.

[dorako321]

i will try to fix this doc/dns.md: line 263 e.g., => e.g.

[akito0107]

I will try to fix test/parallel/test-cluster-net-send.js !