nodejs/code-and-learn

Code and Learn at Node.js Interactive in Amsterdam

mikeal opened this issue · 70 comments

mikeal commented

So, I'd like to resurrect this event to run either before or after the Node.js Interactive events in Amsterdam and Austin.

We need to figure out what this will look like so that we can budget it correctly.

As a reminder, this program was intended as a series of workshops for getting people working on Node.js itself. The goal was that by the end of the workshop they would have a PR open, or be close to it.

We need to put together a list of subjects, people that can properly mentor the subject, and if those people need their travel covered we need to know where they are coming from. Also, for each subject and mentor allocation we need to figure out the maximum size of the attendees.

/cc @dshaw @ashleywilliams @hackygolucky @Fishrock123

MylesBorins commented

I'd be up for helping with this, obviously dependent on me attending the events

bnoordhuis commented

Amsterdam is my home turf (after a fashion) so I'll volunteer. I request remuneration only for the train ticket but I do tend to travel 1st class exclusively so there's that.

iancrowther commented

happy to chip in

mikeal commented

We really need someone to step up to lead this effort at Amsterdam and Austin (doesn't need to be the same person each time).

jasnell commented

I can help as a mentor on site. Going to be fairly busy I think so I don't want to overcommit as 'lead'

MylesBorins commented

@mikeal would this be running during the collaborator summit? Do we have resources to offer individuals, potentially with a focus on under represented groups, continued support after a code and learn?

I think it might be particularly interesting to pair current collaborators with new collaborators. It might also be great if we can offer some sort of scholarship to individuals in need so that they can focus a continued effort on getting involved.

  • 👍1
Fishrock123 commented

Which day is this at in Amsterdam?

I may be able to help but I definitely don't want to lead it.

bnoordhuis commented

I can lead but it depends on what day it is. (EDIT: In Amsterdam.)

mikeal commented

Morning of the 17th.

addaleax commented

I’ll definitely want to be there and can probably also invest some time into it beforehand.

bnoordhuis commented

How morning is morning? 17th works for me but I need to drive up from Groningen. I can be there around 11 am.

bnoordhuis commented

Thanks, that page was still blank last I checked. :-)

I don't think I'll be able to make 9 am, I'd have to leave before 6 am.

mikeal commented

Heya, do you need anything from the events team?

Fishrock123 commented

I won't be there fwiw.

bnoordhuis commented

So who is going to lead?

addaleax commented

If nobody else is available, I can try to, even as a newbie… I’d definitely want to talk with people who have done this before in that case

MylesBorins commented

im up for being involved, let me know what I can do to help

bnoordhuis commented

IRC or email? Don't worry, the biggest thing is to have some 'good first contribution' issues stacked up, everything else flows from there (after walking people through how to build first.)

addaleax commented

hm, how about in IRC in an hour (I’m busy until then), maybe with @thealphanerd too?

jasnell commented

I will be there also and can definitely help in any way that I can but I'm more than happy for @addaleax or others to take the lead if they'd like.

bnoordhuis commented

I have cheese/bacon pancakes and a Danish krimi to attend to first but I'll hop onto IRC afterwards. Make it 1.30h from now.

addaleax commented

oh, have fun! and yeah, works for me :)

MylesBorins commented

ping me when ya’ll are ready, I should be around

On Sep 8, 2016, at 1:12 PM, Anna Henningsen notifications@github.com wrote:

oh, have fun! and yeah, works for me :)

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub #56 (comment), or mute the thread https://github.com/notifications/unsubscribe-auth/AAecV7QpUzKw2TFFAtUhK-7ojlm2T4I6ks5qoEHogaJpZM4IKMKM.

  • 👍1
addaleax commented

oh, yeah, seems I forgot to comment… I should be able to lead this :)

addaleax commented

Okay, so from what I’ve gathered by talking to a few people here, the expectations for Code & Learn are a bit diverse, as are the skill levels of attending people. I’ve gathered a few changes that people might up to do, maybe that’s enough (I’ll definitely gather a bit more), but I’m pretty sure that even if not we’re going to be able to provide attendees with things they’re interested in (@saghul even offered to help people with first libuv contributions if they’re interested in that!).

A few things that I think are okay:

  • lib/internal/bootstrap_node.js
    • The NativeModule._cache can safely be turned into a Map!
    • var -> const substitutions
  • lib/v8.js, src/node_v8.cc:
    • There are a few fields missing on HeapStatistics which have been added in more recent V8 versions:
      • malloced_memory
      • peak_malloced_memory
      • does_zap_garbage
  • lib/util.js:
    • i): The SIMD setup can be turned into a loop and the typeof checks can be removed
    • ii): Support for SharedArrayBuffer can be added here, too.
    • iii): (Maybe) The catch-all try check in tryStringify() for circular objects is probably too unspecific
  • src/udp_wrap.cc, src/stream_base.cc:
    • The uv_buf_t bufs_[16] lines could be MaybeStackBuffers;
  • lib/internal/process.js
    • (Maybe) We might want to emit a warning for signals that users shouldn’t listen on (as listed in the docs warning in nodejs/node#8410)?
  • lib/internal/lazy_transform.js
    • The default encoding here should be configurable and default to crypto.DEFAULT_ENCODING, not latin1?. This seems to kind of have been forgotten in nodejs/node#5522 – It would be a semver-major change, but I think this is not going to be contentious.
  • edit: nodejs/node#3208

And a few test files that could be cleaned up in NodeTodo/@Trott style:

  • test-c-ares.js:
    • A lot of vars that can be const
    • assert.strictEqual should be used instead of assert.equal
    • Wrap all callbacks that should be called once in common.mustCall()
  • test-child-process-buffering.js:
    • A lot of vars that can be const
    • assert.strictEqual should be used instead of assert.equal
  • test-child-process-detached.js
    • A lot of vars that can be const
    • assert.notStrictEqual should be used instead of assert(… !== …)
  • test-child-process-stdio.js:
    • A lot of vars that can be const
    • assert.strictEqual should be used instead of assert.equal, the same goes for notEqual
  • test-child-process-stdout-flush.js:
    • A lot of vars that can be const
    • assert.strictEqual should be used instead of assert.equal, the same goes for notEqual
    • Wrap all callbacks that should be called once in common.mustCall()

Tomorrow we might get a better picure of how many people would be able and willing to attend (maybe @mikeal could ask a quick question when he mentions Code & Learn during the final notes)?

And @mikeal … not sure whether it makes a difference, but maybe we could add your “It's a workshop for people who want to learn how to contribute to Node.js core.” at the beginning and drop the whole “Concerns have also … need to be distributed geographically” part? I think that should clear up everything sufficiently.

/cc a couple of people who are here: @mhdawson @jasnell @bnoordhuis @saghul @mcollina @mikeal @thealphanerd

  • 👍5
  • 🎉3
saghul commented

Great work putting that together @addaleax! FWIW, I tagged some issues as good first contributions to libuv here: https://github.com/libuv/libuv/labels/good-first-contribution So if anyone what's to go for it, I'm happy to help as a mentor.

jasnell commented

wow... amazing list! To this I would add that there are a number of issues tagged as docs-requested that could be accessible to some participants.

sejoker commented

I would like to pickup test cleanup of test-c-ares.js from @addaleax proposed list

wzoom commented

DONE Working on test-child-process-detached.js from the comment above

MichaelTSS commented

Working on test-child-process-buffering from @addaleax's comment

alkiskal commented

Working on test-child-process-stdout-flush.js from @addaleax's comment

pmatzavin commented

Working on test-child-process-stdio.js

AdriVanHoudt commented

@MichaelTSS yeah I raised my hand ^^ I have a PR ready but you can take it if you want

MichaelTSS commented

@AdriVanHoudt keep your PR I'll do something else

AdriVanHoudt commented

@MichaelTSS Thanks and I fear @alkiskal has took it :S

yosuke-furukawa commented

I am working on

Support for SharedArrayBuffer can be added here

AdriVanHoudt commented

Taking var -> const in lib/internal/bootstrap_node.js

lrlna commented

👋 ; I am in test-path.js .

quibusus commented

working on unsafe signal warning in lib/internal/process.js

alistairjcbrown commented

Working on support for SharedArrayBuffer in lib/util.js

thomasvanlankveld commented

on test-tls-connect-any-given-socket

delvedor commented

Working on test-os.js!

akito0107 commented

I am working on test-net-server-address

hohy commented

Working on test-pipe-file-to-http.js

canastro commented

Working on test-http-agent-destroyed-socket.js

gareth-ellis commented

working on lib/v8.js, src/node_v8.cc

wzoom commented

DONE working on test-eval.js

ronaldronson commented

working on test-assert.js

RachBLondon commented

working on test/parallel/test-cluster-disconnect-idle-worker.js

graphicbeacon commented

working on test/parallel/test-async-wrap-post-did-throw.js

RachBLondon commented

working on node/test/parallel/test-cluster-disconnect-unshared-tcp.js

RachBLondon commented

working on node/test/parallel/test-cluster-disconnect-unshared-udp.js

digitalsadhu commented

working on test/parallel/test-fs-read.js

JonathanPrince commented

working on node/test/parallel/test-child-process-send-utf8.js

ilmartyrk commented

working on test/parallel/test-async-wrap-check-providers.js
test/parallel/test-async-wrap-disabled-propagate-parent.js

sejoker commented

working on test/parallel/test-child-process-exec-env.js

zvictor commented

working on lib/internal/streams/lazy_transform.js (referred above as lib/internal/lazy_transform.js)

AdriVanHoudt commented

doing more var -> const in /lib/internal/child_process.js

wzoom commented

done working on test-child-process-disconnect.js

DennisSchwartz commented

Im doing some var to const in test/parallel/test-child-process-stdin.js

ilmartyrk commented

For other beginners like me, who want to run single tests try running in console
/usr/bin/python tools/test.py --mode=release -J \ {path to test}
for example:
/usr/bin/python tools/test.py --mode=release -J \ parallel/test-async-wrap-disabled-propagate-parent

  • 👍1
wzoom commented

DONE Working on test-child-process-cwd.js

bengl commented

Is there a thread to organize this for Austin?

addaleax commented

@bengl I don’t think so, so: #59

  • 🎉1
Trott commented

Closing because this happened and planning for the next one is over in #59. As always, feel free to re-open or comment if you feel this really shouldn't be closed.

AdriVanHoudt commented

@Trott you made a slight typo in Amsterdam ;)

  • 😄1
hackygolucky commented

There is now! #60

On Thu, Oct 13, 2016, 7:02 PM Anna Henningsen notifications@github.com
wrote:

@bengl https://github.com/bengl I don’t think so, so: #59
#59

—
You are receiving this because you were mentioned.

Reply to this email directly, view it on GitHub
#56 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AB46oKRQiRiPtlaqirLpJt4PI1ndUBEKks5qzrh3gaJpZM4IKMKM
.