|
| 1 | +# Node.js Technical Steering Committee (TSC) Meeting 2024-12-11 |
| 2 | + |
| 3 | +## Links |
| 4 | + |
| 5 | +* **Recording**: <https://www.youtube.com/watch?v=NqtxLJ7_S0g> |
| 6 | +* **GitHub Issue**: <https://github.com/nodejs/TSC/issues/1662> |
| 7 | + |
| 8 | +## Present |
| 9 | + |
| 10 | +* Ruben Bridgewater @BridgeAR (voting member) |
| 11 | +* Gireesh Punathil @gireeshpunathil (voting member) |
| 12 | +* Joyee Cheung @joyeecheung (voting member) |
| 13 | +* Marco Ippolito @marco-ippolito (voting member) |
| 14 | +* Matteo Collina @mcollina (voting member) |
| 15 | +* Michael Dawson @mhdawson (voting member) |
| 16 | +* Rafael Gonzaga @RafaelGSS (voting member) |
| 17 | +* Richard Lau @richardlau (voting member) |
| 18 | +* Ruy Adorno @ruyadorno (voting member) |
| 19 | +* Paolo Insogna @ShogunPanda (voting member) |
| 20 | + |
| 21 | +## Agenda |
| 22 | + |
| 23 | +### Announcements |
| 24 | + |
| 25 | +* No news this week |
| 26 | + |
| 27 | +### Reminders |
| 28 | + |
| 29 | +* Remember to nominate people for the [contributor spotlight](https://github.com/nodejs/node/blob/main/doc/contributing/reconizing-contributors.md#bi-monthly-contributor-spotlight) |
| 30 | + |
| 31 | +### CPC and Board Meeting Updates |
| 32 | + |
| 33 | +*Extracted from **tsc-agenda** labeled issues and pull requests from the **nodejs org** prior to the meeting. |
| 34 | + |
| 35 | +* no updates updates this week. |
| 36 | +* Matteo, watch Robin’s talk on budget from NodeConf.eu after it is published. |
| 37 | + |
| 38 | +### nodejs/node |
| 39 | + |
| 40 | +* fs: enable chunked reading for large files in readFileHandle [#56022](https://github.com/nodejs/node/pull/56022) |
| 41 | + * Ruben: do we want the possibility of readfile to read huge files into memory or just accept |
| 42 | + limit and change the error case. Cannot thinking of use case where it makes sense to load |
| 43 | + anything bigger than 2G |
| 44 | + * Michael: Is there some downside to allowing larger sizes? Ruben, on linux it would |
| 45 | + require multiple reads behind the scenes. |
| 46 | + * Matteo: people seem to be upset if you try to prevent people from |
| 47 | + shooting themselves in the foot. Thinks its the right call to keep the limit, but will |
| 48 | + not oppose if people want to allow it. |
| 49 | + * Michael: ok if we leave the limit in place. It is possible for people work work around it |
| 50 | + * Matteo: ok to allow it and crash with out of memory, would accept PR. |
| 51 | + * Joyee: don’t see a reason to throw and error, if they want to take the risk, document |
| 52 | + that there are better ways to do it. Would be fine for throw away scripts |
| 53 | + * Ruben: more worried about people who don’t read the docs, and error would be a way |
| 54 | + For them to learn about streams |
| 55 | + * Joyee: we could emit a warning |
| 56 | + * Ruben: would be ok with allowing and emitting a warning. We could even make the limit |
| 57 | + a bit lower, say 1G? |
| 58 | + * Ruben: should we say recommendation is to emit warning and allow |
| 59 | + * Gireesh, should we emit a warning for a low level function like this? |
| 60 | + * Ruben, can easily cause the process to crash and out of memory issues, this would be |
| 61 | + useful to help users understand why they are out of memory. |
| 62 | + * Joyee could also check how memory is available. Maybe only warn if there |
| 63 | + is not going to be enough memory. |
| 64 | + * Gireesh, should documentation not be enough? |
| 65 | + * Ruben, people don’t read the document often. |
| 66 | + * Matteo: having the warning at the previous limit is fine. We have other places where |
| 67 | + There are similar cases (ex more than 10 listeners) |
| 68 | + * Gireesh: but those are bad coding practice |
| 69 | + * Matteo: I see this case (loading a file bigger than 2G) is similar |
| 70 | + * Joyee: agree, if you want to be lazy ok so we don’t want to prevent, but warning is ok. |
| 71 | + * Ruben: Do we know a legitimate use case for loading a file of this size into memory? |
| 72 | + * Matteo: top of line mac now has 128g of RAM. |
| 73 | + * Michael: warning seems like a compromise between not allowing big files and allowing |
| 74 | + without doing anything. |
| 75 | + * We did not reach consensus on including a warning |
| 76 | + * We will leave on the agenda to discuss next week |
| 77 | + |
| 78 | +* test: improve zlib tests [#55716](https://github.com/nodejs/node/pull/55716) |
| 79 | + * waiting for <https://github.com/nodejs/node/pull/56027> to land as next step. |
| 80 | + |
| 81 | +### nodejs/TSC |
| 82 | + |
| 83 | +* Clarify the Charter so that contractors are explicity counted as affialiated [#1650](https://github.com/nodejs/TSC/pull/1650) |
| 84 | + * Matteo: ok either way did not plan to be that specific, but happy with suggestion from Ulises |
| 85 | + * Matteo, perhaps propose something different. |
| 86 | + * Ruben: it is a rule defined by the TSC for the TSC. We should not overcomplicate things, since we can change this at any point of time |
| 87 | + * Richard: missing why we have the sentence (about affiliation) in the first place. Lets provide |
| 88 | + more context as to the goal and leave the the rest less specific. |
| 89 | + * Matteo: agree doing change in that way would be better. |
| 90 | + * Michael: would prefer avoiding trying to list the specifics as I doubt we can get it right |
| 91 | + * Matteo: will do another pass |
| 92 | + |
| 93 | +* Let's talk about the CI situation [#1614](https://github.com/nodejs/TSC/issues/1614) |
| 94 | + * Richard, Yagiz wanted to talk about the smartOS situation - <https://github.com/nodejs/TSC/issues/1663> |
| 95 | + * We are in the middle of transitioning some machines. Had a bunch sponsored by |
| 96 | + Joyent. They were moved to equinix Metal. Equinix then asked that we move our non-arm |
| 97 | + out of equinix. MNX volunteered to host these machines. Many have been moved over |
| 98 | + but smartOS ones are still in progress. Also tied with getting the Linux Foundation more |
| 99 | + Involved. |
| 100 | + * Matteo suggested we set a deadline. Area of the project that is not getting enough, |
| 101 | + not enough volunteers, so unless people step up. |
| 102 | + * Challenge, is that the changes require latest C++, so can’t just exclude tests. |
| 103 | + * Same PRs -> MacOS and SmartOS are ones that need newer compilers |
| 104 | + * Richard having deadlines likely helps focus. |
| 105 | + |
| 106 | +* Open OpenCollective and Github Sponsors for Node.js [#1553](https://github.com/nodejs/TSC/issues/1553) |
| 107 | + * No objections, going to land. |
| 108 | + |
| 109 | +## Strategic Initiatives |
| 110 | +* Skipped as we were out of time |
| 111 | + |
| 112 | +## Upcoming Meetings |
| 113 | + |
| 114 | +* **Node.js Project Calendar**: <https://nodejs.org/calendar> |
| 115 | + |
| 116 | +Click `+GoogleCalendar` at the bottom right to add to your own Google calendar. |
0 commit comments