Commit 6357ef1
doc: document considerations for inclusion in core
Document the things that are considered when making the determination as
to whether something should or shouldn't be in core. This does not (yet,
at least) attempt to address *how* to include modules in core. (Should
it be in the Node.js code base or vendored in from a separate
repository?) It is limited to *whether* something should be in core or
not.
Closes: nodejs/TSC#1041
PR-URL: #40338
Fixes: nodejs/TSC#1041
Reviewed-By: Robert Nagy <ronagy@icloud.com>
Reviewed-By: Michael Dawson <midawson@redhat.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Michaël Zasso <targos@protonmail.com>1 parent ef1ace7 commit 6357ef1
1 file changed
Lines changed: 55 additions & 0 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
@@ -0,0 +1,55 @@ | |||
| 1 | + # To be or not to be in core | ||
| 2 | + | ||
| 3 | + Should a module be in core? This question arises every so often. This document | ||
| 4 | + explains things to consider when deciding whether a module should be in core or | ||
| 5 | + not. | ||
| 6 | + | ||
| 7 | + ## Strong arguments for including a module in core | ||
| 8 | + | ||
| 9 | + 1. The module provides functionality that is standardized (such as a | ||
| 10 | + [Web API][]) and overlaps with existing functionality. | ||
| 11 | + 2. The module can only be implemented in core. | ||
| 12 | + 3. The module can only be implemented in a performant way in core. | ||
| 13 | + 4. Developer experience is significantly improved if the module is in core. | ||
| 14 | + 5. The module provides functionality that can be expected to solve at least one | ||
| 15 | + common use case Node.js users face. | ||
| 16 | + 6. The module requires native bindings. Inclusion in core enables utility across | ||
| 17 | + operating systems and architectures without requiring users to have a native | ||
| 18 | + compilation toolchain. | ||
| 19 | + 7. Part or all of the module will also be re-used or duplicated in core. | ||
| 20 | + | ||
| 21 | + ## Strong arguments against including a module in core | ||
| 22 | + | ||
| 23 | + 1. None of the arguments list in the previous section apply. | ||
| 24 | + 2. The module has a license that prohibits Node.js from including it in core | ||
| 25 | + without also changing its own license. | ||
| 26 | + 3. There is already similar functionality in core and adding the module will | ||
| 27 | + provide a second API to do the same thing. | ||
| 28 | + 4. A module (or/and the standard it is based on) is deprecated and there is | ||
| 29 | + a non-deprecated alternative. | ||
| 30 | + 5. The module is evolving quickly and inclusion in core will require frequent | ||
| 31 | + API changes. | ||
| 32 | + | ||
| 33 | + ## Benefits and challenges | ||
| 34 | + | ||
| 35 | + When it is unclear whether a module should be included in core, it might be | ||
| 36 | + helpful to consider these additional factors. | ||
| 37 | + | ||
| 38 | + ### Benefits | ||
| 39 | + | ||
| 40 | + 1. The module will receive more frequent testing with Node.js CI and CITGM. | ||
| 41 | + 2. The module will be integrated into the LTS workflow. | ||
| 42 | + 3. Documentation will be integrated with core. | ||
| 43 | + 4. There is no dependency on npm. | ||
| 44 | + | ||
| 45 | + ### Challenges | ||
| 46 | + | ||
| 47 | + 1. Inclusion in core is likely to reduce code merging velocity as the Node.js | ||
| 48 | + process for code review and merging is more time-consuming than that of most | ||
| 49 | + individual modules. | ||
| 50 | + 2. By being bound to the Node.js release cycle, it is harder and slower to | ||
| 51 | + publish patches. | ||
| 52 | + 3. Less flexibility for end users. They can't update the module when they choose | ||
| 53 | + without also updating Node.js. | ||
| 54 | + | ||
| 55 | + [Web API]: https://developer.mozilla.org/en-US/docs/Web/API | ||
0 commit comments