[WIP] Add raw data generation#977
[WIP] Add raw data generation#977chicoxyzzy wants to merge 1 commit intocompat-table:gh-pagesfrom chicoxyzzy:raw-data
Conversation
|
Oh I forgot to generate results for ranges and |
ljharb
left a comment
There was a problem hiding this comment.
Why not gitignore raw-data, since this is published to npm?
package.json
Outdated
| "type": "git", | ||
| "url": "git://github.com/kangax/compat-table.git" | ||
| }, | ||
| "files": ["raw-data"], |
There was a problem hiding this comment.
let's please not use the "files" key - this is a dangerous way to exclude things from npm. In addition, there's no need to exclude anything imo - https://unpkg.com/compat-table@0.0.1/ already has a bunch of other stuff.
| .ruby-version | ||
| _site/ | ||
| .idea/ | ||
| es5/ |
There was a problem hiding this comment.
I don't see any reason why they should be downloaded when you need just test results from npm. Source code is available from GH because npm supports GH and Git
There was a problem hiding this comment.
IMO there should be only test results and environments
There was a problem hiding this comment.
github is for development, npm is for consumption. packages published to npm should have everything - that way it's available when consuming previously-installed-things offline as well.
There was a problem hiding this comment.
The only things that other projects need is raw data. Why publish HTML files?
There was a problem hiding this comment.
Why wouldn't they need the HTML files if they wanted to display them, or if they want to be able to view the HTML pages without an internet connection?
.npmignore
Outdated
| esnext/ | ||
| esintl/ | ||
| non-standard/ | ||
| strict-mode/ No newline at end of file |
There was a problem hiding this comment.
Please ensure all files have a trailing newline
package.json
Outdated
| { | ||
| "name": "compat-table", | ||
| "version": "0.0.0", | ||
| "version": "0.1.0", |
There was a problem hiding this comment.
Version numbers should be bumped only in commits directly to master; not in PRs
|
I'll create new PR instead of this one |
|
Ooh now I see what the goal is. 🤔 It wasn't immediately clear to me in #965. So- you seek the results without the test functions. Maybe they should be separated altogether. Although- that seems like it would add undesirable complication to this codebase. |
|
I think it would. |
I found it much harder to build all data from JSON files (as I wanted in #965) than generating raw data from current js data files. Having js files allows to have autocomplete in editors when using shared data from 'data-common.json'. Also removing tests is not so trivial and it doesn't brings any significant profits.
Also we need to register npm name and setup CI for automated package version bumping and publishing.