KR: Q2-Q3 data quality goal: Real values for all CSS features #3710
Comments
|
@Elchi3 are you using a script to generate the sub-issues? |
|
@connorshea Nothing fancy, just a hacky walking function. (which you're welcome to improve) function traverseFeatures(obj, depth, identifier) {
depth--;
if (depth >= 0) {
for (let i in obj) {
if (!!obj[i] && typeof(obj[i])=="object" && i !== '__compat') {
if (obj[i].__compat) {
let comp = obj[i].__compat.support;
let browser = comp['safari'];
if (!Array.isArray(browser)) {
browser = [browser];
}
for (let range in browser) {
if (browser[range].version_added === null || browser[range].version_removed === null ||
browser[range].version_added === true || browser[range].version_removed === true ) {
features.push(identifier + i);
}
}
}
traverseFeatures(obj[i], depth, i + '.');
}
}
}
}
let features = [];
traverseFeatures(bcd.css, 100, 'css.');
console.log(features.join("\n")); |
|
I think that Chrome Android and WebView Android (as is Opera and Opera Android) can be considered a part of Chrome Desktop's updates. Everything that Chrome has but all the other Chromium browsers don't is taken care of by #3869, mirroring the properties over. I'm gonna do my best to tackle a number of CSS features, get the |
|
In that case, I'd say I'll create another issue for Edge and for the other browsers let's then see if we can get to those real values by using mirroring. |
|
|
|
I am not aware of any feature where CSS varies from Chrome desktop on Android and and WebView. Unfortunately, I can't exclude it either. |
|
Made some modifications to that script you wrote, @Elchi3, to allow easier browser and value changes, as well as a quick fix to printing the identifier. 😉 Edit: the mentioned script has been merged and can be accessed via |
|
Update as of May 3rd, 2019 (two more months to go)
|
|
Which of the other browsers are going into the next sprint (Koko Taylor)?
They need estimates. |
|
Firefox is almost done, so I would imagine that would be the first one. Chrome Android is a mirror of Chrome Desktop, so they’ll both be finished together. |
|
Safari and Chrome are the most important ones for now I would think and work on them will likely continue from this sprint as they aren't completed yet. Might want to add Edge to the next sprint as well. Firefox is already meeting the acceptance criteria of 95%, but I'm going to try to bring it over the real finish line. |
|
Can we get an updated chart? :) |
|
Coming right up! Update as of today (May 27, 2019):
|
|
Q2 is coming to an end and we aren't done with this yet. Please see the above referenced issues for the remaining work that we need to do here. The two gigantic issues for Chrome and Safari weren't really practical and we couldn't see any end with them, so I've re-shuffled and re-estimated this work now. All issues are marked with "help wanted", please assign them to you if you intend to do work on them or comment in the issues which parts of the tasks you want to get done. Any help is appreciated. |
|
Oh and the issues I've filed don't categorize everything, so there is still quite a bit of rest left to do afterwards, but we can get to that once we have the above things done. |
|
Status as of version 0.0.86 (released on 2019-07-11) for web platform features:
|
|
Status as of version 0.0.89 (released on 2019-08-08) for web platform features:
|
|
We've now finished this for the 8 selected browsers. It is now forbidden to add
Note that for webview android we introduced a ranged value "≤37". We have no plans to find out in which exact old version css features are implemented in this browser. If someone has a need for this, please let us know. |
|
This was a huge amount of work and an outstanding achievement. Congratulations on guiding this to a successful conclusion! |
|
A huge thank you and shout out to @ddbeck and @vinyldarkscratch for helping out on this over many months ❤️ |
|
Congratulations! This makes me and my coworkers tremendously happy.
Joe Medley | Technical Writer, Chrome DevRel | jmedley@google.com |
816-678-7195
*If an API's not documented it doesn't exist.*
…On Thu, Aug 29, 2019 at 9:15 AM Florian Scholz ***@***.***> wrote:
A huge thank you and shout to @ddbeck <https://github.com/ddbeck> and
@vinyldarkscratch <https://github.com/vinyldarkscratch> for helping out
on this over many months ❤️
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#3710?email_source=notifications&email_token=AB6S7C2KJZO7RT65VCJFBOLQG7Y3TA5CNFSM4HA6BVJ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5PBALA#issuecomment-526258220>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AB6S7C47HUME2F25BT6UX3LQG7Y3TANCNFSM4HA6BVJQ>
.
|
|
Well done, you BCD folks! Truly an inspiring effort to get this done. |
To support the year goal and those tools that are already consuming compat data, we (the MDN team) would like to get to 100% real values for all CSS features in Q2 (VS Code, WebHint.io, WebStorm are using mostly CSS compat data at this point, so getting CSS data correct has impact on multiple large projects). Also, having one area complete with real values might inform us what could be the next iteration in terms of data quality.
The statistics script can be used to gather the current status here as well (I've just changed the if condition in line 58 to be just
if (data === 'css')) and this is the output:truevaluesnullvaluesThis Q2 goal will probably mean that some peers will prioritize reviews for CSS data updates over other (large) data updates during the quarter (April to June).
Depends on:
The text was updated successfully, but these errors were encountered: