ECMAScript 2017 HTML version - search broken, various 404s and script errors

[farsightsoftware]

The search box in the HTML version of the latest released spec, http://www.ecma-international.org/ecma-262/8.0/index.html, is broken (typing something, say "class", doesn't find it). There are various 404s and script errors as you use the page.

Looks like a deployment issue on the ECMA site, these issues don't happen at https://tc39.github.io/ecma262/.

404s:

• http://www.ecma-international.org/ecma-262/8.0/ECMAScript%C2%AE%202017%20Language%20Specification_files/ecmarkup_002.htm
• http://www.ecma-international.org/ecma-262/8.0/ECMAScript%C2%AE%202017%20Language%20Specification_files/github.css
• http://www.ecma-international.org/ecma-262/8.0/ECMAScript%C2%AE%202017%20Language%20Specification_files/ecmarkup.htm

There's also an unexpected token error:

index.html:23 Uncaught SyntaxError: Unexpected token <

Presumably as a result of the 404s, mousing over various things causes errors like this:

index.html:685 Uncaught TypeError: Cannot read property
'prod-grammar-notation-StatementList' of undefined
    at Object.activateIfMouseOver (index.html:695)
    at HTMLDivElement.<anonymous> (index.html:501)
    at HTMLDivElement.<anonymous> (index.html:520)

(I posted this on es-discuss several days ago, but the suggestion was to open an issue here.)

[IgnoredAmbience]

A couple of other issues:

• An erroneous (and unclosed?) <script> tag has been inserted above the <title> tag.
• The favicon URL is unstable, pointing to https://tc39.github.io/ecma262/2017/img/favicon.ico. The 2017 directory is likely to disappear when 2018 is progressed.
• There's a few absolute links pointing to the draft standard, instead of being relative to the current document. (grep the html source for tc39.github.io/ecma262/#) (these seem to persist in the original source though, so I'll raise a PR as appropriate for this.)

The 404ing URLs appear to be referencing the directory created by a browser's "Save Page" functionality (complete with the © symbol from the page title).

[IgnoredAmbience]

A couple of other issues:

• An erroneous (and unclosed?) <script> tag has been inserted above the <title> tag.
• The favicon URL is unstable, pointing to https://tc39.github.io/ecma262/2017/img/favicon.ico. The 2017 directory is likely to disappear when 2018 is progressed.
• There's a few absolute links pointing to the draft standard, instead of being relative to the current document. (grep the html source for tc39.github.io/ecma262/#) (these seem to persist in the original source though, so I'll raise a PR as appropriate for this.)

The 404ing URLs appear to be referencing the directory created by a browser's "Save Page" functionality (complete with the © symbol from the page title).

[bterlson]

I will look in to this, thank you!

[bterlson]

I will look in to this, thank you!

[tjcrowder]

Bump. I refer people to the specification all the time. Would be better if it were working properly.

[tjcrowder]

Bump. I refer people to the specification all the time. Would be better if it were working properly.

[domenic]

Don't refer them to the outdated snapshot on that website. Refer them to the spec implemented by browsers at https://tc39.github.io/ecma262/ .

[domenic]

Don't refer them to the outdated snapshot on that website. Refer them to the spec implemented by browsers at https://tc39.github.io/ecma262/ .

[tjcrowder]

@domenic - Well, implemented by varying browsers at varying times. That's beside the point. The most recent snapshot should work properly.

[tjcrowder]

@domenic - Well, implemented by varying browsers at varying times. That's beside the point. The most recent snapshot should work properly.

[domenic]

No, ecma262 only includes things which have reached stage 4, and are implemented in multiple browsers.

The snapshot should generally not be referred to, especially by programmers.

[domenic]

No, ecma262 only includes things which have reached stage 4, and are implemented in multiple browsers.

The snapshot should generally not be referred to, especially by programmers.

[tjcrowder]

@domenic - Useful to know (where is that documented?). It's still beside the point. The snapshot should work properly, and it's perfectly valid to need to refer someone to how things were as of ES2017.

[tjcrowder]

@domenic - Useful to know (where is that documented?). It's still beside the point. The snapshot should work properly, and it's perfectly valid to need to refer someone to how things were as of ES2017.

[domenic]

I'll just leave it as "I disagree"; I think the less effort people spend on the snapshot, the better, so as to discourage people from referring to it. This is just my opinion, but I do think it'd be good to tone down your expectation that other people do work for you to fit your opinion.

[domenic]

I'll just leave it as "I disagree"; I think the less effort people spend on the snapshot, the better, so as to discourage people from referring to it. This is just my opinion, but I do think it'd be good to tone down your expectation that other people do work for you to fit your opinion.

[tjcrowder]

@domenic I don't deserve that slap. I've offered to help in the past and would happily help with this if it were something I could help with.

[tjcrowder]

@domenic I don't deserve that slap. I've offered to help in the past and would happily help with this if it were something I could help with.