Re: Testing MathML spec on https://labs.w3.org/pubrules

On 02/04/2015 04:10, Denis Ah-Kang wrote:
> Hi David,
>
> We are working on giving a better feedback to the editors. Sorry for
> that.

No problem, I just thought I'd try a bit of early testing on the current
REC:-)

I'll try to give you some details about each error you encountered
> and point you to the related code.
>
> On 04/01/2015 11:17 PM, David Carlisle wrote:
>>
>> * headers.errata says errata not found
>> There is a link to the errata just below the authors in the frontmatter.
>
> Pubrules is looking for a specific sentence
> "Please check the errata for any errors or issues reported since
> publication."
> and a href on "errata".

OK I'll change our drafts to match.

>
> https://github.com/w3c/specberus/blob/master/lib/rules/headers/errata.js
>
>>
>> * headers.dl says Latest Version missing
>> as previously agreed, the spec has two latest version links
>>
>> Latest MathML 3 version:
>> Latest MathML Recommendation:
>
> This is an exception. Pubrules is actually looking for the text "Latest
> Version":
> https://github.com/w3c/specberus/blob/master/lib/rules/headers/dl.js
>
> We aren't sure yet how to handle these exceptions because it's really on
> a specification basis.

It seems quite common for specs that have been through many versions to
want to do something like this. Check for 'Latest.*[Vv]ersion'  ?


>
>>
>>
>> * headers.translations says we should use
>> http://www.w3.org/2003/03/Translations/byTechnology?technology=MathML3
>> rather than
>> http://www.w3.org/2005/11/Translations/Query?titleMatch=MathML3
>> (I can change it if that information is correct, just asking:-)
>>
>
> That's just a warning because you are not using the recommended text but
> it won't be a problem for echidna. It will not pass echidna if there's
> an error. Warnings are fine.
>
> <https://github.com/w3c/specberus/blob/master/lib/rules/headers/translations.js>
>

It's good for me (if not for echidna) not to see warnings:-)
If the 2003 URI is preferable to the 2005 one I'll change the
drafts to use that (the pages seem to be the same and at the present
time there are no translations) Or is the 2005/11 URI in fact newer
and a good link to use, and the test should not be warning?


>
>> * sotd.mailing-list says no mailing list link
>> the SOTD has a mailto: link and links to the www-math list archives and
>> to the list subscription page?
>
> Pubrules is looking for a text content matching the mailing list like:
> <a href="mailto:www-math@w3.org">www-math@w3.org</a>.
>
> <https://github.com/w3c/specberus/blob/master/lib/rules/sotd/mailing-list.js>
>
>
> It seems like a bug as there are no requirements for that. I've created
> an issue and we'll work on it.
> https://github.com/w3c/specberus/issues/174

OK thanks

>
>>
>>
>> * sotd.processDocument
>> This is I think a new requirement, can add.
>
> It is indeed a new requirement following the publication of the new
> process document.
>
> https://github.com/w3c/specberus/blob/master/lib/rules/sotd/process.js
>
> Note, the old pubrules has also been updated to check that process
> document.
>
> Hope that helps and thank you for your feedback. Do let us know if you
> have other questions.
>

David


The Numerical Algorithms Group Ltd is a company registered in England and Wales with company number 1249803. The registered office is:

Wilkinson House, Jordan Hill Road, Oxford OX2 8DR, United Kingdom.



This e-mail has been scanned for all viruses by Microsoft Office 365.

Received on Thursday, 2 April 2015 08:59:47 UTC