Close <head> tags #177
Close <head> tags #177
Conversation
The `.editorconfig` set the indent style to 4 spaces, while the `.jshintrc` and templates were using two spaces.
Update editorconfig to use 2 spaces indention
Test runner fix for Windows
|
What do you mean by "max compatibility"? They will be parsed correctly in all browsers without being self closed. |
|
Allows you to set the 'application/xhtml+xml' media type in the HTTP response without breaking the site. Please consider reopening this pull request as it is very useful to use XHTML media type for older mobile browsers. |
|
I'd argue that's an edge case. It's easy enough to change them yourself. |
|
And what's the down-side of closing the tags in the template? |
|
We don't need them, that's why this doesn't get pulled in. You think we should add |
|
@samaxes Just curious: what older browsers? While it'd be nice to support something like this (for those edge cases), I think we'd need to do a lot of work to make these changes. I don't personally want to be accountable for maintaining XHTML support, especially since the trend seems to be away from it. Yeoman is opinionated, and I think it's fine for us to have "HTML5 doctype" as an opinion. |
|
No, I don't as they are optional in last HTML spec. |
|
@btford I'm talking about the serialization/media type here, not the spec. HTML5 spec allows you to use both HTML5 or XHTML5 serialization: http://html5doctor.com/html-5-xml-xhtml-5/. |
|
|
|
In addition, Yeoman in general builds on top of HTML5 Boilerplate as far as that is applicable within the scope of the generator. |
|
@kevva Exactly, you can serve HTML5 documents (using HTML5 doctype) but using XHTML5 serialization with the correct media type (which is |
|
@passy I don't get the relationship with the issue at hand. |
|
We try to minimize differences if possible, this creates new ones. |
|
@passy which differences? Currently the template closes some tags in the head element and not the others. Do you have a reason for that? |
|
@samaxes That's a valid point. Feel free to submit a PR that removes the trailing slash for the inconsistent ones. :) |
|
@passy That can be done. But I believe that you, like me, would prefer the opposite ;) |
|
I see your point, but it doesn't make sense to partly support XHTML. There would still be errors as I said previously. We would have to maintain a polyglot version of this inside a separate branch to fully support it or maybe have an option like That's a lot of work to cater for a very small group of people. |
|
Forget about XHTML, because as you say, it's a lot more complicate that this and you would have to do a lot more stuff to be "XHTML compliant" (like using data-* attributes: |
|
@samaxes, exactly. That's my point. So why should we pull this in? What says it's a best practice? |
|
@kevva We use the Polyglot Markup spec for best practice guidelines:
|
|
Yes, according to that spec it's the best practices. But we're using HTML5 which says you can use both. And as @passy said above, we're trying to keep the differences from h5bp as few as possible hence no self closing tags. |
|
Polyglot Markup spec is the HTML5 best practices guidelines. The HTML5 spec does not define best practices for web authors, it is more oriented for user agents. |
|
@samaxes This is probably the wrong place, though. You should lobby HTML5 Boilerplate instead. If they change to XHTML5, we will certainly consider changing the base templates. |
|
I agree that we should be consistent in how we handle self-closing tags, but I don't feel strongly one way or another. I think @passy is right about sticking to what H5BP does for now. But I don't want to dismiss @samaxes use case. I'd love to hear from @necolas or @paulirish and see if they think H5BP can/should adhere to the Polygot Markup Spec. At very least, I plan to support multiple markup languages in the near future (see #74). With this in place, it'd be easy to add an option for this. |
|
I do agree that if you use H5BP as the base for your HTML templates you should stick with them. |
|
We get |
|
Who is writing XML-based web apps? |
|
You tell me. An edge case to say the least. |
|
That's my point. You never need self-closing tags when you're writing web applications. |
|
Exactly, that's why I disagree with this change. I wouldn't consider the Polyglot spec best practice for authors either. Maybe if you want your document to be both valid HTML and XML which, to my knowledge, is very rare. |
|
@necolas So you believe that XHTML5 serialization should be removed from the HTML5 spec and the use cases for it to be included do not make any sense? |
No. |
|
My point is that the amount of effort needed to implement the best practice guidelines from the Polyglot Markup spec is negligible. However, allowing the web app author to freely choose the serialization he likes the most is a huge benefit. |
|
Please continue this discussion elsewhere. Thanks. |
searchPath parameter for wiring.appendFiles
Lower case the user email field in when passport queries to find a user during login. Also lowercase the email on user creations. Fix yeoman#177
No description provided.