Re: [whatwg] use of article to markup comments

[Silvia Pfeiffer]

On Mon, Feb 18, 2013 at 12:19 AM, Nils Dagsson Moskopp <
n...@dieweltistgarnichtso.net> wrote:

> "Bruce Lawson"  schrieb am Sat, 26 Jan 2013 13:30:18
> -:
>
> > In short, why should the spec suggest any specific method of marking
> > up comments?
>
> As someone who is interested in semantics and tired of scraping
> content and applying scrappy heuristics: If it is clear that an
>  within an  represents a comments one can easily:
>


 in  could be a comment. Or it could be something else
entirely. Your heuristic may work in many cases, but certainly not in all.
If we really wanted to be sure where to find the semantic concept of a
comment, we should introduce a meaningful element for it such as .

Silvia.

Re: [whatwg] use of article to markup comments

[Nils Dagsson Moskopp]

Adrian Testa-Avila  schrieb am Sun, 27 Jan
2013 10:31:06 -0800:

> So, maybe a better question is why should the spec suggest only one 
> specific method?

Answer: As someone consuming HTML and XML content, I find it extremely
unhelpful if equivalent semantics are expressed in mand different ways.

-- 
Nils Dagsson Moskopp // erlehmann

Re: [whatwg] use of article to markup comments

[Nils Dagsson Moskopp]

"Bruce Lawson"  schrieb am Sat, 26 Jan 2013 13:30:18
-:

> In short, why should the spec suggest any specific method of marking
> up comments?

As someone who is interested in semantics and tired of scraping
content and applying scrappy heuristics: If it is clear that an
 within an  represents a comments one can easily:
  * programmatically find article comments in HTML
  * write interoperable stylesheets for comments, using the selector
“article > article”
  * use HTML fragments in a document store for content management (I
wrote a blog software with a git backend yesterday and plan to add
this feature)

Without having one interoperable way all that becomes a lot harder.

-- 
Nils Dagsson Moskopp // erlehmann

Re: [whatwg] use of article to markup comments

[Gordon P. Hemsley]

List elements and sectioning elements both represent hierarchical
relationships. They differ in how they emphasize that relationship:
lists emphasize the hierarchy outside the content, while sectioning
emphasizes the hierarchy within the content.

If the question is specifically about how to mark up comments on a
blog post or something, there's no reason you can't combine the two
methods: Each comment is a self-contained , with
relationships between comments represented by .

One example:
http://jsbin.com/edewoy/1

That example presumes you consider blog post comments (or replies to
comments) as a section within the content that is being commented on
(or replied to). You could also modify the markup to have two
s (one for the blog post and one for the comments) packaged
within a single parent , but the principle is the same.

Note that the key here is that there is no restriction on combining
lists and sectioning elements, and thereby no need to modify the
semantics of  or  (as proposed in [2] in the root message).

Gordon

On Mon, Jan 28, 2013 at 12:13 PM, Steve Faulkner
 wrote:
>> Brucel wrote:
>>
>> On Sat, 26 Jan 2013 10:56:10 -, Steve Faulkner
>>  wrote:
>>
>>
>> > Lists are appropriate for indicating nested tree structures. The use
>> > of lists to markup comments is a common mark up pattern used in
>> > blogging software such as wordpress. The code verbosity is not
>> > dissimilar to  the use of article, less so even option end  tags
>> > are omitted. Besides comments are generated code not hand authored so
>> > I don't see a problem with code verbosity
>>
>> [...]
>>
>> >
>> >> (It makes some sense, I suppose, to think of comments as a "list", but
>> >> *unordered*? If you're going to group them at all, wouldn't the order
>> >> be important? Bruce Lawson (
>> >> http://lists.w3.org/Archives/Public/public-html/2013Jan/0111.html)'s
>> >> observation that comments are "heavily dependent on context" would seem
>> >> to support the idea that it *is* important, especially since some
>> >> comments are responses to others.)
>> >
>> > agreed it would be better to use order lists.
>> >
>>
>>   Wordpress blogs, for example, have comments like
>>
>> "Bob Smith said at 9.55 on 31 Febtember: LOL"
>>
>> Thus, every comment has a link that a UA can use to jump from comment to
>> comment. The order is implied via the timestamp. So what's wrong with
>>
>> 
>> Witty blogpost
>> lorem ipsum
>>
>> 
>> 35 erudite and well-reasoned comments
>> Bob Smith said at 9.55 on 31 Febtember: Can
>> I use DRM in Polyglot documents?
>> Hixie said at 9.57 on 1 June: What's your
>> use case?
>> ...
>> 
>>
>> 
>>
>> In short, why should the spec suggest any specific method of marking up
>> comments?
>
> Good question, in the case of  recommended tomarkup comments
> it seems like it's an element in search of a use case.
>
> For users who consume article semantics it appear to cause issues when
> used for any piece of content ranging from a one sentence comment to
> an article containing thousands of words or an interactive widget.
>
>
> regards
> SteveF



-- 
Gordon P. Hemsley
m...@gphemsley.org
http://gphemsley.org/ • http://gphemsley.org/blog/

Re: [whatwg] use of article to markup comments

[Steve Faulkner]

> Brucel wrote:
>
> On Sat, 26 Jan 2013 10:56:10 -, Steve Faulkner
>  wrote:
>
>
> > Lists are appropriate for indicating nested tree structures. The use
> > of lists to markup comments is a common mark up pattern used in
> > blogging software such as wordpress. The code verbosity is not
> > dissimilar to  the use of article, less so even option end  tags
> > are omitted. Besides comments are generated code not hand authored so
> > I don't see a problem with code verbosity
>
> [...]
>
> >
> >> (It makes some sense, I suppose, to think of comments as a "list", but
> >> *unordered*? If you're going to group them at all, wouldn't the order
> >> be important? Bruce Lawson (
> >> http://lists.w3.org/Archives/Public/public-html/2013Jan/0111.html)'s
> >> observation that comments are "heavily dependent on context" would seem
> >> to support the idea that it *is* important, especially since some
> >> comments are responses to others.)
> >
> > agreed it would be better to use order lists.
> >
>
>   Wordpress blogs, for example, have comments like
>
> "Bob Smith said at 9.55 on 31 Febtember: LOL"
>
> Thus, every comment has a link that a UA can use to jump from comment to
> comment. The order is implied via the timestamp. So what's wrong with
>
> 
> Witty blogpost
> lorem ipsum
>
> 
> 35 erudite and well-reasoned comments
> Bob Smith said at 9.55 on 31 Febtember: Can
> I use DRM in Polyglot documents?
> Hixie said at 9.57 on 1 June: What's your
> use case?
> ...
> 
>
> 
>
> In short, why should the spec suggest any specific method of marking up
> comments?

Good question, in the case of  recommended tomarkup comments
it seems like it's an element in search of a use case.

For users who consume article semantics it appear to cause issues when
used for any piece of content ranging from a one sentence comment to
an article containing thousands of words or an interactive widget.


regards
SteveF

Re: [whatwg] use of article to markup comments

[Adrian Testa-Avila]

On 01/26/2013 05:30 AM, Bruce Lawson wrote:





(It makes some sense, I suppose, to think of comments as a "list", but
*unordered*? If you're going to group them at all, wouldn't the order
be important? Bruce Lawson (
http://lists.w3.org/Archives/Public/public-html/2013Jan/0111.html)'s
observation that comments are "heavily dependent on context" would seem
to support the idea that it *is* important, especially since some
comments are responses to others.)


agreed it would be better to use order lists.



 Wordpress blogs, for example, have comments like

"Bob Smith said at 9.55 on 31 Febtember: LOL"

Thus, every comment has a link that a UA can use to jump from comment 
to comment. The order is implied via the timestamp. So what's wrong with



Witty blogpost
lorem ipsum


35 erudite and well-reasoned comments
Bob Smith said at 9.55 on 31 Febtember: 
Can I use DRM in Polyglot documents?
Hixie said at 9.57 on 1 June: What's 
your use case?

...




In short, why should the spec suggest any specific method of marking 
up comments?


I think examples are useful for clearly illustrating the spec.  An 
example in the spec shouldn't be construed as "the only right way" of 
doing things, of course.


So, maybe a better question is why should the spec suggest only one 
specific method?



--
Adrian Testa-Avila
adr...@custom-anything.com 
follow on facebook 

Re: [whatwg] use of article to markup comments

[Bruce Lawson]

On Sat, 26 Jan 2013 10:56:10 -, Steve Faulkner  
 wrote:




Lists are appropriate for indicating nested tree structures. The use
of lists to markup comments is a common mark up pattern used in
blogging software such as wordpress. The code verbosity is not
dissimilar to  the use of article, less so even option end  tags
are omitted. Besides comments are generated code not hand authored so
I don't see a problem with code verbosity


[...]




(It makes some sense, I suppose, to think of comments as a "list", but
*unordered*? If you're going to group them at all, wouldn't the order
be important? Bruce Lawson (
http://lists.w3.org/Archives/Public/public-html/2013Jan/0111.html)'s
observation that comments are "heavily dependent on context" would seem
to support the idea that it *is* important, especially since some
comments are responses to others.)


agreed it would be better to use order lists.



 Wordpress blogs, for example, have comments like

"Bob Smith said at 9.55 on 31 Febtember: LOL"

Thus, every comment has a link that a UA can use to jump from comment to  
comment. The order is implied via the timestamp. So what's wrong with



Witty blogpost
lorem ipsum


35 erudite and well-reasoned comments
Bob Smith said at 9.55 on 31 Febtember: Can  
I use DRM in Polyglot documents?
Hixie said at 9.57 on 1 June: What's your  
use case?

...




In short, why should the spec suggest any specific method of marking up  
comments?


--


Bruce Lawson
Open standards evangelist
Developer Relations Team
Opera

http://dev.opera.com

Re: [whatwg] use of article to markup comments

[Adrian Testa-Avila]

On 01/25/2013 03:44 AM, Steve Faulkner wrote:
Over on the HTML WG list [1] we have been discussing the use of the 
article

element to mark up comments.

I have sketched out a few alternative possibilities to the current
recommendation in the spec of using the article element as it has been
indicated by users who consume the semantics that its use is suboptimal.

I am bringing this over here as well as one of the suggestions I have [2]
made may include implementation changes



[1]
http://lists.w3.org/Archives/Public/public-html/2013Jan/thread.html#msg109 


[2] http://www.html5accessibility.com/tests/comments.html


I would be concerned that recommending the use of  would simply 
confuse matters further, and lead to messier, unnecessarily verbose 
markup.  A clean list of comments is fine, but most comments end up 
tree-structured, not lists.  Tree-s are quite verbose compared to 
simply nesting s.


(It makes some sense, I suppose, to think of comments as a "list", but 
*unordered*?  If you're going to group them at all, wouldn't the order 
be important?  Bruce Lawson ( 
http://lists.w3.org/Archives/Public/public-html/2013Jan/0111.html)'s 
observation that comments are "heavily dependent on context" would seem 
to support the idea that it *is* important, especially since some 
comments are responses to others.)


Robin Berjorn ( 
http://lists.w3.org/Archives/Public/public-html/2013Jan/0115.html) made 
a suggestion to wrap comment  elements inside a  
element, which seems like a good approach, both in terms of semantics 
and practical result.


*However*, I don't see anything wrong or confusing about nested 
s.  I think it makes perfect sense.  Someone (can't find it 
now) wrote that the goal of AT applications is to "read the contents of 
the main article, without comments or other distractions" - the fact 
that a "comment"  is nested inside another automatically 
implies that it is *supportive*, and not  an integral part of the main 
article itself (it can be excluded without negative impact).




--
Adrian Testa-Avila
adr...@custom-anything.com 
follow on facebook