[uf-new] hSet survey findings and future direction

[Martin McEvoy]

On Wed, 2007-05-30 at 14:03 -0400, Manu Sporny wrote:
> Here are the current findings based on the hSet survey and input from
> the list. These findings impact how we approach/solve the grouping
> problem. Here's what we know so far:
> 
> 57% think the grouping problem should be solved.
> 14% do not believe there is a grouping problem.
> 
> Sparse and ordered grouping were the two most important grouping
> sub-problems according to the survey respondents.
> 
> 29% thought that grouping should be solved on a
>     Microformat-by-Microformat basis.
> 56% thought that grouping should be solved in a generic way.
> 
> 50% favored the option #3 solution.
> 50% favored the option #7 solution.
> 
> In general, most respondents are convinced that there is a grouping
> problem that needs to be solved in a generic way.
> 
> Here is a link to the grouping-brainstorming page:
> 
> http://microformats.org/wiki/grouping-brainstorming
> 
> It seems that the dot-notation class name approach (option #3) is
> contested because of the dots. What if we use ':' instead of dots -
> colon-notation? Names would look like this - hset:group:subgroup

> 
> What if we retracted Option #3 as a possible solution? Is there anybody
> that would strongly oppose Option #7 at that point? I can't say that I
> strongly oppose #7 as it only has a couple of small downsides compared
> with Option #3.
> 
> Is there anybody against adopting option #7 as the grouping problem
> solution?
"only 7 people that filled out the survey" wow so clearly not many
people think that grouping is a problem? but lets press on anyway ...

Your problem solution for hAudio would look like this yes? 

<div class="haudio">
<span class="collaborator vcard fn">Gogol Bordello</span>
	<span class="hset fn" id="gypsy-punks-underdog-world-strike">Gypsy
Punks: Underdog World Strike</span>
</div>
	[...]
<div class="haudio">
   <span class="hset gypsy-punks-underdog-world-strike fn">09 Start
Wearing Purple</span>
</div>
   [...]

for me that is way to much code and way to messy, notice I marked up the
full title of the group in this case an album title, l-o-n-g title yes!

I can take it away "gypsy-punks-underdog-world-strike" and the id
attribute which would leave me with this

<div class="haudio">
<span class="collaborator vcard fn">Gogol Bordello</span>
	<span class="hset fn">Gypsy Punks: Underdog World Strike</span>
</div>
	[...]
<div class="haudio">
   <span class="hset fn">09 Start Wearing Purple</span>
</div>
   [...]

you can clearly see that you don't need the bloated markup for you to
get the idea that this is a hset or grouping of sorts already, the only
thing you would need to mark up is which is the group and which is a
member,

<div class="haudio">
<span class="collaborator vcard fn">Gogol Bordello</span>
	<span class="hset collection fn">Gypsy Punks: Underdog World
Strike</span>
</div>
	[...]
<div class="haudio">
   <span class="hset item fn">09 Start Wearing Purple</span>
</div>
   [...]

this is the title of your group or collection [Gypsy Punks: Underdog
World Strike]
<span class="hset collection fn">Gypsy Punks: Underdog World
Strike</span>

this is a member or item [09 Start Wearing Purple]
<span class="hset item fn">09 Start Wearing Purple</span>

NO ID attributes
NO .: separators
Uses Simple naming conventions
Fits into the Microformat way
Uses Dublin core collection terms
http://dublincore.org/groups/collections/collection-application-profile/
The haudio objects are sparsely grouped 
And yes I have proposed this method before.

Now Manu, you may have noticed that just by looking at the name of the
item member is NOT possible to determine the Group just by looking at:

class="hset item fn"

To me I don't think it is necessary only because the declaration
class="hset collection fn" has "a special meaning" on a page and that
the contents is the grouping title, 
also a browser/phraser/whatever knows that there may (or may not) be
member items on a page the contents of which are member item titles.

Im sorry guys Ive been using Microformats since early 2006, and the
thing that makes it useful, attractive, and easy to use is the fact that
microfomats use "brief, descriptive class names" and anything beyond
this ie hset.group.subgroup or hset:group:subgroup (which can be
confused for Qnames) is not really a microfomat to me this is something
else that should be developed elsewhere. 

A question what standards are you using in your proposals?
http://microformats.org/wiki/naming-principles#For_Other_Standards.2C_Prefer_Older_to_Newer
or are you all just making them up?

the only possible use for "." breadcrumbs or ":"qnames I can see in the
very very distant future for microformats may be to pass values in a
class and keep the machine data out of the human readable area such as a
title="[...]" ie

class="updated.20070530T2243" or
class="dtstart.20070530T2243" or
class="duration.203"

but even this may be beyond the scope of microformats.


-Martin McEvoy-


> 
> -- manu
> _______________________________________________
> microformats-new mailing list
> microformats-new at microformats.org
> http://microformats.org/mailman/listinfo/microformats-new
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 2171 bytes
Desc: not available
Url : http://microformats.org/discuss/mail/microformats-new/attachments/20070530/9ab6ac94/smime.bin