On Lee | 3 Oct 2009 20:25
Picon
Favicon

[foaf-dev] foaf extension

http://danbri.livejournal.com/data/foaf shows the following:

1.      <foaf:member_name>anjeve</foaf:member_name>

2.      <foaf:image>http://l-userpic.livejournal.com/54077021/20017</foaf:image>

3.      <foaf:tagLine>brad's life</foaf:tagLine>

 

But I couldn’t find them on http://xmlns.com/foaf/spec/ (http://xmlns.com/foaf/0.1/)

 

Are those outdated or extension properties?

Also, what are the guidelines/policies for extending foaf (if there is any)?

 

Best,

-- On Lee

 

_______________________________________________
foaf-dev mailing list
foaf-dev@...
http://lists.foaf-project.org/mailman/listinfo/foaf-dev
Stephane Corlosquet | 11 Oct 2009 17:09
Picon
Gravatar

[foaf-dev] Drupal user profiles in RDFa

Hi all,

I'm checking the RDFa markup we have chosen for Drupal 7 and want to make sure it's optimal, given that it won't be able to change it after the code freeze (which in one week!). In particular I'd like to discuss the case of the user markup which is used to describe the author or a page or comment. "User profile" in this email refers to the concept of sioc:User / foaf:OnlineAccount. Drupal has 3 main cases which all have different markup.


1. If the administrator of the site decided to make the user profile pages public, then the author will be a link to the author profile page http://example.org/user/23. The original markup is:

<a href="/user/23" title="View user profile." class="username">John</a>

We have 2 options to add RDFa to this link which both return the same RDF data:

<a href="/user/1" rel="sioc:has_creator foaf:maker" title="View user profile." class="username"><span resource="/user/1" typeof="sioc:User" property="foaf:name">Henry</span></a>

or

<span rel="sioc:has_creator foaf:maker"><a href="/d7sprint/user/1" typeof="sioc:User" property="foaf:name" title="View user profile." class="username">Henry</a></span></span>

Question for the RDFa ninjas reading this: is there a way to embed all this information without adding a span tag either inside or outside the a existing a tag. I'm asking this because adding extra markup is frown upon by Drupal themers as it might break some CSS rules or have unintended effects, so if we can do without it it's best. To understand the risk of adding an extra span, take a long at the second option above: if there is a CSS rule on "a span" path targetting other markup on the site, then name of the author will suddenly be affected by this rule, while it would not in the original non-RDFa markup. Unfortunately there is no "blank" tag which maybe would have been helpful here by not interacting with CSS. Such "blank" HTML tag would have been useful in this case to wrap RDFa markup and ensure it is not targeted by any CSS rule (this could be a convention at least).


2. If the user profile pages are not accessible then only the name of the author will be displayed and we are missing the URL of her profile page, or rather I should say we can generate a URI for the user profile, but resolving it would lead to a 403 Access denied. The default Drupal markup in this case is simply:
<span class="username">John</span>

I see 3 alternatives to annotate this in RDFa:

  2.1 Use a markup similar to the one above, but the cons is that the user profile URI will not be dereferencable (and hereby breaking one of the Linked Data principles). It will return a 403 Access denied.

<span rel="sioc:has_creator foaf:maker" class="username"><span resource="/user/1" typeof="sioc:User" property="foaf:name">Henry</span></span>

  2.2 Same as 2.1 but without the resource attribute hence generating a bnode. I always try hard to avoid generating these, but if this is no work around, then we'll have to use it.

<span rel="sioc:has_creator foaf:maker" class="username"><span resource="/user/1" typeof="sioc:User" property="foaf:name">Henry</span></span>

  2.3 We don't introduce the concept of user profile and simply link the page to a name (literal).

<span class="username" property="foaf:maker">Henry</span>

I don't like 2.3 because it's not as meaningful from an RDF perspective. Also, we don't have the same choices in the RDF properties we can use due to the domain which is now a literal, which would be a problem in the context of Drupal since the page -> author relationship gets the same set of RDF properties no matter what the case is (between 1, 2 or 3). So I'd rather use 2.1 or 2.2. but we've got to decide between the best of the two: a non dereferencable URI or a bnode? I like the non referencable URI because we can uniquely identify the author outside the scope of the page, but I wonder how the parsers will react when trying to resolve the URI.


3. In the case of a non registered user leaving a comment, Drupal offers to leave her name, homepage and email address (though the email address is not displayed for privacy reasons). The default markup is:

<a href="http://openspring.net/" rel="nofollow" class="username">Stephane Corlosquet (not verified)</a>

We don't have a user profile URI here, but a homepage which is usually linked to a foaf:Person. Here we have multiple options again to describe the author of a comment. I'm not sure we should directly link a page to a foaf:Person, should we?. Do we have to generate a foaf:OnlineAccount /sioc:User URI here based on the homepage by adding #user to it? use a bnode? opinions?

Stephane.

_______________________________________________
foaf-dev mailing list
foaf-dev@...
http://lists.foaf-project.org/mailman/listinfo/foaf-dev
Nicholas Shanks | 12 Oct 2009 12:06

Re: [foaf-dev] Drupal user profiles in RDFa

On 11 Oct 2009, at 16:09, Stephane Corlosquet wrote:

> Hi all,
>
> I'm checking the RDFa markup we have chosen for Drupal 7 and want to  
> make
> sure it's optimal, given that it won't be able to change it after  
> the code
> freeze (which in one week!). In particular I'd like to discuss the  
> case of
> the user markup which is used to describe the author or a page or  
> comment.
> "User profile" in this email refers to the concept of sioc:User /
> foaf:OnlineAccount. Drupal has 3 main cases which all have different  
> markup.

I am not keen on redundancy. I would rather let the Web of Data handle  
things like saying "a user is someone with an online account".
Also, I believe OnlineAccount is going to be renamed to Account soon  
(with the old term relegated to an appendix and made same as or  
subclass of the new one).

> 1. If the administrator of the site decided to make the user profile  
> pages
> public, then the author will be a link to the author profile page
> http://example.org/user/23. The original markup is:
>
> <a href="/user/23" title="View user profile." class="username">John</ 
> a>
>
> We have 2 options to add RDFa to this link which both return the  
> same RDF
> data:
>
> <a href="/user/1" rel="sioc:has_creator foaf:maker" title="View user
> profile." class="username"><span resource="/user/1" typeof="sioc:User"
> property="foaf:name">Henry</span></a>
>
> or
>
> <span rel="sioc:has_creator foaf:maker"><a href="/d7sprint/user/1"
> typeof="sioc:User" property="foaf:name" title="View user profile."
> class="username">Henry</a></span></span>

again sioc and foaf ways of saying the same thing aren't really  
necessary. IMHO you should choose one primary vocabulary that best  
meets as many of the triples you want to describe as possible, and  
stick to that one wherever you have a choice. (i.e. become "a SIOC  
document" or "a FOAF document")

> if there is a CSS rule
> on "a span" path targetting other markup on the site, then name of the
> author will suddenly be affected by this rule, while it would not in  
> the
> original non-RDFa markup. Unfortunately there is no "blank" tag

Actually span and div *are* the blank tags of HTML. They should not  
really be used in CSS selectors.
I would contend that a selector of the form "a span" is already broken  
if it was intending to style something more specific than a bunch of  
semi-random inline content.
A more appropriate user selector might be "article a > .hostname" or  
whatever the use case was.
As such, I would not worry about it, and perhaps write an op-ed  
article about how not to abuse CSS when creating themes!

> 3. In the case of a non registered user leaving a comment, Drupal  
> offers to
> leave her name, homepage and email address (though the email address  
> is not
> displayed for privacy reasons). The default markup is:
>
> <a href="http://openspring.net/" rel="nofollow"  
> class="username">Stephane
> Corlosquet (not verified)</a>
>
> We don't have a user profile URI here, but a homepage which is usually
> linked to a foaf:Person. Here we have multiple options again to  
> describe the
> author of a comment. I'm not sure we should directly link a page to a
> foaf:Person, should we?. Do we have to generate a foaf:OnlineAccount
> /sioc:User URI here based on the homepage by adding #user to it? use a
> bnode? opinions?

I suggest foaf:Agent + rdfs:seeAlso

the user could be representing their company, and put their company  
URL, for example:
<a href="http://acmesprings.example/" rel="nofollow"  
class="username">Acme Springs, Limited</a>

– Nicholas
Attachment (smime.p7s): application/pkcs7-signature, 2456 bytes
_______________________________________________
foaf-dev mailing list
foaf-dev@...
http://lists.foaf-project.org/mailman/listinfo/foaf-dev
Toby Inkster | 12 Oct 2009 13:42
Picon
Gravatar

Re: [foaf-dev] Drupal user profiles in RDFa

On 11 Oct 2009, at 16:09, Stephane Corlosquet wrote:

> We have 2 options to add RDFa to this link which both return the  
> same RDF data:
>
> <a href="/user/1" rel="sioc:has_creator foaf:maker" title="View  
> user profile." class="username"><span resource="/user/1"  
> typeof="sioc:User" property="foaf:name">Henry</span></a>
>
> or
>
> <span rel="sioc:has_creator foaf:maker"><a href="/d7sprint/user/1"  
> typeof="sioc:User" property="foaf:name" title="View user profile."  
> class="username">Henry</a></span></span>

foaf:maker is probably not what you want. foaf:maker's range is  
foaf:Agent, whereas a sioc:User is not a foaf:Agent, it's an account  
that a foaf:Agent holds.

> Question for the RDFa ninjas reading this: is there a way to embed  
> all this information without adding a span tag either inside or  
> outside the a existing a tag.

<a href="/user/1" title="View user profile."
    about="/user/1" typeof="sioc:User"
    rev="sioc:has_creator" resource=""
    property="foaf:name">Henry</a>

>   2.1 Use a markup similar to the one above, but the cons is that  
> the user profile URI will not be dereferencable (and hereby  
> breaking one of the Linked Data principles). It will return a 403  
> Access denied.
>
> <span rel="sioc:has_creator foaf:maker" class="username"><span  
> resource="/user/1" typeof="sioc:User" property="foaf:name">Henry</ 
> span></span>

I think this is the right thing to do.

> 3. In the case of a non registered user leaving a comment, Drupal  
> offers to leave her name, homepage and email address (though the  
> email address is not displayed for privacy reasons). The default  
> markup is:
>
> <a href="http://openspring.net/" rel="nofollow"  
> class="username">Stephane Corlosquet (not verified)</a>
>
> We don't have a user profile URI here, but a homepage which is  
> usually linked to a foaf:Person. Here we have multiple options  
> again to describe the author of a comment. I'm not sure we should  
> directly link a page to a foaf:Person, should we?. Do we have to  
> generate a foaf:OnlineAccount /sioc:User URI here based on the  
> homepage by adding #user to it? use a bnode? opinions?

Assuming that you're still wanting to avoid that <span> element, the  
following would be reasonably consistent with the markup for people  
who have accounts:

<a href="http://openspring.net/"
    about="_:bnode_001" typeof="sioc:User"
    rev="sioc:has_creator" resource=""
    property="foaf:name" content="Stephane Corlosquet"
    >Stephane Corlosquet (not verified)</a>

This does *not* use the address <http://openspring.net/> in any  
triples though. It's ignored. If we include that <span>, then we can  
use the openspring.net address:

<span rel="sioc:has_creator">
   <a typeof="sioc:User"
      rel="foaf:page" href="http://openspring.net/"
      property="foaf:name" content="Stephane Corlosquet"
      >Stephane Corlosquet (not verified)</a>
</span>

(Note that I've used foaf:page instead of foaf:homepage because the  
latter is a very strong assertion, being an IFP. Stick with foaf:page  
for user-submitted content, otherwise OWL inference on your data will  
start to throw up some very strange conclusions.)

--

-- 
Toby A Inkster
<mailto:mail@...>
<http://tobyinkster.co.uk>
Toby Inkster | 12 Oct 2009 13:51
Picon
Gravatar

Re: [foaf-dev] Drupal user profiles in RDFa

On 12 Oct 2009, at 12:42, Toby Inkster wrote:

>>   2.1 Use a markup similar to the one above, but the cons is that  
>> the user profile URI will not be dereferencable (and hereby  
>> breaking one of the Linked Data principles). It will return a 403  
>> Access denied.
>>
>> <span rel="sioc:has_creator foaf:maker" class="username"><span  
>> resource="/user/1" typeof="sioc:User" property="foaf:name">Henry</ 
>> span></span>
>
> I think this is the right thing to do.

You want  <at> about here though.  <at> about is for setting subjects,  
 <at> resource is for setting objects. (It's possible to use them the  
other way around, but not a good idea usually.)

My point though, was that, yes, I think it's fine to use URIs here  
which currently resolve to 403. That's an accurate representation of  
the situation (that the data exists but access is denied).

--

-- 
Toby A Inkster
<mailto:mail@...>
<http://tobyinkster.co.uk>
Stephane Corlosquet | 13 Oct 2009 21:17
Picon
Gravatar

Re: [foaf-dev] Drupal user profiles in RDFa

Hi,

Thank you all for your feedback, that's excellent! :)

On a side-note people still haven't agreed if the URIs of profile
pages can be used as the URIs of the accounts but I wouldn't worry too
much about that.

Yes, I asked the same question to Dan Brickley during DrupalCon Paris, and he said it was ok to do that.That's also simpler to grasp for people with little background in RDF.


Alternatively you could get rid of the span by just not declaring the
name [...] and/or the type on this page.
Instead you just link the maker with the appropriate property. The
type can be inferred from the range of the property and the name could
be supplied when dereferencing the profile. This is of course
problematic with the case you describe below where the profile is not
accessible.

That was also my first take on this, but I changed my mind as some plugins or UI tool can take advantage of the typeof values. There exist plugins which target specific attributes on RDFa pages and format them depending on their value, e.g. sioc:User of a page in a special form (with a logo, popup or whatever - I know Michael Hausenblas or Alexandre Passant have an example for this). Relying on reasoning and dereferencing is too much to expect and would slow down such interface.


I'd like to get more opinions on the 3 third case where anonymous visitor can leave comments. I agree that using foaf:homepage here is inappropriate due to the IFP, and since we do not control what URI people will put in there we should be careful. Options suggested so far:
1. bnode typed sioc:User or foaf:Agent to cover the case where it's a link to a company homepage.
2. link this bnode to the "homepage" URI with foaf:isPrimaryTopicOf, foaf:page or rdfs:seeAlso.

foaf:isPrimaryTopicOf is also an IFP, so I think we should avoid it. rdfs:seeAlso is good though somewhat generic. foaf:page seems the most appropriate here and does not imply strong assumptions. What do you think?

Stephane.

On Mon, Oct 12, 2009 at 7:48 AM, Steven Pemberton <Steven.Pemberton-rh8NL+sEX9E@public.gmane.org> wrote:
On Sun, 11 Oct 2009 17:09:52 +0200, Stephane Corlosquet <scorlosquet <at> gmail.com> wrote:

Hi all,

I'm checking the RDFa markup we have chosen for Drupal 7 and want to make
sure it's optimal, given that it won't be able to change it after the code freeze (which in one week!). In particular I'd like to discuss the case of the user markup which is used to describe the author or a page or comment.

Thanks for the chance to feedback!


"User profile" in this email refers to the concept of sioc:User /
foaf:OnlineAccount. Drupal has 3 main cases which all have different markup.


1. If the administrator of the site decided to make the user profile pages public, then the author will be a link to the author profile page
http://example.org/user/23. The original markup is:

<a href="/user/23" title="View user profile." class="username">John</a>

We have 2 options to add RDFa to this link which both return the same RDF
data:

<a href="/user/1" rel="sioc:has_creator foaf:maker" title="View user
profile." class="username"><span resource="/user/1" typeof="sioc:User"
property="foaf:name">Henry</span></a>

or

<span rel="sioc:has_creator foaf:maker"><a href="/d7sprint/user/1"
typeof="sioc:User" property="foaf:name" title="View user profile."
class="username">Henry</a></span></span>

I'm assuming the /dsprint7 is a typo here, otherwise they wouldn't have the same triples.

So the triples they produce are like:

 <http://www.w3.org/basedoc.html> <http://sioc.org/has_creator> <http://www.w3.org/user/1> .
 <http://www.w3.org/basedoc.html> <http://foaf.org/maker> <http://www.w3.org/user/1> .
 <http://www.w3.org/user/1> <http://foaf.org/name> "Henry" .
 <http://www.w3.org/user/1> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://sioc.org/User> .

Of the two I slightly prefer the second, with the outer <span>, since the URI doesn't have to get repeated.

If it weren't for the <at> typeof, you could do it all in one go:

<a href="/user/3" rel="sioc:has_creator foaf:maker" title="View user
profile." class="username" property="foaf:name">Steven</a>

but of course the <at> typeof changes the subject.


Question for the RDFa ninjas reading this: is there a way to embed all this
information without adding a span tag either inside or outside the a
existing a tag. I'm asking this because adding extra markup is frown upon by
Drupal themers as it might break some CSS rules or have unintended effects,
so if we can do without it it's best.

At the cost of repeating the URI, but with the advantage that it doesn't risk messing up your CSS, you don't have to nest the extra span:

<a href="/user/1" rel="sioc:has_creator foaf:maker" title="View user profile." class="username" property="foaf:name">Henry</a>
<span about="/user/1" typeof="sioc:User"></span>

So this is a non-nested version of your first solution.


2. If the user profile pages are not accessible then only the name of the
author will be displayed and we are missing the URL of her profile page, or rather I should say we can generate a URI for the user profile, but
resolving it would lead to a 403 Access denied. The default Drupal markup in
this case is simply:
<span class="username">John</span>

I see 3 alternatives to annotate this in RDFa:

 2.1 Use a markup similar to the one above, but the cons is that the user
profile URI will not be dereferencable (and hereby breaking one of the
Linked Data principles). It will return a 403 Access denied.

<span rel="sioc:has_creator foaf:maker" class="username"><span
resource="/user/1" typeof="sioc:User"
property="foaf:name">Henry</span></span>

 2.2 Same as 2.1 but without the resource attribute hence generating a
bnode. I always try hard to avoid generating these, but if this is no work
around, then we'll have to use it.

<span rel="sioc:has_creator foaf:maker" class="username"><span
resource="/user/1" typeof="sioc:User"
property="foaf:name">Henry</span></span>

 2.3 We don't introduce the concept of user profile and simply link the
page to a name (literal).

<span class="username" property="foaf:maker">Henry</span>

I don't like 2.3 because it's not as meaningful from an RDF perspective.
Also, we don't have the same choices in the RDF properties we can use due to
the domain which is now a literal, which would be a problem in the context
of Drupal since the page -> author relationship gets the same set of RDF
properties no matter what the case is (between 1, 2 or 3). So I'd rather use
2.1 or 2.2. but we've got to decide between the best of the two: a non
dereferencable URI or a bnode? I like the non referencable URI because we
can uniquely identify the author outside the scope of the page, but I wonder
how the parsers will react when trying to resolve the URI.

I think you're worrying too much about the 403s. The triples are true, whether or not you try to dereference the URIs:

 <http://www.w3.org/basedoc.html> <http://sioc.org/has_creator> <http://www.w3.org/user/1> .
 <http://www.w3.org/basedoc.html> <http://foaf.org/maker> <http://www.w3.org/user/1> .
 <http://www.w3.org/user/1> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://sioc.org/User> .
 <http://www.w3.org/user/1> <http://foaf.org/name> "Henry" .

What you may want to avoid is making them *easy* to dereference. Then I would use a span+ <at> resource instead of an a:

<span resource="/user/1" rel="sioc:has_creator foaf:maker" title="Inaccessible user profile" class="username" property="foaf:name">Henry</a>
<span about="/user/1" typeof="sioc:User"></span>


3. In the case of a non registered user leaving a comment, Drupal offers to leave her name, homepage and email address (though the email address is not displayed for privacy reasons). The default markup is:

<a href="http://openspring.net/" rel="nofollow" class="username">Stephane
Corlosquet (not verified)</a>

We don't have a user profile URI here, but a homepage which is usually
linked to a foaf:Person. Here we have multiple options again to describe the author of a comment. I'm not sure we should directly link a page to a
foaf:Person, should we?

No.


Do we have to generate a foaf:OnlineAccount
/sioc:User URI here based on the homepage by adding #user to it? use a
bnode? opinions?

In the absence of other information I prefer to use rel="foaf:isPrimaryTopicOf" to identify a person from their homepage.
(See http://www.w3.org/MarkUp/2009/rdfa-for-html-authors)

Hope this helps.

Best wishes,

Steven Pemberton


Stephane.


_______________________________________________
foaf-dev mailing list
foaf-dev@...
http://lists.foaf-project.org/mailman/listinfo/foaf-dev
Julie Letierce | 15 Oct 2009 19:01

[foaf-dev] survey: How Semantic Web researchers use Web 2.0 to communicate about their work?

(Apologises for cross-posting. Please forward to interested colleagues and students)

Dear All, 

In order to study the habits and motivations of the Semantic Web community to publish and share contents online using Web2.0 services, I have launched a short online survey "How Semantic Web researchers use Web 2.0 to communicate about their work?"

If you are researching Semantic Web technologies, I would really appreciate if you can take 10 minutes to complete this questionnaire. And If you have already completed the survey, thank you very much! 

The survey is online at http://tinyurl.com/semweb-survey

If you have any questions in relation to the survey, feel free to contact me.


Thanks a lot for your help.
Best Regards.

-- 
Julie Letierce

_______________________________________________
foaf-dev mailing list
foaf-dev@...
http://lists.foaf-project.org/mailman/listinfo/foaf-dev
Story Henry | 17 Oct 2009 13:46

[foaf-dev] Social Web Bar Camp in Santa Clara Monday November 2nd

Please forward to interested parties, tweet, blog!

There will be a Social Web Camp in Sun Offices in Santa Clara on  
Monday November 2. It's is being hosted by SUN Microsystems and  
organized by Henry Story of SUN Microsystems and Daniel Appelquist of  
Vodafone, co-chair of the W3C Social Web XG.

Register at: http://barcamp.org/SocialWebCamp-Santa-Clara

When: Monday, 2nd of November 9:00 am to 5:00 pm

Where:
	The Auditorium at Sun's Campus,
	Santa Clara,
	95054 California

All topics related to the Social Web will be covered, in a bar-camp
style. It will be broad and open to all types of participants.

There are quite a number of complimentary activities that week in the  
Bay Area, for those looking for other reasons to come:

- We recommend the Internet Identity Workshop [1] taking place the  
rest of the week,
- The W3C is running it's Technical Plenary [2], where the Social Web  
XG is  meeting on Tuesday
- ApacheCon is also taking place in the Bay Area [3], and could be of  
interest

[Signature]

[1] http://www.internetidentityworkshop.com/
[2] http://www.w3.org/2009/11/TPAC/
[3] http://www.us.apachecon.com/c/acus2009/

Social Web Architect
Sun Microsystems		
Blog: http://blogs.sun.com/bblfish
Dan Brickley | 18 Oct 2009 12:32

[foaf-dev] SocialWeb anti-pattern wiki? today's annoyance: name too long/short/weird to fit

Greenpeace's site just refused to let me sign up, because my first
name - Dan - is too short. See screenshot at
http://www.flickr.com/photos/danbri/4022025450/ ...

OK so my passport and family use 'Daniel' but I've been using 'Dan'
for over twenty years. Who are they to say that's not my name?

I know plenty of people who go by 'Jo' as a first (given) name, which
is a mere two letters. Or a Jan (like my mum); or a Ted (cc:'d). Sure
these can be considered contractions of some original longer form, but
I'm sure many people have < 5 chars (or < 3 chars even) on their birth
certificates. Not that a Web site has any reason to demand to know
what's on your birth certificate!

See also http://www.slideshare.net/danbri/danbri-foaf-talk-social-web-camp-www2009
for a couple of examples from the ill-fitting Charles McCathieNevile:

Charles tried to sign up to Facebook ('Facebook helps you connect and
share with the people in your life - Sign up, it's free any anyone can
join"), only to be told:  “Name contains too many capital laters.”

So, off to Twitter.com instead, to "Join the conversation", as they
say. But sorry: “Name is too long (max is 20 characters)”.

This is what happens when you put computer programmers (many of whom I
count as dear friends) in a role where they get to control (often
casually and without a lot of care) what end users can say about
themselves. See http://geeks-bearing-gifts.com/gbgContents.html for a
lot more detail on this kind of mess.

What can we do to help programmers who are put in this role to stop
and think; to remember when they're building so-called 'social Web'
sites that not everyone understands 'first name' as meaning the same
thing, and that the rule "all first names must have 5 < x < 20 chars"
makes no more sense in Croydon than in China.

Articles like http://adactio.com/journal/1357/ ("the password
anti-pattern") and
http://factoryjoe.com/blog/2009/01/02/twitter-and-the-password-anti-pattern/
("Twitter and the Password Anti-Pattern") have been quite effective in
getting more software engineers to understand why certain kinds of
technical lazyness is disrespectful of the poor end users. In this
case, the habit of asking users for passwords for another site teaches
users to behave in ways that may leak their passwords.

So I wonder whether we (the w3c socialweb incubator group, the foaf
list, or anyone interested) might usefully start collecting up a list
of anti-patterns (probably in a wiki), so that busy overworked coders
can be helped to avoid some ugly little mistakes which exclude users
or force them into narrow ill-thought-out schemas...

cheers,

Dan
Norman Gray | 18 Oct 2009 14:21
Picon
Picon
Gravatar

Re: [foaf-dev] SocialWeb anti-pattern wiki? today's annoyance: name too long/short/weird to fit


Greetings.

On 2009 Oct 18, at 11:32, Dan Brickley wrote:

> not everyone understands 'first name' as meaning the same
> thing, and that the rule "all first names must have 5 < x < 20 chars"
> makes no more sense in Croydon than in China.

And just in case anyone is rash enough to think that that's all there  
is to i18n, there's an entertaining discussion of names at
<http://blog.jclark.com/2007/12/thai-personal-names.html 
 >

Norman

--

-- 
Norman Gray  :  http://nxg.me.uk

Gmane