RE: Bulk use of [free] accessibility checkers (don't)

If the folks at Bobby are concerned about abuse of their free services, I
recommend they implement a way to limit the number of these free uses by IP
or site.  If you set a limit of 25 pages per ip/site, you would pretty much
curtail any abuse.  Beyond the 25, the users should know whether or not
Bobby meets their needs.

Just my .02

Jim Fitzgerald

-----Original Message-----
From: Brian Kelly [mailto:b.kelly@ukoln.ac.uk]
Sent: Friday, August 03, 2001 1:24 PM
To: 'Al Gilman'; w3c-wai-ig@w3.org
Cc: chitchcock@cast.org
Subject: RE: Bulk use of [free] accessibility checkers (don't)



> At 08:15 AM 2001-08-03 , Brian Kelly wrote:
> >I am currently carrying out a benchmarking survey of a 
> selection of Web 
> >sites. The survey will look at measures such as the size of entry 
> >points,  quality of HTML, WAI compliance, etc.  I will be using 
> >Web-based tools, so that the method can be seen by all.  
> I'll also be 
> >using freely-available Web services, so anyone can try the 
> survey for 
> >themselves.
> 
> There is a question of business justice, here.  It is not 
> fair to rely on the free Web-interactive services such as 
> Bobby at CAST for survey purposes.  
> 
> Bulk work such as surveys should be performed with resources 
> provided by the activity doing the survey.  This is partly 
> where our clunky 'robot control' system came from.  There 
> have been problems with U.S. activities bringing the free 
> over-the-Web Bobby service to its knees in a de_facto 
> denial-of-service attack by putting production loads on the 
> web-published service.  Use the free service for samples, 
> please.  Not for surveys or other production runs.
> 
> Those who wish to survey should reasonably expect to pay.  
> There is no excuse for assuming that people will render you 
> great gobs of service for free.  That which is appropriately 
> _de rien_ is in the eye of the service provider, not receiver.

Interesting point.

At the last W3C AC meeting I suggested that W3C should flag their
various services (e.g. HTML and CSS validation, XSLT transformation
services, etc.) as to whether they were intended from production use or
as pilots.  I don't think that's been addressed yet.

I don't think you can say that just because something is for free, it
shouldn't be used intensively - the open source lobby won't like you for
saying that!

Also the services are funded somewhere - and it may be that the funding
is to provide a global service.  This is the case for W3C's validation
services.

Other free services may encourage survey use, if it increases traffic to
their site, and therefore advertising revenue.

So I would say it was for the service provider to define their usage
conditions.

Note we should propably distinugush between automated use by programs,
and embedded URLs which are initiated by humans.  I was referring to the
latter use - e.g. see <http://www.ariadne.ac.uk/issue28/web-watch/>.

Brian
---------------------------------------
Brian Kelly
UK Web Focus
UKOLN
University of Bath 
BATH
BA2 7AY
Email: B.Kelly@ukoln.ac.uk
Web: http://www.ukoln.ac.uk/
Phone: 01225 323943 
> Al
> 
> 
>

Received on Friday, 3 August 2001 12:40:23 UTC