- From: Hugh Glaser <hg@ecs.soton.ac.uk>
- Date: Tue, 16 Jun 2009 23:40:11 +0100
- To: Andreas Harth <andreas@harth.org>, Dan Brickley <danbri@danbri.org>
- CC: Brian Manley <brian.manley@gmail.com>, "semantic-web@w3.org" <semantic-web@w3.org>, "stephane.corlosquet@deri.org" <stephane.corlosquet@deri.org>
Great idea. On 16/06/2009 20:23, "Andreas Harth" <andreas@harth.org> wrote: > Hi, > > On Tue, Jun 16, 2009 at 08:19:37PM +0200, Dan Brickley wrote: >> On 16/6/09 19:57, Brian Manley wrote: >>> Cool idea. But "Hall of Shame?" :/ >> >> Yup & yup. It's important not to embarrass new SW adopters. > > we wanted to have something catchy, but I understand the tagline > could be misinterpreted by innocent bystanders. Yes. If you are publishing GBs or even TBs of data, then it is expected that there will be errors. We don't want that to discourage people. I think it may be useful to have some more categories. Many of the reports seem to be type error, but there are also data (instance?) ones. So http://dbpedia.org/page/President_of_the_United_States report is an example. Now it might have been a mistake by the dbpedia people in the acquisition code, a mistake in their data, a delay in harvesting, or a delay in the source data. Or an error in the source data (as with the Afriston/Alfriston problem in unlocode). Encouraging people to help data publishers by encouraging them to find out a bit more about why the problem arose would be really good. I think the really important category is whether it is a source data or SW publisher data problem. By the way, seeing as most of the descriptions are single lines, it would be great to see the first line or so in the summary, so I don't have to keep clicking. Best Hugh > > We'll discuss alternatives while we try to fix the OpenID bug. > If you have other suggestions for a tagline please let us know. > > Cheers, > Andreas. > >
Received on Tuesday, 16 June 2009 22:41:01 UTC