RE: SIMILE document hosting

> The default HTML exporter in Word is poor. Microsoft have a 
> better one here 
> http://office.microsoft.com/downloads/2000/Msohtmf2.aspx
> 
> Of 
> course, diffs won't work if we are using 
> different exporters.

Based on your suggestion I will obtain and use the better exporter.

> I think this approach to versioning is ugly. CVS would be 
> much better. We could use ViewCvs to overcome the problem 
> that people without CVS access can't see different versions.

I suspect that diff on word source will not be very helpful, so we would need to check in both .doc (source) and .html (derived) into cvs.  Diffs on .html are I suspect what people will want.

Let's chime in here with hosting requirements.  I'll start

For CVS:
- globally accessible (i.e. cvs server must reside outside corporate firewalls)
- anonymous read access?
- controlled write access?
	- how secure
		- is simple username/password OK?  is MIT's kerberized CVS ok?
	- setup constraints
		- MIT accounts ok

For ViewCVS:
- how is this deployed?
		

POSSIBLE ALTERNATIVES: (please suggest more)

- MIT Athena
	- secure CVS checkins available, via kerberos and MIT certificates
	  (requires MIT account for each writing user)
	- anonymous updates available
	  (I think, although these may be hosted on a different cvs server and so
	  may require some nightly mirroring process).

- others?


> -----Original Message-----
> From: Butler, Mark 
> Sent: Wednesday, April 02, 2003 7:40 AM
> To: Bass, Mick; 'em@w3.org'
> Cc: 'www-rdf-dspace@w3.org'
> Subject: RE: SIMILE document hosting
> 
> 
> Hi Mick
> 
> > 4. Many of the source documents are Microsoft Word, with
> > variants saved as html, and then published via 
> > http://web.mit.edu/simile.
> 
> The default HTML exporter in Word is poor. Microsoft have a 
> better one here 
> http://office.microsoft.com/downloads/2000/Msohtmf2.aspx
> 
> Of 
> course, diffs won't work if we are using 
> different exporters.
> 
> > 6. Multiple versions of documents are published to the web.
> > Currently this occurs using conventions in the document name 
> > (i.e. "somedoc 0.22.html", "somedoc 0.24.html").  I would 
> > suggest that we continue this practice even if version info 
> > were also maintained via, say, cvs.  Doing so would allow 
> > those without access to the cvs repository to still access 
> > the various versions of a document.
> 
> I think this approach to versioning is ugly. CVS would be 
> much better. We could use ViewCvs to overcome the problem 
> that people without CVS access can't see different versions.
> 
> Dr Mark H. Butler
> Research Scientist                HP Labs Bristol
> mark-h_butler@hp.com
> Internet: http://www-uk.hpl.hp.com/people/marbut/
> 
>

Received on Wednesday, 2 April 2003 08:03:14 UTC