web publishing process

[Travis Ritter]

We currently use FTP from our PCs to the Unix server, but we are 
looking at using Samba.  I have a test scenario up and running, and 
it looks promising.  

In brief, Samba is a package that allows a Unix server to use the 
same networking protocol (SMB) as Windows NT for file and print 
sharing.  This lets you create a Windows-style share on your Unix 
server.  You can then map a network drive from your PC, so that the 
shared directory on your webserver becomes drive X: on your 
Windows machine.

In theory, this would allow people to do live editing on their pages 
using their favorite Windows HTML editor.  In practice, I would 
still recommend that development be done on the local drive and 
the finished product dragged and dropped into the web directory.

Caveat:  I have used Samba extensively to aid me in 
administrative tasks, and it has worked very well; however, this is 
the first time I will be using it to offer resources to a large number 
of users, all with different accounts.  This might add complexities 
or overhead that make Samba an undesirable solution (but I'm 
optimistic).

If you're interested, you might want to look at the Samba website:

www.samba.org.


At 10:44 AM 2/3/99 -0800, David Vose wrote:
>Our library is migrating staff computer accounts from UNIX to NT. Our
>web server will remain on the UNIX box but all web page authoring will
>be done on PCs on our staff NT server. Prior to this change, all web
>authoring was done with pico and pages were quickly and easily made
>live. Of course this process did not provide for html validation. With
>the new arrangement, staff will have to do all editing/authoring on
>their pcs and then it will somehow be sent to the unix web server. 
>
>THE QUESTION: In this split environment, what is the best way to get new
>or updated pages from the staff pcs to the unix web server. One
>suggestion is to have all authoring done on individual pcs (as opposed
>to doing it in a shared directory space similar to what can be done in
>UNIX). When an html file is finished, the author would copy the files to
>a "waiting" directory on the NT server. All files in this directory
>could be validated there (unless the authoring tool does it) and then
>ftp'd to the webserver with an automated process. This seems time
>consuming and less than ideal since there would be three copies of each
>file: one on the author's pc, one in the waiting directory, and one in
>the live space. 
>
>I'd be interested in hearing how others that work in a split environment
>like this publish their pages. 
>
>Thanks, 
> 
>David Vose
>Binghamton University Libraries
>(607) 777-4907
> 
-------------------------------------------------------
Travis Ritter                 
Network Manager                          
University of Oregon Library          
  
Email:  tritter at darkwing.uoregon.edu
Phone: (541) 346-2140
Fax:     (541) 346-3485