Microsoft Wants Patent For Inserting User IDs In Filename

Reader Comments (rss)

(Flattened / Threaded)

1. Feb 28th, 2005 @ 3:40am

   This is the same as mail-merge

   by Pete Austin

   Mail-merge into URLs is an obvious technique used in bulk emailing. For example an image tag like the following would display the logo of your favorite sports team:
   (href="http://domain/teams/{field teamname}.gif" alt="{field teamname}")
   Microsoft's patent seems to be copying this.

   (reply to this comment) (link to this comment)

2. Feb 28th, 2005 @ 5:00am

   No Subject Given

   by Loraan

   Would UNIX's tilde (home directory) count as prior art? What about environment variables in a pathname?

   (reply to this comment) (link to this comment)

3. Feb 28th, 2005 @ 10:26am

   Re: No Subject Given

   by Anonymous Coward

   Yes - countless thousands of unix shell scripts use this technique. This is nothing new.

   (reply to this comment) (link to this comment)

4. Feb 28th, 2005 @ 10:49am

   GFS

   by Anonymous Coward

   I don't know if this predates MSFT's claim, but RedHat's GFS supports variables as filenames -- see http://www.redhat.com/docs/manuals/csgfs/admin-guide/s1-manage-pathnames.html for a description of this.

   (reply to this comment) (link to this comment)

5. Feb 28th, 2005 @ 2:50pm

   Automounter

   by Mark Rasmussen

   We do this all the time w/ the automounter on unix (am-utils)... users, netgroups, all kinds of stuff.

   (reply to this comment) (link to this comment)

6. Mar 1st, 2005 @ 12:19am

   ClearCase

   by Ed

   Adding to the prior art parade...
   
   In ClearCase, one can append '@@' followed by all manner of arcane data to the end of a filename to select a specific version of the file's contents, and this has worked for at least a decade.

   (reply to this comment) (link to this comment)

7. Mar 1st, 2005 @ 12:39am

   Prior art?

   by Anonymous Coward

   OK, maybe I'm missing something here, but isn't this something we've done for decades? Microsoft says:
   

   "One problem with existing systems is that pathnames are essentially static. There is currently no mechanism for pathname components that are customizable or evaluated in the context of a user of the path. Once a pathname has been created, that same pathname will resolve to the same data regardless of which user is logged on, which machine the pathname resolution occurs on, or the like."

   
   Isn't this accomplished with users have different /home directories in Unix? Or "chroot"?

   Or how about the "user home directory" in Windows networking? Or using the environment variable %USERNAME% in the pathname? (After all, the Windows registry has many references to pathnames containing %SYSTEMROOT%, so it would work with %USERNAME% too.)

   Again, maybe I'm missing something, but this hardly sounds like something that "until now, eluded those skilled in the art."

   (reply to this comment) (link to this comment)

8. Mar 6th, 2005 @ 9:15am

   What are they doing?

   by iksobert

   It sounds like M$ is trying to pull the wool over the USPTO's eyes. We all know how little time they have to investigate each claim. So there's M$..."if we make it sound really hard and really neat, they will let us dominate."

   (reply to this comment) (link to this comment)

Add Your Comment