[OpenAFS-devel] History of AFS

Derrick J Brashear shadow@dementia.org
Tue, 30 Jan 2001 21:11:59 -0500


--On Wednesday, January 31, 2001 08:50:05 AM +1000 Kalvinder Singh 
<ksingh@oz.agile.tv> wrote:

> I noticed nobody replied to this question, and looking back at it, I
> guess I was very vague...

The only people who "really know" would be people who'd been with Transarc 
since the beginning, and I don't know if any of them are (still) reading.

> The reason I ask is I know someone who has had experience with the
> original CMU code (he told me it was very unmaintainable), anyway, I was
> hoping that someone here could tell me if the current code is similar to
> the CMU code, or since it was very hard to maintain most of it (or some
> of it) was rewritten...

The CMU-generation code was AFS version 2; This is AFS version 3. AFS2 used 
a different RPC system than the current one, and lacked several of the 
subsystems present in the current codebase. So, it's a good bet a lot of 
the code is new since the CMU generation, and I'd guess that almost all of 
it is. The transition from a research project to a commercial product means 
you have different concerns, and my understanding is a lot of work was done 
to make that transition.

Despite the fact that I work at CMU, and despite my office being in space 
formerly occupied by that project, I don't actually know the answers to 
your questions; AFS was a product before I got to CMU.

-D