[Pipet Devel] Choice of ORB implementations

J.W. Bizzaro bizzaro at geoserve.net
Wed Apr 5 11:14:24 EDT 2000


Brad Chapman wrote:
> 
> I'm confused about how you want to make use of the unix permission system
> in a distributed system.
> I guess you mean that we should be writing our own specific type directory
> service. I'm opposed to this (at the present time) for two reasons:

Oh, I wasn't even considering directory services when I mentioned that.  It
was just a simple suggestion to use the permissions, etc. that are already set
in the Unix filesystem.

But of course, we don't even agree on how nodes should be stored and used.  My
suggestion of making use of the Unix filesystem stems from my idea that the
system should work from within that permanent storage medium.  Your (and
Jarl's) approach is to work from an ephemeral, CORBA-based medium.

I'm not saying 'don't use CORBA'.  I'm saying that the storage medium should
always reflect what is happening in the working medium (CORBA).  This is first
of all important in case there is a catastrophic failure.  And I think the
working medium can glean information from the storage medium, such as who owns
a file and who can use it.

> 2. I would like to stick with open standards. It seems like we have went
> out of our way to do this thus far by using XML and CORBA. If we start
> using our own "proprietary" setup, then we jepordize the interoperability
> of our system.

I'm not suddenly against open standards.  The Unix filesystem is in fact an
older standard than most others :-)

Cheers.
Jeff
-- 
                      +----------------------------------+
                      |           J.W. Bizzaro           |
                      |                                  |
                      | http://bioinformatics.org/~jeff/ |
                      |                                  |
                      |        BIOINFORMATICS.ORG        |
                      |           The Open Lab           |
                      |                                  |
                      |    http://bioinformatics.org/    |
                      +----------------------------------+




More information about the Pipet-Devel mailing list