I settled on Repository::Simple in the end, but I'd still take
feedback on the module name. I have just uploaded it to PAUSE and
would welcome any comments and criticism. It should show up as
Repository-Simple-0.01.tar.gz shortly.
At this point, my focus has been to keep the interface as simple as
possible, probably a bit too simple. It turned out to have less in
common with File::System than I thought it would, but File::System
still provided some initial foundation.
As of this release, the system provides nothing but the ability to
read and the only engine is the FileSystem engine. I used tests to
drive the development, so it should be relatively robust.
I'm now starting to consider write operations and will probably look
at creating the Memory engine. I'm also starting to look forward to
other features, specifically version control since the implementation
of VC will probably have the greatest impact on the API---as in, I'll
probably have to introduce another couple layers to the API to make VC
practical.
I'd appreciate any feedback.
Cheers,
Sterling
On 3/28/06, Andrew Sterling Hanenkamp <[EMAIL PROTECTED]> wrote:
> > You mean they're re-discovering hierarchical databases that were
> > displaced by RDBMS? :-)
>
> Heh. Yes. They've found that the best way to store heirarchical data
> is in a heirarchy.
>
> Anyway, thanks for responding. Repository:: is not a bad idea. I
> think the work I'm currently doing would fall under the name
> Repository::Simple at this point, though it might fit in better with
> the ::Lite crowd better since the idea is pretty tightly bound up with
> XML and ::Lite seems to be preferred there. That and its got quite a
> few packages to be considered "simple."
>
> At this point, I'm working on what I hope is a one-JAPH job in a month
> or two in my spare time by rewiring File::System. It will then be a
> functional Pure Perl implementation of some of the core ideas behind
> JSR 170 Level 1, but the API will bear no particular resemblance.
> Doing this suits my own nefarious purposes best.
>
> I've done the overall design, gotten most of the tests written, and
> I've rearranged and refactored much of the code, but I've got a long
> way to go before all my tests pass.
>
> Cheers,
> Sterling
>
> On 3/21/06, A. Pagaltzis <[EMAIL PROTECTED]> wrote:
> > Hi Andrew,
> >
> > * Andrew & Terri Hanenkamp <[EMAIL PROTECTED]> [2006-02-26 11:05]:
> > >2. If not (1), what's a good name? Content::Repository?
> > >Something else?
> >
> > Something else, most definitely, simply on the grounds that
> > Content:: is a less than self-describing TLNS. I wish I had a
> > good suggestion, though. The concept is so generic that I think
> > it will be hard to name well. I am thinking that Repository::
> > should be the TLNS, but I dunno.
> >
> > >Another reason for me taking this seriously is because there are
> > >some in the industry trying to hype this one on the level of
> > >importance of the emergence of RDBMS in the 1970's.
> >
> > You mean they're re-discovering hierarchical databases that were
> > displaced by RDBMS? :-)
> >
> > Regards,
> > --
> > #Aristotle
> > *AUTOLOAD=*_;sub _{s/(.*)::(.*)/print$2,(",$\/"," ")[defined
> > wantarray]/e;$1};
> > &Just->another->Perl->hacker;
> >
>
