[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re2: Please cast your votes for Sorcerer's Apprentice



Christopher Fry asks:

>Sorcerer's Apprentice sounds pretty damn good, providing your claim of
>reliability is accurate.
>
>Please clarify:
>
>- You say it uses MCL CLOS but has been under development for 3 years. This
>is before MCL CLOS was generally available. Did you use PCL, some non-Mac
>machine, or non-clos for the earlier phase of development?
We used MACL 1.3.2 + PCL initially (and were very pleased by the performance
increase with MCL 2.0 and native CLOS -- as well as not having to "warm up
the method caches" before saving apps :-)

>- Does the Oracle DB interface use DAL? If so, how did you deal with the
>255 character limitation of records?
We didn't use DAL, but Oracle's OCI (Oracle C Interface) library.  Had to
fuss some to get long columns working, but they are.  We always anticipated
that a port to DAL or other, RDBMS vendor specific interfaces would be
straightforward.  Has this problem with DAL been fixed yet?

>- What's the documentation like?  One of the big issues here is "how easy
>is it going to be for someone to use it?" If you write or at least help
>design code yourself, you're in a very good position to be an ace user of
>it. This is one of the things "Growing our own" has going for it. But of
>course only a small percent of the potential user base will have much to do
>with a home-grown App Framework, so this issue boils down to modularity,
>general elegance of the modules and ... documentation.
Documentation?  What documentation?  Seriously though, what SA has in this
area is:
1) Good, clean architecture and code.  So reading the code isn't all that bad.
2) Some video taped chalk talks.
3) Class, method and slot documentation browser (based on the metainfo stuff
in CLM/GINA -- we needed it while still using PCL and much before MCL 2.0's
introspective MOP functions were available.  Could be moved to the MOP quite
easily.
4) Test examples of usage at the end of most code modules.
5) The sample applications, if available with SA, would be very useful.

>I can imagine that SA could at least use a good tutorial. Writing this may
>be no easy task, but is something that should be done for a home-grown App
>Framework too. If we're faced with the choice of write code + doc or just
>write doc for some already good code, the choice is easy. A good tutorial
>can certainly be a collaborative project involving code writing for
>examples, testing them, english editing, and testing that on actual users.
>If SA is really up to snuff, we should be able to write a tutorial and
>reference manual in SA itself and distribute that as both an example and
>documentation.
Yes.  What SA needs is an overview doc, a tutorial and a style guide.

>- If we really get SA source, we can, of course, make improvements. Sounds
>like that's a big if so any details about it would help making this
>proposal more realistic.
I don't think SA will be useful, UNLESS it is distributed in source form.
I also don't think it'll be any harder to obtain source rights to it than
just binary, but that's just a guess and a matter for private discussions
between Apple (or whomever) and PI.

Thanks for your interest in SA.

Mike Wirth