🚀 go-pugleaf

In article <UYBGaHy00WB50oe7cW@andrew.cmu.edu> rs5o+@andrew.cmu.edu
(Randall Knowles Smith) writes:
>Comments on 4th Dimension:
>The manual looks very good, and is quite authoritatively wrong or
>misleading in some parts.  Not many, just enough to drive a beginner
>mad.  Experts in Database programming probably wouldn't have any
>problems.  Also, the program can be very slow at times; especially when
>running in a multi-user environment off a server.  And this leads me into:

That's not all. Often the manual is right and the program's wrong. Like
when it crashes. Which it does frequently, especially in multi-user.
Also, you're wrong. Even experts can go nuts trying to piece together
the crazily implemented input loop.

>Questions:
>This faster version of 4th Dim--Any due date at all?  Is it DEFINATELY
>coming out?   I'm somewhat desparate, because, you see, I'm a college
>student, just wrote my first 4th Dim. application, and it's SLOW.

It's definitely coming out. It will definitely be faster. It will
definitely almost certainly probably I think maybe hopefully might just
barely make it out before 1990. :-)

Actually, I don't know what the holdup is. It should have been out a while
ago, but maybe they decided to debug this version.

>It's a grading program, and the TA's using it (6 of them) are already
>complaining.  Next semester we'll have 10 times as many TA's and
>students.  My name is known, I don't want to be lynched.  So, I need
>help speeding the thing up.  A new version of 4th dim. would be nice,
>but barring that, can anything be done?  The major bottleneck at the
>moment is disk access.  Ideas on how others have solved this problem
>(loading everything into memory, etc) would be greatly appreciated


Well. You can get a faster disk. The best ones are the full-height CDC
(now known as Imprimis) Wrens. Use accelerators, if you can.

The only real answer, though, is the upgrade I took. I scrapped 4D in
the middle of a big project, wrote off a month of my time, and used
FoxBase.  That was a VERY BIG DECISION. Then I made up my entire loss in
TWO WEEKS!  That made me feel a lot better. The difference is not to be
believed.  Fox really is 6 to 200 times faster than 4D. The most amazing
thing is that this difference also applied to coding time (though not
quite as dramatically). A large project that would have taken me three
months to do in 4D got done in one month in FB. And the result was MUCH
better. Faster, bug-free, didn't crash because the database engine went
to lunch, and the user interface was lots better.

Fox's big problem was the weak (read IBMish) report generator. But I had
no major problems producing macish reports by programming. This is not
hugely desireable, though, which is why they put a most amazing report
generator into V2.0. The first Beta is finally here (Yeah!) and it's
quite stable (unlike the Alphas). It should be out in early May. And Fox
always meets their shipping deadlines.

Realize that I've been hearing "Wait for 4D 1.1... 1.5... 2.0" for longer
that FoxBase/Mac has existed!  In that time Fox has introduced a major
product and released one bug-free major upgrade. In four weeks time they
will introduce another equally significant upgrade.

Y'know, I almost feel sorry for Guy...

---
Alexis Rosen
alexis@ccnysci.{uucp,bitnet}