home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Visual Basic Source Code
/
Visual Basic Source Code.iso
/
vbsource
/
scrtes
/
readme
< prev
next >
Wrap
Text File
|
1993-07-25
|
4KB
|
102 lines
COMMENTS ON SCRTEST - GENERAL PURPOSE DEMO OF BROWSER/SCROLLER 26/July/93
============================================================== ==========
Hi All,
First the excuses :
-------------------
I'm new to V.B. so this is probably NOT the best way to do it, I really
would like suggestions on any improvements. I've just ported the code
from the way I do it in Clipper, I've had to use cheats to implement
the extra bits required to support Scroll-Thumb directed movements.
Listbox / Scroll-bar synchronization
------------------------------------
Keeping the Listbox and the Scroll-bar in synch drove me *NUTS* because
altering the Listbox-index from the scroll-change event triggered a
Listbox-click event and altering the Scroll-value from the Listbox-click
event triggered a scroll-change event. This resulted in some pretty amazing
recursion loops. I got round the problem by keeping a 'recursion-depth'
counter and ingnoring events below the surface level. There must be a better
way !
Key Searches
------------
The keypress seeks are really nice when you're in a big file and want to get
to a specific entry fast. You might want to add some extra feedback in a label
or text box so the User can see how the Composite key is building up and know
when the search fails. I keep a truncated data key in the file that has the
characters converted to upper-case.
Thumb directed reads
--------------------
I've implemented this because people seem to expect it in windows apps, I'm
not at all convinced that it serves any real purpose because the fastest way
to get what you want has to be a keyed search. What I do is have a record-number
stored in the file that reflects its relative position in the file. This can
only be maintained as a 'batch' process so it's not so good for constantly
changing files.
Obvious improvements
--------------------
It's not too efficient only loading 10 records at a time, some sort of buffer
coding could enable more to be loaded at a time but the code would be more
complicated.
Questions
---------
Can I index a field on its contents converted to upper case ?
Can I jump directly to a relative position in an indexed file - eg. the 1,275th
record (measured from the start and INDEXED) of a 5,000 record file ?
Can I jump directly to the first or last record of an indexed file without
cheating with a SEEK ( the equivalent of a GO TOP or GO BOTTOM in dBase) ?
Are there any limitations on the range of ascii characters allowed to be in
a database field ?
The application
---------------
I have an application written in Clipper that has the UK pop charts for every
week since 1960. Access is by Group, Title and Date. I'm currently trying to
port it across to V.B. because :
1) It's a good way to learn V.B.
2) There is a market for the product ported to windows
3) I prefer working from windows these days (it's so much NICER)
Regards,
Ross Whittaker
CIS:100275,317
~~~~~~~~~~~~~~
p.s.
Can I have lots of feedback PLEASE, all criticism gratefully received
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
SCRTEST.ZIP - File contents
---------------------------
README - This file
SCRTEST.MAK - The project file
SCRTEST.FRM - The form file
AFILE.MDB - The data file (a subset of 60's pop groups if you must know)
AFILE.LDB - The data control file (?)
The test file data structure
----------------------------
Database = AFILE
Table = ATABLE
Fields = AFIELD X(30) (main data field)
ARECNO X(5) (relative position)
AKEY X(10) (upper(mid(afield,1,10)))
Indexes = AINDEX1 = ARECNO UNIQUE PRIMARY
AINDEX2 = AKEY DUPS-OK NOT-PRIMARY
==================== end of file =========================