home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Amiga Format CD 40
/
Amiga Format CD40 (1999-05-11)(Future Publishing)(GB)(Track 1 of 3)[!][issue 1999-06].iso
/
-seriously_amiga-
/
misc
/
megabooksr1
/
megabooksr1.readme
< prev
next >
Wrap
Text File
|
1999-03-29
|
4KB
|
68 lines
Short: The ULTIMATE Amiga Address Book (Update)
Author: fosters@dvalley.demon.co.uk (Tom Bampton)
Uploader: fosters@dvalley.demon.co.uk (Tom Bampton)
Version: 4.2 Service Release 1
Type: biz/dbase
Requires: biz/dbase/AAA_MegaBook.lha
=============================
MegaBook 4.2 SR-1 Information
=============================
Welcome to MegaBook 4.2 SR-1. This is a service release of MegaBook to fix
a couple of bugs and introduce a few changes. All the stable code from 4.3
has been merged with the 4.2 code to create this release. Not enough has
been done to warrant a complete release, but I have decided to release this
version mainly as a bug fix.
To install this version of MegaBook, you must have the original MegaBook 4.2
archive. If you dont, obtain AAA_MegaBook.lha from biz/dbase on Aminet.
Check before you download that the version number in the readme is 4.2Release
if it is not, then you do not need to install this version.
Once you have MegaBook 4.2 installed, you can simply extract this archive
directly into the MegaBook directory. Alternatively, you can copy the files
over after extracting somewhere else. Whichever way you use, you must make
sure the directory structure remains intact.
The main changes in this version are as follows:
MegaBook is now accesses the end of the database in a different way, giving
a 100% speed increase on this part of the database. The first half is still
accessed in the same way so there is no speed increase. For those that are
interested, this is how it works. From day 1 MegaBook has used an exec list
to store its database. This has the advantage of being easy on small
databases and is easily attached to a ListView, but has the disadvantage of
being slow on the end part of large databases. This would be better done as
a hash table, but that would mean a major rewrite of large parts of code
and therefore hasnt been done yet. MegaBook now stores a pointer to the
middle of the database, and does all searches above this point from that
pointer instead of from the beginning of the list. On large databases the
difference is phenomanl. We created a 32768 record database for testing.
On the old MegaBook it was noticeably slow to click on any middle to end
record of the database, also goto was slow. This showed its ugly head in
the form of MegaBook hanging for long periods. The new version saw this
cleared up totally. It is still a little slow on the second to middle
record, but this is barely noticeable. There are ways I can speed it up
even more but on less then 1000 records it isnt worth doing. If anyone
uses MegaBook for huge databases over 1000 records in size, please let me
know.
There was a major problem with the font code. If the font you selected didnt
exist or was corrupted, MegaBook would get stuck in an infinite loop. I
fixed this and added large informational messages to tell you exactly what
is happening with regards to font problems. Also, I added an intelligent
recovery to font problems. If the font you have specified does not open,
MegaBook tries to use your system default font. If that doesnt open, it
tries to use topaz 8, if that doesnt open (unlikely!) MegaBook fails to
run. In low memory situations this can cause a wealth of requesters, as
MegaBook tells you everything that is going on as it does it. This i feel
is the lesser of two evils. If you use MegaBook in a low memory situation
frequently, please let me know and Ill disable those requesters for you.
An MBIO for OnyxBase 1.0 is also now included. Thanks must go to Stefan
Blixth for the information on the file format.
A number of smaller bug fixes and improvements have also been made. Please
see the release notes for more information.