Testing on multiple platforms

Shari gypsyware at earthlink.net
Mon Nov 11 08:15:01 EST 2002


>A word of advice: If you plan to release software for various platforms, and
>provide features and support for those platforms, you'd better plan on
>owning those platforms.

I have one OS.  Some people have my OS, others do not.  It has always 
been this way.  Things change not only from platform to platform, but 
OS version to OS version.  I've heard it's common on Windows, the 
change from Win 95 to 98 to 00 to NT, apps break frequently.  Most OS 
upgrades break apps.  A change from OS 9.1 to 9.2 can break an app.

It does not matter which OS I have.  My testing will always be on 
that one specifically.  Unless you have 5 or 10 computers, you CANNOT 
test on all OS's.

A Window's developer would need at least 4 to test on the various 
flavours of Windows.  Minimum.

A Mac developer would need at least 3.  Minimum.

A cross platform developer would need a minimum of 7 computers, if 
you own all the flavours that your program might run on.

I do not know ANYONE, who has this many computers running different 
OS's.  Even the most successful shareware authors, I doubt own this 
many computers, or this many OS's.

For a shareware developer, it is very simple folks...

You get to test before you buy.  This is a KEY point for shareware. 
Not like software you buy in the store, that you are stuck with 
whether you like it or not, shareware programs allow you to try them 
out first.  If an anomoly in the program is undesireable, you don't 
have to buy it.  If the features are undesireable, you don't have to 
buy it.

There is a difference between a bug, and an anomoly specific to a 
platform or OS.  An added menu is an anomoly.  An icon whose colors 
are not the same on different platforms is an anomoly.  5+5=12 is a 
bug.

If you can own one computer, and put out a program that is stable on 
the various platforms, this is sufficient.  If it functions properly, 
this is sufficient.

Perhaps for those who are not shareware developers,  who are 
contracted by someone to create a program for a specific platform, it 
is not sufficient.  I suspect many on this list fall into that 
category.

Even Metacard itself, is a try before you buy proposition.  If you 
don't like the way it works, you move on to another tool.  I like 
Metacard.  It does have its quirks.  But it is a top notch 
development tool, and I give it the highest praise.  I would 
recommend it to anyone.

Shari C
-- 
--Shareware Games for the Mac--
http://www.gypsyware.com



More information about the metacard mailing list