enigma-devel
[Top][All Lists]
Advanced

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

Re: [Enigma-devel] Re: Changes Between 0.81 And 0.92


From: Daniel Heck
Subject: Re: [Enigma-devel] Re: Changes Between 0.81 And 0.92
Date: Tue, 24 Jan 2006 23:37:55 +0100

On Mon, 2006-01-23 at 21:30 +0000, Tarim wrote:
>    A good philosophy for new versions (unless you're IBM or Microsoft, of 
> course) is that they should be upward compatible with previous versions 
> wherever possible.  It's great to introduce new features - but make it so 
> that they have to specified as new objects, or new attributes - not that 
> you change the default actions of existing things.

Actually, there are two different kinds of incompatibilities between
0.81 und 0.92:

1) those that slipped in more by mistake than on purpose

2) those that broke backwards-compatibility intentionally in an attempt
to fix the internals of Enigma "once and for all"

Whether the rotor returns to its home position or not belongs to the
first category, as does the problem with the fire extinguisher.

A few incompatibilities were unavoidable, however, and make a
backwards-compatible 0.93 almost impossible.  For example, I made a
couple of changes to the collision response code and the numerical
integrator.  Do you remember the annoying sound that 0.81 generated when
your marble was pulled into a nearby block? Or how sluggish 0.81 ran
with many marbles on the screen at the same time?  You certainly don't
want these problems back, I guess! 

Unfortunately even subtle changes in these parts of the source code tend
to have unexpected consequences: suddenly friction does not behave the
same way as before, objects get stuck in each other, rotors and spinning
tops start moving erratically, etc.  I don't know whether you ever tried
a the developer's version during this time; if you did, I think you
wouldn't have expected that the 0.90 releases would turn out to be as
stable as it now is.

Anyway, the main reason why I made these fundamental changes back then
was that I always hoped for a 1.0 release in the near future.  My aim
was to produce a 1.0 release that would give level designers a stable
platform.  The XML-based level format was part of this scheme, as were
the changes to the physics engine.

>    Again, thanks to everyone who actually does the work, unlike Tarim who 
> just sits around complaining about a few of the details. :-)

:-) You know, that's *exactly* why every self-respecting software
company has its own QA department, to have people who just sit around
and complaing about a few of the details ;-)


- Daniel






reply via email to

[Prev in Thread] Current Thread [Next in Thread]