1  |   Order by default

Jul 14, 1998, 7:37pm

Hi,

My main wishlist item, is true support for 3Dfx Voodoo/Voodoo2
graphics. That way, we can have more silky smooth ActiveWorlds
at 30 frames per second. Perhaps a full-screen 3D browsing keypress
toggle, for the immersive experience too!

Thanks,
_______________________________________________________________________
____
Mark D. Rejhon WinNT.Linux.Win95 \ / mailto:marky at ottawa.com
http://www.marky.com/ C.C++.VB.Shell \/ AlphaWorld Home 10S 15W

"A friend is someone who will be there without asking anything of you.
A friend is someone you know that knows you, and accepts you."
_______________________________________________________________________

Jan 20, 1999, 3:25am

Hi,

I know that Direct3D can be enabled by an unsupported option in the
AWORLD.INI file, at least in a past version of ActiveWorlds.
(if it's no longer supported in current versions).

However, my number 1 desire for ActiveWorlds is official Direct3D
support. Right now, ActiveWorlds 3D severely is becoming more and
more outdated, compared to many of today's newer 3D games such as
Unreal. At the very least, Direct3D (DirectX6) support should be
implemented directly without RenderWare as an intermediary step.
I believe this would vastly improve the framerate in ActiveWorlds
for many users who use fast computers (450 Mhz boxes, as well as
overclocked Celeron boxes) and fast 3D accelerators (such as the
Riva TNT, 3Dfx Banshee, etc).

My number 2 desire for ActiveWorlds is a full-screen-mode toggle,
where I can hit a button and then suddenly get full screen 3D for
browsing. Or a keypress shortcut such as Alt-Enter, like those
used in some Windows video games or DOS boxes, to toggle between
windowed and full screen. This would be great for sightseeing,
or public chat. (Complete chat and building support isn't that
necessary for full screen mode) That is, there won't be a chat or
building interface, but you would be able to see text floating above
other people, and be able to type what you want to say publically
Just like the "T" chat key in Quake 2 - just very basic chat support
during full screen 3D browsing mode.

My number 3 desire is better 1st-person/3rd-person support. While
I can use the Home/End keys relatively easily, I think its behaviour
needs to be changed so that it is easier to set a certain viewpoint.
As it stands now, it's almost impossible go into 3rd person and have
the "camera" hover just a foot or two behind you - it just jumps
at least 20 feet back at the lightest and briefest tap of the
"End" key - making my avatar look as small as surrounding avatars.

- I think that "End" should be a "zoom-out" only key, while
the "Home" key should be a "zoom-in" only key.
- Hitting "End" for the first time would switch the camera instantly
to hover behind your avatar by about 2 or so feet.
- Tapping "End" or holding down "End" will zoom the viewpoint outwards
incrementally (no acceleration - zooming shoudld stop instantly
after releasing "End" key so that you don't overshoot(
- Tapping "Home" or holding down "Home" should zoom the viewpiont
inwards incrementally. (Once you get as close as 2 feet behind
the avatar, it would snap back into 1st person view instead of
3rd person view)

In general, that would give people more control over the 3rd-person
viewpoint, and be more easily able to adjust exactly the distance they
want from the avatar.

_______________________________________________________________________
____
Mark D. Rejhon WinNT.Linux.Win95 \ / mailto:marky at ottawa.com
http://www.marky.com/ C.C++.VB.Shell \/ AlphaWorld Home 10S 15W

"A friend is someone who will be there without asking anything of you.
A friend is someone you know that knows you, and accepts you."
_______________________________________________________________________

Jan 20, 1999, 4:47am

BTW,

I apologize for posting in the "beta" newsgroup earlier. I tried
to cancel the post when I read the rules in the beta newsgroup.

I'm just a very infrequent visitor and should have read the rules
first before posting in the "beta" newsgroup.

Thanks for understanding,

Mark Rejhon
http://www.marky.com/3d/ - see my ActiveWorlds home here!

Jun 7, 2001, 3:19am

Hi,

> I believe other users requested this as a wish back a few years ago when
> they were using old 3dfx Voodoo technology. It still has not been
> implemented.

One of those would be me ;-)
Although now I use Geforce2 GTS (Geforce3 when price goes down, of course..)

I still want this feature too. Even a rudimentary way of doing this would
be OK. Such as one Alt+Enter toggle to automatically do this:

* Maximize AW window
* Disable all tabs
* Disable web window
* Disable toolbars
* Maybe disable (or reduce size) of chat area

This should be doable with the existing AW architecture, although this
would be a pseudo-fullscreen mode. This is relatively simple to
implement since you can already do this by various different hotkeys
(except for toolbars and chat area)

_______________________________________________________________________
____
Mark D. Rejhon Win2k.Linux.Win98 \ / mailto:marky at ottawa.com
http://www.marky.com/ C.C++.VB.Shell \/ AlphaWorld Home 10S 15W

"A friend is someone who will be there without asking anything of you.
A friend is someone you know that knows you, and accepts you."
_______________________________________________________________________

Jun 7, 2001, 2:44am

Suggestion: Avatars being able to move with moving objects!

I know you get this suggestion a lot, but forward this to the programmers
anyway because I am including some novel suggestions on how to make it
possible.

I am a computer programmer too, so I am going to make this suggestion unique
by suggesting an algorithm to make it possible without increasing data
traffic. I know you get this suggestion a lot, but a possible programming
algorithm could be the following:

Avatar movement is currently recorded using world coordinates and vector
movement. My suggested solution is to instead make coordinates relative to
the object you're standing on, and leave it to the PC's responsibility to
keep the avatar attached to the object at 30 frames per second (even if
position updates only occur every 1 second)

(1) Whenever an avatar steps onto a moving object (rotating or sliding), the
avatar is "flagged" as being attached to that object. The avatar's XYZ
co-ordinates and vector motion would be locked to the object.

(2) No frequent avatar position updates would be required. Once the server
broadcasts the avatar's new "attachment" to a moving object, it would just
transmit one set of coordinates RELATIVE to the moving object itself (not
the world).

(3) It would be the personal computer's responsibility to keep the avatars'
position continually updated in realtime, based on the last time the server
broadcast an avatars' attachment to a specific object, and the position
co-ordinates relative to origin of object. Thinking of it another way,
it's as if the object would have its own set of "world coordinates". That
way, the server does not have to constantly retransmit co-ordinates of
avatars everytime the avatar is moving along with the object.

(4) This would only work with synchronous moving objects, for obvious
reasons (since everybody must be seeing the same object in the same
position)

(5) Server will just periodically rebroadcast avatar position, vector
movement data would be relative to the origin and orietniation of the moving
object. The position broadcasting can be just as infrequent as 0.5 to 1
second, as it currently does now. This does not increase data traffic at
all!

(6) Avatar affilation with a moving object will only occur with gravity
contact (that is, firmly standing on the moving object - not contacting the
side of a moving object, or bumping into it while flying)

(7) Server will broadcast when the avatar steps on the moving object, or
steps off the moving object.

(8) Since moving objects can consist of multiple synchronized moving
objects, stepping from one moving object to another can be tricky. I leave
this exercise to your own discretion, but you could simply just recommend
world builders to make one moving object into a single model, for
simplicity.

Imagine it...! Moving buses, trolleys, boats, monorails. Even cruise ships
with 100 people on them, perfectly stationary on its deck with intermittent
movements. As if a cruse ship was a separate world in itself, i.e. a moving
world (boat) within a stationary world (water).

Thanks,
Mark Rejhon
Citizen #6831
http://www.marky.com
marky at ottawa.com

Aug 27, 2001, 3:59pm

Suggestion: Avatars being able to move with moving objects!

I know you get this suggestion a lot, but forward this to the programmers
anyway because I am including some novel suggestions on how to make it
possible.

I am a computer programmer too, so I am going to make this suggestion unique
by suggesting an algorithm to make it possible without increasing data
traffic. I know you get this suggestion a lot, but a possible programming
algorithm could be the following:

Avatar movement is currently recorded using world coordinates and vector
movement. My suggested solution is to instead make coordinates relative to
the object you're standing on, and leave it to the PC's responsibility to
keep the avatar attached to the object at 30 frames per second (even if
position updates only occur every 1 second)

(1) Whenever an avatar steps onto a moving object (rotating or sliding), the
avatar is "flagged" as being attached to that object. The avatar's XYZ
co-ordinates and vector motion would be locked to the object.

(2) No frequent avatar position updates would be required. Once the server
broadcasts the avatar's new "attachment" to a moving object, it would just
transmit one set of coordinates RELATIVE to the moving object itself (not
the world).

(3) It would be the personal computer's responsibility to keep the avatars'
position continually updated in realtime, based on the last time the server
broadcast an avatars' attachment to a specific object, and the position
co-ordinates relative to origin of object. Thinking of it another way,
it's as if the object would have its own set of "world coordinates". That
way, the server does not have to constantly retransmit co-ordinates of
avatars everytime the avatar is moving along with the object.

(4) This would only work with synchronous moving objects, for obvious
reasons (since everybody must be seeing the same object in the same
position)

(5) Server will just periodically rebroadcast avatar position, vector
movement data would be relative to the origin and orietniation of the moving
object. The position broadcasting can be just as infrequent as 0.5 to 1
second, as it currently does now. This does not increase data traffic at
all!

(6) Avatar affilation with a moving object will only occur with gravity
contact (that is, firmly standing on the moving object - not contacting the
side of a moving object, or bumping into it while flying)

(7) Server will broadcast when the avatar steps on the moving object, or
steps off the moving object.

(8) Since moving objects can consist of multiple synchronized moving
objects, stepping from one moving object to another can be tricky. I leave
this exercise to your own discretion, but you could simply just recommend
world builders to make one moving object into a single model, for
simplicity.

Imagine it...! Moving buses, trolleys, boats, monorails. Even cruise ships
with 100 people on them, perfectly stationary on its deck with intermittent
movements. As if a cruse ship was a separate world in itself, i.e. a moving
world (boat) within a stationary world (water).

Thanks,
Mark Rejhon
Citizen #6831
http://www.marky.com
marky at ottawa.com

Aug 27, 2001, 3:59pm

Hi,

While I am not a 3.2 beta tester, I read the preliminary manual at
http://www.activeworlds.com/help/aw32/ with extreme interest.

I observed that Software mode is making a comeback and is the
new default mode.

I think this is a bad idea because new users are the ones most likely
to be using newer computers with less problems.

You will get many new people with new computers turned away from
ActiveWorlds; because they observe slow performance and are too
impatient to reconfigure the browser. Which means slower revenue
growth for AWLD.

Why not:

- Autodetect the prescence of DirectX 8 or later
(This is a good indicator that the computer's recent or been kept up to
date)

or

- Autodetect an nVidia/RADEON graphics card and other high-compatibility
cards
and select Direct3D automatically.

or

- Query the Direct3D driver for supported capabilities.
If multitexturing and T&L is supported, then automatically enable Direct3D
(The prescence of T&L is a good indicator of a recent graphics card;
and recent drivers; which means fewer problems)

I urge you to keep Direct3D the default, or at least automatically detect
the problem cases. Nowadays, there's just two main cards in new computers:
RADEON or nVidia Geforce2/3 .... They are usually very problem-free nowadays
so it is a prudent idea to at least automatically detect this (query
graphics
driver maybe, and search for "RADEON" / "Geforce" in driver name string?)

Just making an observation!

Thanks!
Mark Rejhon

Aug 27, 2001, 4:10pm

Hi,

In a nutshell, my post solves:
- Good Out of Box Experience for New Users (business goal)
*and*
- Keeping Tech Support Volume Down (programmer goal)

It is easy for a programmer (like myself or Roland) to forget business goals
which is essential in the long term - there needs to be major improvements
to the immersion / out of box experience to attract an audience .... And
going to software mode default is a definite step backwards in a business
perspective.

Easiest solution: Automatic detection for common cases of problem-free 3D
configurations. (Thankfully we're down to two main competitors: nVidia and
ATI, so autodetect is easy)



[View Quote]

Aug 27, 2001, 7:29pm

ActiveWorlds needs to make money; most new computers and new users
will get impatient with slow 3D, especially if they also play 3D
videogames....

So there's incentive to cut tech support complaints by BOTH the 2.2
upgraders,
the completely-new users who expect a good "out of box" experience
(that requires enabling 3D acceleration), the 3.0 users who needed to
reinstall
Windows and reinstalled ActiveWorlds complain about loss of performance;
etc.
You have to look at the WHOLE picture ;-)

At the very least, detect for the common cases of recent problem-free
configs.

Roland could easily err on the side of caution; but you could at least
make sure that Direct3D mode is defaulted on most computers
purchased today from DELL and Gateway (basically, recent drivers /
recent DirectX / recent Radeon or Geforce2/3).
I have observed that it's starting to get pretty problem-free
with recent nVidia/ATI systems as I buy new computers for friends
and family members. Things seem to run right off the bat on these
2001-purchased computer systems.

///////////

Here's another idea for idiot-proofing:

Automatically revert to software on the next execution.
Basically, you would simply set an AWORLD.INI setting called
"StartupCount" upon startup and "ExitCount"
upon exit, to automatically detect (1) whenever AW starts for
first time, (2) whenever AW crashed on previous execution.

Basically, if you start AW for the second time, and it
finds StartupCount=1 and ExitCount=0 that means it crashed
on the last time it ran (because it never had a chance to increment
ExitCount). You can then execute a "safe mode" ActiveWorlds
configuration in this event! And people won't have to
spend hours trying to figure out how to fix their installation.
The only problem, is that it might have crashed for other
reasons. So you might want to prompt a confirmation
dialog "Your ActiveWorlds did not shut down properly on
the previous execution. Do you want to reset ActiveWorlds
configuration to maximum compatibility (slowest performance)
settings? [Yes] [No]"

Could cut down tech support volume...


[View Quote]

1  |   Order by default