[standards-jig] On the Subject of Avatars
bob at wyman.us
Mon Jul 21 20:50:55 UTC 2003
Evan Prodromou wrote:
> My experience with avatars in other IM
> systems is that they are in fact _rarely_ changed.
It is entirely possible that the reason they aren't changed more
frequently is because of a lack of protocol or application support to
facilitate more frequent changes. Thus, you may be seeing the effect of
a failings in the current systems rather than an indication of actual
I think if I had a system that allowed me to change the
appearance of my avatar whenever I wanted to, it is likely that I would,
in fact, tend to change it frequently. For instance, I might tie the
avatar into my current "state" so that if I was "away" it would have one
appearance yet change to something else when I was present. Also, I
might like to change it from a "smiling face" to a "frowning face" by
clicking on a mood button in response to what I was seeing in an IM
session. Then again, I might like to tie my avatar's appearance to the
local weather report and have a "sunny" avatar when the weather was good
but something else if it was raining. If I was playing a game on the
side, I might want the avatar to reflect my current status (strength,
character name, current level, etc.) in the game...
Providing for asynchronous, automatic synchronization of avatars
might enable new applications of avatars and greatly increase the
utility of the things. On the other hand, it might just give people
another way of doing irritatingly cute things that waste bandwidth.
It is always dangerous, when trying to determine requirements to
allow past experience to influence you too much. Much of what we do in
this business is predicting the future -- that cannot be avoided.
> <evanp> Hey, Bob, I just changed my avatar. Check it out!
> <bobw> [click]
> <bobw> Cool!
> <evanp> Thanks!
I think this might be a bit different in real life... I would
<evanp> Hey, Bob, I just changed my avatar. Check it out!
<bobw> Evan, if you want to publish an update why not use PubSub!
Asking me to do work to see your latest image is just a bit
> my experience with questions about avatars on
> this list is that people want a working solution
> now, not a year from now. DSAs can be implemented
> today, by anyone.
It is always the same argument it seems... We can have 80% now
rather than waiting just a little bit longer and getting more. So, some
folk argue for doing it "right" and others argue for doing it "fast and
good-enough." Then, the decision is to support the "fast" solution as a
temporary solution while the "do it right" folk work out the "real"
solution. But, then, once the "do it right" folk are finished with the
design, everyone starts to argue that "we can't change now" since there
are too many existing implementations.
I've been doing protocol design for almost 25 years now and I've
rarely seen a "fast" solution that didn't end up causing more trouble
than it was worth.
More information about the Standards