[JDEV] Emoticons: guidelines

Mattias Campe mattias.campe at rug.ac.be
Fri Apr 19 09:51:38 CDT 2002


 > ----- Original Message ----- From: "Mattias Campe"
 > <mattias.campe at rug.ac.be> To: <jdev at jabber.org> Sent: Thursday,
 > April 18, 2002 6:44 PM Subject: Re: [JDEV] Emoticons: guidelines
 >
 >
 >
 >> Richard Dobson wrote:
 >>
 >>> Well what about creating a new x element for message which
 >>> allows people
 >>
 > to
 >
 >>> define emoticon text replacements e.g.
 >>>
 >>> <message type="chat" to="medds at me.com" from="friend at friend.com">
 >>> <body>This is a emoticon containing message :) (L) ;) :D</body> <x
 >>> xmlns="jabber:x:emoticon"> <emoticon text="(L)" icon="love"/> <emoticon
 >>> text=":)" icon="smile"/> <emoticon text=";)" icon="wink"/> <emoticon
 >>> text=":D" icon="grin"/> </x> </message>
 >>>
 >> [...]
 >>
 >> I'm not sure if I understand it, but doesn't this mean that the
 >> client has to support this x element? So if the client doesn't
 >> want to support that element, people using that client will get
 >> "This is a emoticon containing message :) (L) ;) :D" I'll
 >> understand the ":) (L) ;) :D" part, but what about (L)?
 >>
 >> And won't this cause to much overhead? In the example you gave me
 >> the body itself contains less characters than the x-element.
 >>
 >> Just my thoughts...
 >>
 >> ..m.
 >>
[...]
Richard Dobson wrote:
 > Yes it does mean that the client has to support this x element, it
 > means that it works fine with clients that dont support it, this
 > method also allows the message sender complete control of where and
 > when the emoticons are displayed, so the message will be displayed as
 > the sender intended, and not as happens in msn messenger certain
 > strings getting replaced when you didnt want them to, and not relying 
 > on the recipient client to determine if
 > a particular string should be changed into an emoticon or not. Also
 > as jabber users will not all be using a single particular client it
 > allows better cross-client compatibility. Also the problem of it
 > creating lots of overhead, well the syntax of the x element could be
 > cut down, I did it long and readable so it was easier to understand
 > for people here when reading my example.
 >
 > My 2p
 >
 > Richard

I thought about it a bit (hmm, actually a lot ;) ) and I started to like 
the idea. So to reduce the overhead the message could be sth. like:

<message type="chat" to="medds at me.com" from="friend at friend.com">
	<body>This is a emoticon containing message :) (L) ;) :D</body> 	
	<x xmlns="jabber:x:econ">
		<econ text="(L)" icon="luv"/>
		<econ text=":)" icon="smile"/>
		<econ text=";)" icon="wink"/>
		<econ text=":D" icon="grin"/>
	</x>
</message>

I should even make it shorter by omitting the smileys from the x element 
because they explain themselves. In this example the overhead would be 
reduced to 3/4!! So the example would be:

<message type="chat" to="medds at me.com" from="friend at friend.com">
	<body>This is a emoticon containing message :) (L) ;) :D</body> 	
	<x xmlns="jabber:x:econ">
		<econ text="(L)" icon="luv"/>
	</x>
</message>

I use a smiley maybe at least every 5th message so that would reduce 
overhead a lot!


Now, why my sudden change to like the x-element thought too (too = like 
the colon method):

pros:
~~~~~

+ _internationalisation_ : I think a lot of people here on 
jdev at jabber.org have English as their native language. So "I :love: you" 
sounds very familiar, still, for sb. like me (speaking Dutch) it would 
be "Ik :love: van jou" or French "Je t' :love:" or German "Ich :love: 
dich" it looks a little bit silly doesn't it? E.g. French people are 
very keen on their own language (e.g. they use "un ordinateur" for "a 
computer").

+ _scalability_ : if the colon method would be used we are stuck too 
whatever is decided to be standard. Suppose we all agree to use :luv: to 
show a heart as graphic. So, it becomes a standard, but after a few 
weeks people start complaining about not getting it, WTF is :luv: ? 
"Euhm, it means :love:, sorry folks, but we can't change it :("

+ ... (plz add if you see some more)


Still I see some problems too, "the cons":
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

- _time_ : It will take time before all clients can support this new 
x-element. The colon method doesn't need time because it explains itself.

- _non-supporting-emoticons-clients_ : Will a client developer 
developing a client that doesn't (want to) support emoticons want to 
implement the new x-element to translate e.g.

<message type="chat" to="medds at me.com" from="friend at friend.com">
	<body>This is a emoticon containing message :) (L) ;) :D</body> 	
	<x xmlns="jabber:x:econ">
		<econ text="(L)" icon="luv"/>
	</x>
</message>

to: "This is a emoticon containing message :) :love: ;) :D"

instead of just: "This is a emoticon containing message :) (L) ;) :D" ?

Because that is the whole point here: making it clear to people who 
don't (want to) use emoticons...

- ... (plz add if you see some more)

~~~~~~~//end of cons


Theo, if you like some of the ideas and want to use them to put on 
http://www.theoretic.com/?Jabber_Client_Guidelines , go ahead!


To put things a little bit easy,* too easy actually, but anyway: "Give 
me a good and stable client with working file transfer and I can 
convince my younger sister (15 years old) in 1-2-3 to switch to Jabber, 
give me the same + cooler emoticons than MSN and I can do it in 1-2 (and 
if you can also give me a client that is fully skinnable (winamp3 like) 
I can do it in 1)..."


Btw I won't be checking jdev at jabber.org this weekend, just that you know...*


greetz
.m.




More information about the JDev mailing list