[standards-jig] Constraining standards? (was: DiscussiononJEP-0016: Server-Based Privacy Rules)

Iain Shigeoka iainshigeoka at yahoo.com
Thu Jan 24 17:53:15 UTC 2002


On 1/23/02 4:28 PM, "Shawn Wilton" <shawn at black9.net> wrote:

> No, the primary complaint against ms products are code bloat.  Everything
> crashes and their office suite is hundreds of megs in size.  if it were 30
> megs no one would be complaining.

Right.  But most of the code bloat is from feature bloat.  More features
almost always equals more more code.  There's really no way around it.  MS
has traditionally made higher feature count the primary goal for their
products.  The sacrifice has been code size, bugs, and security problems...
Bloat.

There is a chance that Jabber could fall into the same quicksand.  If you
want to say you support all Jabber standards, and we accept every possible
protocol suggestion under the sun as "Jabber recommendation standards"
Jabber software will necessarily become bloated.  More features can be a
very seductive lure.  If I want to create a "universal" Jabber client, I
would be slitting my throat if I didn't implement every recommended standard
out there...

> The rendering suggestion I put up is just that a suggestion.  I should not
> have used the word standard.  Poor choice on my behalf.  I just want a list
> of these items that client authors can go to w/o having to reverse engineer
> everyone else's client.  :-)

I know.  It falls into the "Jabber recommendation standards" that we've been
tossing around lately.  These would be Jabber standards that aren't in the
"core" set that are required for the "Jabber Powered" logo.  However they
would be standards because they've been accepted as such by the JSF.  There
are benefits to having them (as you point out).  JEP-0016 is sort of just
unfortunately being used as an example because it is a perfect example of a
"Jabber recommendation standard".  :)

My rant has been, how many of these are we going to create/accept?  Should
we even try to constrain the number of these recommendation standards or is
more better.  My personal stance so far has been that "less is more".

I'm not saying "my way or the highway."  I just want to bring up the issue
and hold a debate on it so we're aware of what we're doing (and not just
accidentally falling into any particular pattern).  The debate will
hopefully help the JSF council make decisions about the future of Jabber as
part of their standards process.

-iain


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com




More information about the Standards mailing list