[extropy-chat] Agreement on principles. (was Microsoft)

[J. Andrew Rogers]

On May 22, 2006, at 8:20 AM, KAZ wrote:
> They screamed and cried that Mosaic, and eventually netscape, were  
> evil because they supported graphics, which were not necessary for  
> sgml-driven data webs, and did not "pay for" the bandwidth (by the  
> amount of information delivered) the way text did. "EEEvil! Sound  
> the Conch!"


Oddly, I do not remember this and I was there.  Yes, bandwidth was  
limited back then, but the only discussions I remember were about  
making *efficient* use of graphics, not eliminating them.


> Like people saying "You can't have a four line quote header,  
> because you're reducing effective bandwidth!"
>
> /snort
>
> Yeah, we're SO short of bandwidth, these days.


ClueTime:  the bandwidth people are trying to save is their own, not  
their DSL or ethernet drop.  It takes humans economically significant  
amounts of time to sift through piles of irrelevant text for that  
nugget of wisdom.

For example, with few exceptions, people that quote more than a  
pageful of text before responding frequently get their email dumped  
straight to /dev/null by me.  I receive many hundreds of emails a  
day, and if the potential value of the post is not immediately  
obvious either by content or sender, I often trash it.  Are they  
wasting my Internet bandwidth?  Yes, but I really could not care  
because I have far more bandwidth than most such that the cost is  
insignificant.  On the other hand, the many additional keystrokes and  
time spent scanning with the Mark I eyeball is damn expensive when  
you have as much text to go through as I do in an average day.

It is not about computer bandwidth, because that gets faster every  
year.  It is about human bandwidth, which is painfully finite and  
still not getting any faster.

[...much poorly focused flamebait elided...]


> The "broken system" is that of sticking blindly to what some  
> bureaucrats announce, even when the marketplace has pissed all over  
> it and moved on.


If the marketplace has moved on then stop claiming an implementation  
of the old standard.  If you really believe in this anarcho- 
capitalism, it should not bother you that a huge chunk of the market  
disagrees with your assertions.  If defective implementations of de  
facto standards are obviously superior, then they will win over the  
long term. This is not about new standards, but making claims of  
conformance to an old standard when that is clearly not in evidence.


> I wonder...WERE you one of the guys complaining about Mosaic,  
> saying a lynx/gopher type was better for the net's limited  
> bandwidth potential?


Strawman.  It really has nothing to do with network bandwidth  
consumption.

I use a fancy modern GUI email client that seems to render everything  
just fine no matter what perversion has been applied to the function  
of sending a text message.  I still adamantly support a clean plain  
text email implementation.  I have no problems at all with modifying  
standards, de facto or otherwise, provided that one can demonstrate  
an obvious and necessary improvement versus the existing standard.  I  
am not affected by these choices in practice, but I still recognize  
when a choice is a bad choice and remain an advocate of good choices  
by default.  There is a cost to breaking widely accepted standards,  
so there damn well better be a benefit for that cost.

More often than not, when people think a carefully engineered  
protocol needs to be "improved" to support some minor pet feature,  
all it really shows is that they do not understand the protocol under  
discussion or the theory behind the protocol.  For a classic example  
of stupid protocol design through ignoring literature and not  
understanding existing equivalent protocols, one only needs to look  
at XML.   Wildly popular, but with serious fundamental deficiencies  
that make it useless for many of the problems it was supposed to  
solve, despite these same problems having been well characterized and  
solved long before.

The engineers and scientists of prior generations were not barely  
functional idiots with nothing but the shallowest of insights.   
Engineering brilliance is not a recent phenomenon.  The collective  
ingenuity of those old minds may in fact be missing the mark, but  
only a fool would so casually dismiss the invested wisdom and  
experience of those designers without very carefully considering the  
nature of the problem they were trying to solve.


J. Andrew Rogers