more on comments? Does faster broadband really matter?

[David Farber <dave () farber net>]

Begin forwarded message:

From: Mary Shaw <mary.shaw () gmail com>
Date: December 28, 2005 2:28:56 PM EST
To: dave () farber net
Cc: ip () v2 listbox com
Subject: Re: [IP] more on comments? Does faster broadband really matter?

Brian says, correctly, that the question shouldn't be whether current  
web surfing can take advantage of faster broadband but what  
applications in the future could use it well.

I say the future is now. For example, I store my working files on a  
central server so that I can use the current versions of files from  
my office, from home, or on the road.  I've tried keeping multiple  
synchronized copies of files, and it just doesn't work. So I keep one  
master copy in a central location close to a high-speed link (this  
also gets me other services, like automatic backups).  I haven't  
measured actual effective bandwidth recently, but my machine reports  
10Mbps.

Now, many of my files are multiple megabytes (some of my class  
lectures are a bit over 2MB, for example).  I definitely notice the  
time it takes to open and save files.  Autosave to the central server  
is a serious nuisance.  And this solution simply doesn't work for  
photographs.at current bandwidths -- it's not unusual for the working  
copy of a Photoshop file with some image manipulation in progress to  
be 20-30MB, and when I had to move a couple of 120MB bitmap files I  
started the copy when I quit for the night.

With applications moving to this central-service model, more and more  
people will find what I've already discovered -- bandwidth matters a  
lot, and upstream bandwidth is every bit as important as downstream.

While I'm on the subject, continuous service is absolutely critical.   
My current ISP is pretty casual about rebooting routers, and several  
times this has corrupted the file I'm editing.  If all the ISP is  
thinking about is downloading web pages, occasonal short  
interruptions might not matter, but for me it matters a lot. Of  
course, this happens when I'm working, late in the evening when the  
ISP's customer service, such as it is, has gone home.

Mary Shaw


On 12/27/05, David Farber <dave () farber net> wrote:

Begin forwarded message:

From: Brian Sniffen <bts () alum mit edu >
Date: December 27, 2005 6:53:18 PM EST
To: dave () farber net
Cc: dewayne () warpspeed com
Subject: Re: [IP] comments? Does faster broadband really matter?

> The gist of his argument is that most online activities, like
> standard websurfing, are not significantly sped up by high-bandwidth
> connections, and the few that are, such as downloading, are not
> typically time-sensitive anyway:

Sure.  The introduction of Single Side Band hasn't done much to
increase Morse transmission, either---and the Interstate Highway
System hasn't made the pony express any faster or any cheaper.  In
fact, it seems to have caused a counterintuitive decrease in buggy
whip purchases!

Just so, increasing the commonly available symmetric crosscut
bandwidth won't make web browsing more popular---just like commonly
available T1s didn't make UUCP more popular and Fast Ethernet didn't
make Gopher any more popular.  But AOL usage---"what everyone wants"
in 1994---was under 20 kbps, for use on phone lines.  Mr. Malik says
the web currently works at about 1 Mbps.  Of course 100 Mbps symmetric
connections to the home won't make a difference in 1 Mbps web
surfing---just like they don't make a difference in 20 kbps AOL usage.

But what else will they enable?  Multiple simultaneous video streams?
Immersive applications?  Real-time collaborative software?  Cheap,
easy reuse of spare cycles, a sort of NeighborsExcel@Home?

This seems intuitively obvious: As long as I'm regularly being limited
by the bandwidth I have, I don't have enough!

It's a little harder to dispose of the server-side-throttling
argument.  Again, if the use is quite different from the current day,
why should we expect it to use central servers?

We really are AOL users, or perhaps telegraph operators, trying to
predict the Web.

-Brian

--
Brian Sniffen                                       bts () alum mit edu


-------------------------------------
You are subscribed as mary.shaw () gmail com
To manage your subscription, go to
http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting- 
people/


-------------------------------------
You are subscribed as lists-ip () insecure org
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/