Google speed bump draws scorn

(continued from previous page)

it collects and what Google does with the information.

"The tool offers a plausible consumer benefit. But it makes me uncomfortable because it's Google collecting yet more information about everyone and doing it in a way that's not necessary," said Ben Edelman, a Harvard University researcher who investigates software applications.

For example, he said, it's unclear whether Google will tie information collected from click-stream data to its cookies. Cookies are tiny tracking tags used by most Web sites to associate a specific computer or user with his or her activity online. Often, cookies are used to remember passwords or log-ins, as well as information such as the user's geographic location or past preferences to better serve Web pages.

"The most important issue they don't address is what click-stream data is tied to the Google cookie," said Computerbytesman.org's Smith. "My recommendation: Purge Google cookies and often."

Mayer said that that click-stream data from Web Accelerator is not associated with the computer's cookie.

"To date, we're not doing anything with this data in terms of market research. We have no plans, but should that change we would aggressively notify our users and give them some escape hatch," Mayer said.

To address the security flaw, Mayer said the company is deactivating the mechanism that caches vulnerable Web pages. Mayer said the problem happens only on a small number of sites, typically discussion groups, because those sites are not passing the proper no-cache header information. She said the company is also contacting Web masters to work with them on that issue. In addition, Google is looking at possible mathematical algorithms to prevent the caching from happening in the future.

Yet another contingent, Web designers, has criticized the application for deleting Javascript commands or pages of their sites within the software environment. That means that people using Web Accelerator may not be able to hit the back button to return to a previous page. Google spokesman Nate Tyler said the company is also looking into that issue.

Web acceleration tools were popular years ago during the dot-com heyday, when most people accessed the Web with dial-up connections. Internet service providers such as America Online have offered them for free to their dial-up customers as a means of improving the surfing experience. Market research firms such as ComScore Networks have also used the tools as a means to entice new subjects for research panels, watching their behavior online to estimate the popularity of Web sites, for example.

But in an era of widespread broadband usage--more than 50 percent of households in the United States have a high-speed connection--the value of such tools has diminished. Peter Christy, co-founder of market research firm Internet Research Group, said that despite that perception, a good Web accelerator can mitigate packet loss, or latency, as information is sent from router to router. It can also optimize how an object-heavy Web site is compressed and sent to a visitor.

Addressing the privacy issues, Cristy said there's always a basic trade-off between getting a service and getting absolute privacy.

"If you look at Google, this fabulously useful company, they make their money by selling people ads," he said. Besides search, "the way Google becomes useful is in building some model of who I am and what I'm interested in and delivering me ads. That's either really useful or very sinister."

the secure language "HTTPF"

Is this supposed to mean the encryption protocol HTTPS?

How does Google get access to passworded discussion boards anyway?

[SmokieUK]

RE:

Google lists password protected discussion sites because (as it was briefly stated in the article):

"[i]... the problem happens only on a small number of sites, typically discussion groups, because those sites are not passing the proper no-cache header information.[/i]"

[aabcdefghij987654321]

Author is confused about HTTPF and HTTPS

"Web Accelerator does not cache secure Web sites in the programming language "HTTPF" such as banking or credit card pages"

There is a fledging "HTTPF" filtering proxy (it's a web proxy which strips potentially executable code) and perhaps is what google are using.

And, yes, banking and credit card pages always use HTTPS.

I suspect the author knows little about either and has confused the two. I stopped reading about that point.

[ron williams]

Changing the Internet

What are all forums supposed to change and insert the "no cache" header. Whats going to happen if my members install this. Google needs to not be Microsoft and downplay there errors and release software too early. View my forum at <a class="jive-link-external" href="http://holycow.audionine.com" target="_newWindow">http://holycow.audionine.com</a>

[OneWithTech]

What the F$%&k is [httpf]

I'm going to shed a little somethin' on those who, maybe don't understand web code, let alone what exactly [httpf] is!

Here is [httpf] explained in detail:
<a class="jive-link-external" href="http://httpf.sourceforge.net/" target="_newWindow">http://httpf.sourceforge.net/</a>
Here is another little snippet on [httpf]:

--------------------------------------------
[html] httpf A WWW Security Proxy (in C using POSIX threads) for filtering HTTP and HTML to only forward allowed/harmless content.
--------------------------------------------

In English this means JACK. It's not a protocal in the sence of "true web" protocols. This protocol is only used in the circumstances stated above. And 95% of the web-developers out there will tell you that: "They've Never Used It"!

So that leads me to answer the last question for everybody. Google is very good at quite a few technologies, but also very good with the PR.

You see the Circle C that denotes a copyright at the bottom of our web pages. This nomenclature is used to protect our "Finished Works". The USPTO (United States Patent and Trademark Office) ensures that complying with the rules set by them our "publications","pictures",and "overall copy" are protected under these rights.

What if Google is violating the Copy Protection of every site that it "Caches" on there servers? My programmer was a little pissed that a "Vanilla Envelope" with the tag link "Do Not Bend" shows up at my house address to him at Matrix Structures Online. Funny, American Express just did that to me not to long ago.

Apparently it's OK to start using that data you've been mining for the last couple of years. All that mining at my expence and the expense of everybody else that likes to keep there copy protected.

How convienent for Chase, American Express, and Google, to use "My Copy" for there profit!

Hence the PR runaround!

www.tech01.net

[OneWithTech]

A better httpf definition

httpf is a kernel module to allow queueing of people for busy sites. When web site is overloaded with number of connections, new people are assigned a tracking number and accepted as their time has come...
httpf A filtering proxy which processes HTTP- and HTML-traffic to enhance the security: Remove javascript calls, check document types, remove client infos (operating system, browser version)

www.tech01.net

[dmehus]

It's a typographical error

Below is a copy I've sent to CNET News.com editors to alert them of the typographical error in the third paragraph of this story. To send your own reports of typos or factual inaccuracies, send e-mail to: <a href="#">[e-mail]</a> and a copy editor or the CNET News.com managing editor, Jon Skillings, will respond to your report.

&lt;Begin&gt;
Dear CNET News.com Editors:

There appears to be a typographical error in the third paragraph of "Google speed bump draws scorn" located at <a class="jive-link-external" href="http://news.com.com/Google+speed+bump+draws+scorn/2100-1032_3-5698447.html," target="_newWindow">http://news.com.com/Google+speed+bump+draws+scorn/2100-1032_3-5698447.html,</a> where "HTTPF" should be corrected to read "HTTPS". Or, ideally, instead of mentioning the acronym, mention secure sockets layer transfer or something similar. However, simply updating it from "HTTPF" to "HTTPS" would more than suffice.

Cheers,
Doug M.
dmehus AT telus DOT net
&lt;End&gt;

Another error

Also, read the first sentence of this paragraph:

The complaints center on the search giant's Web Accelerator, which was released on Wednesday. Downloadable software for broadband users, Web Accelerator is intended to speed access to Web pages by serving up cached or compressed copies of sites from Google's servers.

It seems like there should be something after the "released on Wednesday" part.

[rockerrb]

Privacy advocates need to give it a rest

Google's new web accelarator doesn't work any differently than any other service on the web: It's free and it's use is not mandatory. If your privacy is such a big concern to you then don't use it. Targeted web marketing is the past, current and future of the internet. Goolge and it's stock will continue to grow for a very long time.

[rockerrb]

Privacy advocates need to give it a rest

Google's new web accelarator doesn't work any differently than any other service on the web: It's free and it's use is not mandatory. If your privacy is such a big concern to you then don't use it. Targeted web marketing is the past, current and future of the internet.
Goolge is a huge multibillion dollar company. Any minor technical problems should be quickly and easily resolved. Goolge and it's stock will continue to grow for a very long time.

Its in beta

In my opinion google is doing nothing wrong and has not promoted or presented this app incorectly. The thing is in beta. Of course there will be errors. Of course there will be security leaks. betas should not be entered into lightly by anyone. When you use a beta you take responsibility for your own information and security because the software is known to be faulty. Its in beta to fix the faults and find out what is desired for the program. This is a non news item. Sheesh

[Jon Skillings]

The typo has been corrected

As several of you have pointed out, HTTPS is the correct term. The story has been fixed. Thanks for calling the typo to our attention.