Forum Moderators: Robert Charlton & goodroi
System Requirements
Operating System: Win XP or Win 2000 SP3+
Browser: IE 5.5+ or Firefox 1.0+
Availability: For users in North America and Europe (during beta testing phase)
Press Release:
Google Web
Accelerator significantly reduces the time that it takes broadband users to
download and view web pages. The Google Web Accelerator appears as a small
speedometer in the browser chrome with a cumulative "Time saved" indicator.Here's how it works. A user downloads and installs the client and begins
browsing the web as she normally would. In the background, the Google Web
Accelerator employs a number of techniques to speed up the delivery of
content to users.
Looks like some of the Mozilla hires are paying dvidends.
Email, blog, guestbook and formmail spammers are going to absolutely love it.
Immediately blocked from all the sites on my servers. No need, no way.
Google you are getting carried away with yourselves.
btw, AOL has been using a proxy cache for years and it's the first thing users want to bypass.
Why would any broadband user want a 1%-5% speed improvement?
A. It will make the Google Browser more appealing when it is released!
It impresses me to no end that a company representative would actually take the time to not only read what we have to say, but to offer clarification, suggestions and advice.
Thank You.
Justin
End Off Topic
"Another focus that I know of is on compressing data that goes over the connection."
Now, this is the way to go.
"Prefetching" pages is just plain dumb waste of bandwidth and creates more problems than what it is worth.
A competitive broadband accelator that I use uses Content Sensitive Compression (CSC) to individually compress each element of a web page.
The speed increases on using the competitive accelator on DLS broadband have been averaging 40% (and I'm not even using the maximum compression settings, as they decrease the quality of images too much).
Over GPRS I've achieved 200%-300% speed increase.
What is their ultimate goal with pre-fetch?
not nice thought:
"behaviour targeted ads" -? - i had a conversation with a web marketing exec just yesterday and that side of the industry is saying that looking past "contextual" is "behavioural" - i.e. tracking surfing patterns etc... something others have been doing for awhile...
so targeting ads based on where someone has been and where they might go would give more useful targeting options for advertisers than just showing ads related to what page the surfer is currently on etc...
linking this in to AdWords somehow would produce this sort of option...
nicer thought:
seriously doubt this "behaviour tracking" would be employed by Google... and WebAccelerator is probably just another genuine app to offer users like Picasa...
Enjoy."
There's nothing in there about an easy ability to prevent the automated requests to our sites. Just a well worded attempt to explain it.
Would Google be satisfied if I "addressed" automated queries to them that they found problematic merely by explaining to them what I was doing?
I can see it now....
"I was sending you automated requests. I was doing so to your search pages but did not touch your PPC ads. This can have the effect of increasing load to your servers without the ability for you to monetize the traffic. You can identify my traffic by the header 'X-moz: you are acting discourteously google' but do note that this is a non-standard header that may change*".
*That's no joke. Read the link on their FAQ.
"As a User"
I have to say I was very impressed, all pages were loading faster and smoothly.
My first visit to WW after installation required me to log in and my surfing IP address was Googles IP address.
Although I was highly impressed with the results I had to un-install simply because of 'Privacy'.
The idea of sending uncrypted data, cookies and form submissions through Google won't help my 7 hours of sleep at night.
This is a great tool but saving a few seconds for the price of privacy is just not worth it.
It's not spyware due to the option of uninstalling; however I feel uneasy about potential attacks google may come under, plus the serving of ads based on the data I send to Google.
I prefer my ISP saving 3 months of my surfing habbits then Google caching every bit of data for an unknown period?
"As a webmaster"
I'll monitor the effects this may have but at this moment in time I'm not taking any action until we see more clear evidence of bandwidth abuse.
It will help google find our sites and index the content that is for sure, but if this tool grows to be popular we can't just block the IP due to the site being potentially un-available to the users.
So in the end we may well be forced to allow the users to surf our pages with the tool.
Overall:
Another great tool from the labs and I'm very impressed, however their is a BUT! A users privacy is the price, thats the price of free software and I won't use it for this reason.
I've gotten the impression that there's less emphasis on prefetching than on smart proxy caching including incremental changes to pages.
Ah that's alright then. If it's only a small a part of it that's, just disable the pre-fetching and everybody's happy!
[edited by: mrMister at 8:45 am (utc) on May 5, 2005]
It's not spyware due to the option of uninstalling
By this logic, Spam wouldn't be spam if it has an unsubscribe option.
How does this thing handle web applications where prefetching could essentially fire off an action within the application that was not desired?
You mean something like an admin system for your web site, where at the press of a button you can add, edit and delete pages.
Yes, Google WA was more than capable of "prefetching" these delete events for you, thus deleting pages on your web site for you!
From what I can tell though, Google seem to have taken steps to prevent this from happening now. The prefetch doesn't seem to be working in the same way as it did when I first analysed it.
<link rel="noprefetch" href="http://url/to/get/">?
I wish it were that simple
I can't believe google is going the Way of the "gator" like businesses
"Download our Free, Appealing but useless little App" and we will be "In Control" w'll know exactly what you do on the web and when, thus better targeting you with our ads down the line....
Google : You don't need to do this kind of Stuff, better continue to concentrate on Pure, Plain Web Search.
Second, it is both a Scraper, and Spyware, and a Proxy.
Third, it does not honor robots.txt and does not even specify an User-Agent string for that file.
Forth, it generates extra useless traffic and wastes your bandwith.
I have only two things to say about this
No, this is not negociable. If Google wants to have access, they will have to have a different view on users than Gator / Claria. There is absolutely nothing even remotely positive to say about this product
[edited by: claus at 11:19 am (utc) on May 5, 2005]
any quality proxy must pass the ua unchanged. Thus, it will be the same ua as your browser.
[showcase.netins.net...]
about 100 pages surfed: total saving 9 seconds. It still isn't as fast as raw Opera...
Has anyone been able to identify it via User Agent yet
As Brett stated it uses your browsers UA so it can't be blocked via UA string.
I think the only way to block it is through the IP range of 72.14.192.0 - 72.14.207.255 via .htaccess . I send them to a custom 403 on another domain (the .net,.info etc of the same domain name) in which I added one of the reasons they may have been blocked was by using the Google's webaccelerator.
As Brett stated it uses your browsers UAAs a Google employee was quoted as saying on SEW, it does use a Google User Agent.
See, this is part of what has me so concerned.
WebAccelerator could well be a fine product with benefits to users and webmasters alike.
But the constant stream of contradictory information makes me wonder why Google seems to be going out of its way to obfuscate these issues.
If someone gives me the impression they're hiding something I always assume the worst possible scenario. Am I alone in this?
With apologies to GoogleGuy, what has he really told us about WebAccelerator?
http://showcase.netins.net/web/phdss/environ2.htmabout 100 pages surfed: total saving 9 seconds. It still isn't as fast as raw Opera...
Yep, I still use the old Opera 6.06 and have yet to find a faster browser for a Windows based PC ;)
You need to show this thread to your colleagues as well:
[webmasterworld.com...]
Seems like Google Accellerator becomes a referral spam application when the prefetched page contains some sort of referral URI.
I have had 6 referrals from Webmasterworld to a website of mine since I posted the first time to this thread. I have never disclosed the website URI to Webmasterworld.
News Flash: Google hires hires lead Opera programmer. Ex-Mozilla programmers, re-join the Mozilla team :).
We might not be their target market though. AOL stil has some 20 million users, not to mention netzero, earthlink etc., who still use the modem.
Claus: Could you please post the complete (and most compact) required code to block this thing so mod_rewrite novices like me don't wipe out their traffic.
Could you also post the relevant text of your custom 403 so we can help Google users get the message.
Ta!
And rightfully so. Robots.txt is designed for spiders with automatic methods to determine which link to crawl next. With Google's Web Accelerator, however, each request is directly triggered by the behaviour of a human user.
So, it's a proxy. Google aren't the first in this field: my ISP has a transparent proxy, yours does too. AOL have been doing something like this since... well, forever.
Rather than worry about the proxy accessing your site, I think it would be more useful to look at the wider consequences of the proxy in terms of Google's operations, be it advertizing or search. Google is above all else a data-collection and data-mining company. So simply put, what does this new data bring?
Google have been recording search habits for years: cookies, toolbar, GDS... The proxy will give them a great dataset of browsing habits. A couple of examples. A proxy user searches for widgets, selects a result. Google can continue tracking the user via the proxy: do they find what they are looking for? How many pages do they visit on the target site? They could not easily follow that kind of activity before. After that comes the browsing cache itself: Google can use HTTP referrers to see what pages requested by users appear to be bookmarked (no referrer) rather than via a link. Google can also compare the cache made by the proxy to the cache made by Googlebot to see the difference. Google can find the "dark web" of pages not readily accessible via Googlebot as users are acting as a super-bot with full Javascript/cookies/referrers capability (and multiple user agent strings to boot).
Overall, with the proxy Google are getting a taste of the web from the user's perspective, rather than a bot's perspective. The user data is giving eyes to the bot. It could bring their search to a whole new level. It is the cloakers who are screaming loudest today, mostly about the (undoubtedly important) privacy concerns, but their real fear is coming from the massive breach in their technology that the accelerator is bringing.
More accurately, it is a compressing proxy.
Dialups do not benefit because modern 56k modems all compress data to/from their isp anyway. The rest of the web is still stuck in the uncompressed dark ages of the 70's. A compressing proxy does indeed stand to speed up the web. One thing we know for absolute certainty is that bandwidth requirements and page sizes are going to increase. I believe average page sizes will increase 3 fold in the next 5 years.
I agree in principle with claus's feelings on the product, but I disagree on some of the specifics:
> a made up consumer need
I do think there is a small justifiable (aka: excuse) need to speed up the web. In theory, the web should have been 100% compressed data years ago, but we are still living out that 1960-70's uncompressed legacy. Compressing data - makes sense for all the major isps, and Google is just muscling in on that territory itself.
> Second, it is both a Scraper, and Spyware, and a Proxy.
How is it a scraper? I see it as a human still at the kb.
Spyware? Agreed, but that is not new - it is just *more* of the same thing they have with the toolbar and all their sources of data now.
Proxy? So what?
> it does not honor robots.txt
It is a proxy on behalf of a human, it isn't a bot.
> User-Agent string for that file.
It can't claus, it *has* to pass the UA unfettered.
> wastes your bandwidth.
Agreed. how much it uses is open for debate.
I think the take away here is that if everyone would just install GZip on their websites, we would have the same effect in ALL browsers and not just in IE/Moz.
