ITF and the Google Web Accelerator
So I like Google, and I’m always interested when they come out with something new and cool. Hence, I jumped on the Google Web Accelerator bandwagon.
Then as a service to all of you I asked my OTA what his thoughts were on the matter:
On Google Web Accelerator: Google Blogoscoped has a good introduction to GWA.
One of my questions has been: Why is Google doing this? It takes a lot of bandwidth and storage space to run a proxy server for everyone on internet (granted, they aren’t at that point yet.). TechDirt has an excellent theory that makes a lot of sense: Google uses your surfing as their spider.
Right now, a search engine works by using an automated program, called a spider, to “crawl” the internet. As it goes, it collects information about each website it comes to. It ranks that website for different individual search terms using several factors, the two most prevalent being #1 other people linking to it and the #2 the content of the site. This spider tries to emulate a random surfer that starts at a random starting point and goes for a while and then eventually gets bored and stops.
It would, of course, be ideal if Google had real people, instead of computers, surfing randomly. With Google Web Accelerator, they have that. As you go to websites, they can index that site. They know how you got there, so they know what search terms are relevant for that site. They know that it is a popular site if a lot of their users are going there. It would be a very nice increase in the accuracy of their search results.
However, it has been having some problems. People are being logged into forums under other usernames (This is because everything is run through a proxy. The forum software sees the multiple users as a single user. When you log in to your account, it logs in for everyone.) and it is deleting emails by pre fetching (One of the ways that GWA speeds up your surfing is by looking at the links on the page that you are on and choosing the ones it thinks you might click on and then starting to download them even before you click on them. You know in your webmail how you have those little delete links? Well GWA was prefetching those too and in the process deleting your emails.).
Of course, why do you need an accelerator with a broadband connection? (GWA is only recommended for use with broadband.) Broadband is pretty snappy.
Google has stopped offering downloads of GWA because “We have currently reached our maximum capacity of users and are actively working to increase the number of users we can support.” Google Blogoscoped thinks that “This may or may not be a polite way of saying there were too many bug reports being thrown at them, and that they now need time to check them.”
There are always privacy concerns with this kind of thing.
I have an intense interest in all things Google, but I don’t have the knack for the kind of research Hans knocks out. This is some great stuff. The only issue I’d take up with him, and this is strictly opinion, is about the need for a web accelerator. I like the idea of speeding up anything that can be sped up, and you can never surf the web too fast!
For a long time a major criticism against Google has been their nebulous and ambiguous privacy policies and practices. It looks like the GWA just brings more of the same.
Hans finishes up with his official recommendation:
I have not been happy with GWA. If anything it has slowed down my surfing not sped it up. It has made some pages inaccessible. It still has some bugs. It has privacy concerns. I uninstalled it. I would suggest that you don’t install it, or if you have, uninstall it. The risk/problems to rewards ratio is just too high.”
So I sadly uninstalled it. I wanted it to be a good thing, but I managed to put aside my pie-in-the-sky ideals and do the right thing.
10 Responses
May 25th, 2005 at 12:28 pm
Did it seem to speed anything up for you, Tom?
May 25th, 2005 at 1:19 pm
This looks like the right place for a technical question. :)
I’m just getting into site feeds/newsreaders since I’m reading checking multiple blogs, but I have an issue with the whole deal. Hans, you have two feeds, one for posts and another for comments. By sorting your site in my reader by date, I can see posts and comments in chronological order. Simply splendid. But few people have two feeds, so if I want to check for comments, I have to log onto the site and peruse it the old-fashioned-way anyway. Does this mean as long as Tom doesn’t create a comment feed I have to check the entire page full of posts to see if there are any new comments? I’m sure others deal with this scenario too. Get on the site, no new posts, look for new comments, look at comment numbers, try to remember what they were last time I was here, open post, oops, no new ones there, next post, oh heres one, next post…
What are the options for making this ridiculously convenient newsreader a little more convenient? Oh us spoiled Americans…
May 25th, 2005 at 1:36 pm
I thought it did (I wanted it to!) but I never ran any scientific experiments to see if it really did.
I could be wrong, but I don’t think Blogger offers a feed for comments. Can you set one up some other way Hans?
May 25th, 2005 at 2:21 pm
Scientific experiments!! Maybe ’cause you’re not a scientist!! That was good Tom!! I still chuckling. :))))
May 27th, 2005 at 11:35 am
Lol Dave! I can definitely identify: Looking thru the posts, trying to remmeber how many comments there were last time, etc…
Unfortunately, Blogger does not provide comment feeds. My blogging platform, b2evolution, does. Tom, I think you should look at moving to a custom blogging platform. It wouldn’t be as difficult, because you have your site hosted somewhere besides Blogger’s servers. I would suggest, however, that you go with WordPress, a more advanced branch of the b2evo source code, if you do go that route. It does take a lot of work to switch, but now, being off for the week, would be the perfect time.
If you do go that route, I would be glad to step you thru it. It does take a lot of work… There is a plug-in that will import all your old blogger posts. (Hmmm… not sure about the comments…)
Anyway, Dave, hope that answers your question.
May 27th, 2005 at 11:58 am
I posted the wrong link for b2evo. Here is the corrected one: b2evolution
May 27th, 2005 at 12:24 pm
I know, I really should. I considered something like that at the very beginning, but I didn’t have an OTA then and it was a little intimidating. I thought I’d start a blog with Blogger once and see what happened.
I wish I would have done some other things different too, but that’s how you learn. The blog took off more than I really expected, I think. I probably will make the switch at some point in the future, but now my week is about over!
Believe me, I know what you guys are talking about with keeping up with comments. I’ve done/I do that a lot. But with ITF the comments get emailed to me! Much nicer.
May 31st, 2005 at 9:03 pm
I would like to point out to your readers that the deleted e-mails were a cause of poorly-designed web site, and were not solely the fault of Google. The HTTP specifications (RFC 2616) specifically states:
[Quote]
Implementors should be aware that the software represents the user in their interactions over the Internet, and should be careful to allow the user to be aware of any actions they might take which may have an unexpected significance to themselves or others.
In particular, the convention has been established that the GET and HEAD methods [“links” in the vernacular*] SHOULD NOT** have the significance of taking an action other than retrieval. These methods ought to be considered “safe”. This allows user agents to represent other methods, such as POST, PUT and DELETE, [“buttons” in the vernacular*] in a special way, so that the user is made aware of the fact that a possibly unsafe action is being requested.
Naturally, it is not possible to ensure that the server does not generate side-effects as a result of performing a GET request; in fact, some dynamic resources consider that a feature. The important distinction here is that the user did not request the side-effects, so therefore cannot be held accountable for them.
[/Quote]
Although any software that pre-fetches should consider issues such as these, using a link to delete an e-mail is inferior to using a button. Unfortunately, a large number of web sites do not carefully consider these issues, which makes it particularly difficult for Google and other caching software to work around this issue.
-“Anonymous Coward”
* For all the technical gurus, yes, “links” and “buttons” are very loose approximations.
** This phrase, or the phrase “NOT RECOMMENDED” mean that there may exist valid reasons in particular circumstances when the particular behavior is acceptable or even useful, but the full implications should be understood and the case carefully weighed before implementing any behavior described with this label. –RFC 2119
May 31st, 2005 at 11:08 pm
Thanks for that, AC. Very interesting, and makes a lot of sense.
June 1st, 2005 at 11:34 am
AC: I understand what you are saying and that same (very valid) explanation is all over the internet. I am aware that many web developers write very sloppy code, but if it is a common (wrong) practice, Google needs to do something about it. Google has tons of super-bright people. They can figure something out.
My “no” recommendation is based on multiple factors, prefetching deletion only being one of them:
#1 No noticeable increases in speed.
#2 Breaks some session cookie based apps
#3 Privacy
#4 Yet another app on your machine
#5 It’s still in beta; the casual user should wait*
#6 It breaks (prefetch deletion) many badly coded web apps (I know it’s the developers of the apps’ fault, but it really doesn’t matter who’s fault it is: if it works without GWA and doesn’t work with GWA, you have to do what’s best for you)
* I know. Google has lots of stuff in beta for a long time. Nevertheless, this app feels particularly immature to me. In this case the beta title is more deserved than, say, Froogle.