Wednesday 26th October 2005
It’s not the technical details
Adam posts about Google’s web accelerator, and gives a fairly simple explaination about what’s going on. Or even more simply, because the web accelerator “clicks” on everything on your behalf, when you log into a website, google might click on things you rather it didn’t. Like logout, or delete, or reset, or so on.
The argument is now going back and forth on technical points (like RFCs), but that’s missing the point. Basically, Google are acting like a prick. They’re doing something that they know is going to, rightly or wrongly, cause a lot of trouble. Hence they’re being pricks. Plain and simple. I don’t care whether indeed what they are doing is right or wrong, what the interpretation of “SHOULD NOT” might be. But the fact that the arguments are about details like that is just an indication that the argument is occuring between techies who lack basic social skills.
I want to upgrade all the machines at work to SP2. If there’s spyware on the machine, the upgrade fails and screws the machine up. There shouldn’t be spyware on the machine. So technically, I should just upgrade all the machines, and blame the clients for their spyware problem if everything screws up. But I don’t, because I’m not a prick. So we’re working at getting rid of the spyware first.
A non-socially retarded policy by Google would be to assume that logged-in applications aren’t web-accelerator safe, since many of them aren’t. Then they could examine some open source ones, help identify and fix the broken ones, and declare them “WA-approved”. The application could then stick a tag in the html to encourage accelerators to crawl their app. Closed source vendors could use the seal of approval as a marketing ploy - “Get our new, fantastic, version 6.5 of Another Web App - WA-approved”. And then the end game will occur - most web apps will be accelerator safe, and everyone will have time to have throw some shrimps on the barbeque and drink beer. And it’ll happen without any of the bad feelings that are being generated by some socially inept techies.
The story about the website was important and the message was this: The public web used to be this way. The same arguments happened there and all the search engines were pricks and they crawled anything.
Could they have been nice and waited for sites to opt in for crawling? Sure, but you wouldn’t have search engines today. The feedback loop (search is useful which motivates sites to add themselves to search, which makes it so useful…) would have been very slow and you would have had a polite web which was mostly useless.
GWA has causes far more blog inches than real effects. It’s far from the only program which does that sort of thing and the others have more installs. But its simple existence is causing people to change. Sure, they’re repeating themselves ten years later about how terrible search engines^W^W web accelerators are, but (very) slowly does it and by the time GWA launches it should be safe.
Comment by Adam — 26/10/2005 @ 2:54 pm
One difference is that you would expect crawlers in public areas, but not in private ones (unless you invite crawlers into your restricted areas specifically, which would be weird).
And since when were blog inches supposed to correlate with real importance?
Comment by Andy — 27/10/2005 @ 9:19 am