i know i'm missing something, but i'm having a REALLY hard time getting spun up about Google Apps.
i get what this is. It's a way to run your app on Google, tie in to easily google services and a nifty way to deploy and monitor with scale. It's in python, which is fine, you get a tiny little sandbox to play in for free, which is also fine, but i'm really not seeing the earth shatter and a new world emerge from this.
Honestly, i'd love some enlightenment here.
This is what i'm getting so far. A developer signs up, pulls down the SDK and can start plugging code in. They get a protected (stripped down) version of Python which may or may not have things like input filtering for things like UTF-8 and XSS attacks, process some things and then dump back the data. Applications are hosted off of google, who's pretty much free to do with it as they see fit.
It's not that there aren't some clever bits in there. Going over the sdk libraries, the parts i thought were most interesting were the Data Abstraction bits (where you don't care where the data comes from) and the simplified User bits (where you don't care how the user is validated). Both of those are really nice, but don't require google at all. Frankly, a lot of the fervor and buzz i've been reading is more confusion with the sort of things that Python can provide vs. where google is doing any work.
i'm also a bit worried about the user access bits. Right now, you request a user to login, they do, and you get back an object containing their email address, nickname (which may or may not be an email address) and the auth domain. Yeppers, their email address. Good thing that gmail has good spam blocking, huh? i was tempted to build an app and have folks log in so that the div containing the last n visitors would be set to "display:block" instead of hidden.
The whole thing reminds me of GBase, google's effort to create a mass storage set. When it came out, folks were doing everything from predicting the end of Craigslist and Ebay to pointing out the dawn of a new digital age. How many services are you aware of that use GBase? (Tip: Google shopping does. If you want to list your data on Google shopping, you have to use GBase, so it's not really an optional element and may explain a reason why Google Shopping hasn't quite outpaced Amazon.
i was talking with someone today about this and i kept feeling that a lot of this is more pendulum swinging. (well, in this case, it's google wanting to own the pendulum) As things swing back and forth from Server Based to Client Based, folks are missing the actual win scenario here. i'm a big fan of keeping data cloud based and function client based. This… doesn't really solve that. It just gives folks a playground and makes things easier for Google when someone comes up with an idea worth acquiring or copying.
As it is, i'm just not as blown away by this as everyone else seems to be, and not because i don't like Google. i'd feel the same if this was a Yahoo product, or Mozilla, or BBC or anyone else. Honestly, that's my checkpoint for anything. If this was being built by Yoyodyne Industries, would i still be excited about it?
Right now, i'm not. i'll stick to working on my own server on my own code with my own data stores. Yeah, i'll probably hit scaling issues in the future and i'll have to address them. If i build my abstractions correctly though, my code won't care. i like to imagine i can do that.
Ok, that I can understand. I'm also sure that a great deal of the hyperbole and hyperventilation is mostly the Valley echo chamber.
Perhaps I should just remind myself of this.
Save This Page


Seems to me it's just a way to try a new app without worrying about operating costs and certain kinds of infrastructural issues, like setting up machines at a colo, etc. Who is saying it's anything more?