The don't be Evil Empire?

December 24, 2006 7:44 PM

people aren't happy about Google canceling it's API and replacing it with an AJAX search widget you're supposed to put on your page (and one that's deliberately obfuscated and difficult to extract data from). Google's obsession with secrecy isn't going unnoticed, and some employees are starting to feel like the magic is gone and they're being treated like children. Oh, and just to keep things in perspective Microsoft just filed a patent on RSS
posted by delmoi (35 comments total)

Er, the "obsession with secrecy" link should point here, sorry.
posted by delmoi at 7:45 PM on December 24, 2006

I love FPPs with a punch line. Oh Microsoft, will your antics ever end?
posted by Alex404 at 7:48 PM on December 24, 2006

Google's obsession with secrecy... (corrected link)

Though the page "Not Found" it went to made me laugh.
posted by hal9k at 7:57 PM on December 24, 2006

I will preview before posting. SPANK! I will preview before posting.
posted by hal9k at 8:01 PM on December 24, 2006

It's misleading to point to Aaron Swartz's thing saying "employees are starting to feel...". Aaron Swartz is not a Google employee.
posted by Nelson at 8:01 PM on December 24, 2006

I think google will uncancel the soap api.
posted by empath at 8:08 PM on December 24, 2006

I think google will uncancel the soap api.

I sure hope they do. I use it to provide a free, quality search engine for my department's web site. I'd hate to have to use any of the free-and-crappy server-side alternatives.
posted by Blazecock Pileon at 8:18 PM on December 24, 2006

one that's deliberately obfuscated

There are technical reasons why the code might not be readable. The code might be compressed to improve download time or the code might be compiled from another language as the referenced page suggests.
posted by bravada at 8:32 PM on December 24, 2006

It sounds like they're starting to reign in free ranging api access in the maps area too. Oh well, MSN has a search API still, doesn't it? I think yahoo does aswell...

Are there others?

If you wanna see something kinda funny, check this out: gigablast.com. They've got an XML API as well... not as terrible as you might think.
posted by ph00dz at 8:35 PM on December 24, 2006

The API never produced accurate results for me - seriously. I found it easy to use, but when I ran searches it gave me back incorrect SERP rankings for domain names. I can only think the API was checking one particular data centre and that centre produced vastly different results to the centres my own manual searches were interrogating (via google.com).

Ultimately, the API did not work for me.

Sad to say, but screen-scraping is the most accurate way to find ranking positions in Google SERPs - just scrape a page set to display the top 100 results and you can at least determine if your website is in the top 100 for a particular search. I don't use any more of Google's bandwidth than if I did a manual search.

I suppose there might be some posters here who think screen-scraping in anyway is unethical. Well, consider that Google's entire business is based on scraping (as all search engines are), and that the vast majority of website owners haven't even heard of the robots.txt file, let alone know how to stop Google's bots from scraping their site (and using their bandwidth).....
posted by FieldingGoodney at 9:05 PM on December 24, 2006

AfterMASH.
posted by dhartung at 9:13 PM on December 24, 2006

> I'd hate to have to use any of the free-and-crappy server-side alternatives.

Or be compelled to shell out for a Google box.
posted by ardgedee at 9:13 PM on December 24, 2006

Or maybe they deprecated it because SOAP is a magical pile of ass shavings. Seriously, it's CORBA for web nerds who missed out on the first suck-go-round. I wouldn't want to maintain a SOAP API either, and this is coming from somebody who has to code for the rapidly-whirling shit tornadoes that pass for PDA application platforms nowadays.

There are plenty of things wrong with Google, but that's because it's a large enterprise, and nothing works perfectly in a large enterprise. This seems like a good decision to me, and the worst complaint I can see here is that they decided not to maintain an avowedly experimental API.

and frankly, the only response I have to that complaint is "my, you haven't been developing software for very long, have you?"
posted by phooky at 9:20 PM on December 24, 2006 [1 favorite]

Or be compelled to shell out for a Google box.

I'd be happy to buy one if it was cheaper. As it is we don't deal with that number of documents to justify the hardware cost of a dedicated appliance (or the associated maintenance costs).
posted by Blazecock Pileon at 9:27 PM on December 24, 2006

well, this makes for a pleasant news day. i always love it when corporations file patents on Concepts other perple have designed.
posted by localhuman at 9:29 PM on December 24, 2006

Yeah, but the point is there isn't a good API for it at all now. It's not that they dropped a crappy experimental interface, they dropped the only interface.

This is why I use Yahoo's APIs for the app I work on. They're not amazing, but they're generally good. I'm pretty sure you can get JSON back, too, which is Awesome.
posted by heresiarch at 9:29 PM on December 24, 2006

Or maybe they deprecated it because SOAP is a magical pile of ass shavings. Seriously, it's CORBA for web nerds who missed out on the first suck-go-round. I wouldn't want to maintain a SOAP API either, and this is coming from somebody who has to code for the rapidly-whirling shit tornadoes that pass for PDA application platforms nowadays.

Having worked for a year to clean up and finish an 'almost complete' enterprise system built around crappy SOAP code, I agree wholeheartedly.

I'm sure it's been used well. But sweet Jesus, it's a kludge of pain.
posted by verb at 9:33 PM on December 24, 2006

I call bullshit on the "Microsoft patenting RSS" story. If you read past the first couple pages before making your frothy-mouthed blog post, it becomes obvious that the only thing being patented is the system-wide RSS API in Vista.

For fuck sake internet, learn how to read patents.
posted by cillit bang at 9:36 PM on December 24, 2006

Has no one been paying attention to gData? I feel like I'm the only one that's seen it. Look its RESTful! It's over Atom and (somewhat) RSS!
posted by jmhodges at 9:50 PM on December 24, 2006

Sorry, "GData".
posted by jmhodges at 9:50 PM on December 24, 2006

Or maybe they deprecated it because SOAP is a magical pile of ass shavings. Seriously, it's CORBA for web nerds who missed out on the first suck-go-round.

I've seen the SOAP spec, and I agree. But it's being removed and replaced with nothing.
posted by delmoi at 10:09 PM on December 24, 2006

Let me see if I have this straight: Google used to give away a service which consumed bandwidth and CPU time on their server farm, for which Google got nothing whatever in return: no direct revenue, no advertising exposure, nada.

Now they've decided they're tired of doing that. They're still permitting people to put Google search on non-Google web pages, but Google wants to attach advertising to it so that Google can get some revenue to offset the expense of offering that service.

Exactly why is it that this is evil?
posted by Steven C. Den Beste at 10:39 PM on December 24, 2006

someone needs to calm down a bit and stop trying to be the #1 tiger on metafilter
posted by dminor at 11:42 PM on December 24, 2006

SOAP does not need to be difficult, at least in the case of the Google API.

The developer of the Google API posted to this thread. Guess who it is.
posted by bravada at 11:49 PM on December 24, 2006

Well theres always EvilAPI for all your screen scraping needs. Its brand new, but it'l probably be mostly useful when google decides to drop the SOAP altogether.
posted by Mach5 at 8:21 AM on December 25, 2006

Nelson: "It's misleading to point to Aaron Swartz's thing saying "employees are starting to feel...". Aaron Swartz is not a Google employee."

And the Google employees who do respond to his article tell a very different story.
posted by PeterMcDermott at 8:31 AM on December 25, 2006

Aaron's postings have gotten a bit odd lately…
posted by blasdelf at 8:36 AM on December 25, 2006

Lately?
posted by IshmaelGraves at 9:12 AM on December 25, 2006

Exactly why is it that this is evil?

Those services are not remotely comparable.
posted by dhartung at 12:17 PM on December 25, 2006

As an aside, I find it strange that people are (or were-- no one talks about it anymore) concerned that the NSA is spying on us, but comparatively fewer people care that Google has an astounding amount of data on you, even if you never use their search engine.

I'm not tech savvy, but it seems obvious that if you visit a webpage that has adsense, then google knows your IP, from where you came, and where you go next.

I think we should be much more concerned about a "Big Brother" government, and more about corporate feudalism.

How inconceivable is it that Google, Yahoo, etc, releases Senator Brownback's search history, or visited sites? Or the opposite in Verizon's case? And God forbind anyone in government use a gmail account-- or email to someone esle with a gmail account.

I mention this because of the anger against canceling API. Google was giving it away for free, for God's sake.
posted by TheLastPsychiatrist at 1:06 PM on December 25, 2006

> Oh, and just to keep things in perspective Microsoft just filed a patent on RSS


No, they didn't. But nice FUD.
posted by Ayn Marx at 1:12 PM on December 25, 2006

Reading the comments to this post by Aaron Swartz, I'm pretty amused. There seems to be a point at which his post is forwarded around the company a bit, which results in some really choice comments:

Apparently our secrecy does keep the general public clueless about Googles interworkings. You have no idea how great it is to be integrated like you become at Google. To sit down at a table and eat lunch with people who are not only brilliant, but share your interests. Its really too bad we cant say much more because of NDA's but oh well. Im happy with my job and you can just keep your speculations.#

I think I'm okay with not being "integrated". I didn't quote the condescending "you are wrong" and "[$6k/month] entry-level programmer's salary" comments, but they're lovely too.
posted by blacklite at 2:29 PM on December 25, 2006

cillit bang: For fuck sake internet, learn how to read patents.

Here's claim 1 of the first application:

1. A computer-implemented method comprising:
presenting a user interface having a subscription control to enable a user to subscribe to a feed;
receiving, via the user interface, a user selection of the subscription control; and
responsive to receiving the user's selection, initiating a feed subscription process.

I haven't read the entire application, but isn't this pretty much what an RSS feeder (or whatever you call them) does?

On the other hand, this application does NOT show that the patent system is broken. One thing that most people don't understand is that it's just a patent APPLICATION. It is not yet a granted patent. The application, in its current form, will most likely be rejected, because the claimed subject matter most certainly isn't new. People who don't understand the difference between an application and a granted patent, like this Pete Cashmore dude, shouldn't be allowed to comment on patent issues.
posted by sour cream at 11:36 AM on December 26, 2006

I haven't read the entire application, but isn't this pretty much what an RSS feeder (or whatever you call them) does?

Yes. And you'll notice that the rest of the application refers to RSS as a pre-existing technology. That should be enough to show that Microsoft has no intention of claiming RSS for itself, and that taking Claim 1 literally is the wrong interpretation.

My understanding is that the terms used in Claim 1 can only be interpreted in the context of the description section. For example, it only covers a "method" as set out by the description, not all possible methods.
posted by cillit bang at 1:19 PM on December 26, 2006

...taking Claim 1 literally is the wrong interpretation.

I am not a US patent attorney, but my understanding of US patent law is that to ascertain the scope of the claims, you actually do start by taking them literally. Only if there is doubt about the precise meaning of terms, then you look at the description next. When there is still doubt about the meaning of terms after having looked at the description, you go for dictionaries etc. (extrinsic evidence).
In the case at hand, claim 1 appears to be clear enough and I'd be surprised if it isn't intended to mean what it appears to be saying.
In other words, it looks like claim 1 is very broad and shouldn't be granted (because it appears to be anticipated by RSS software that has been around for years).
posted by sour cream at 2:54 PM on December 26, 2006