I've been doing a lot of thinking and writing (at work) about Web platforms and Web APIs recently. Just yesterday I was talking to some folks about Web APIs and an interesting question came up about the Google Web APIs. Basically we couldn't come up with good reasons for Google to have exposed them or for people to use them in their current form. I left the conversation wondering whether I just wasn't smart enough to see the brilliance of their strategy. Thanks to a post by Alex Bosworth's entitled Search Engine APIS it seems I may not be so dumb after all. He wrote

Recently, the Search engines have been trying to get into the mix, with new pushes to gain developer followers. Google recently launched a new subdomain emphasizing their committment to not being a technology black hole. Less recently, Yahoo launched their own developer site, in an attempt to build 3rd party value around their product.

However so far these 3rd party applications haven't found widespread appeal and had breakout success. This is in large part because of crippling artificial restrictions placed on Google/Yahoo/etc APIs.

It's not to say that applications made on platforms haven't been fairly cool ie: Yahoo News In Pictures, or clustering Yahoo search results.

But these will never progress beyond cool toys, and I'm not sure that Yahoo/Google/etc realize that. Talking to Google engineers, I was informed that Google thinks of their 3rd party API program as a complete failure and only a couple people have made anything vaguely useful from it. The reason why is they have no real committment to making these programs work. For example their terms and conditions are so draconian and laden with legalese there is no motive for developers to work with them.

This confirms some of the suspicions I had about their search engine API. Of course, this isn't to say I think providing programmatic access to search engine results is a bad idea. I just think Google's approach is very weird.

For example, I like the fact that I can get the results of a query on recent news items for terms such as "G-Unit Game beef" as an RSS feed from MSN Search.  Or that I can subscribe to the results of searching for my name on Feedster in my favorite RSS reader. As an end user I get a lot of use out of them and as a developer they are a cinch to work with.

There's a difference between cool apps and useful apps. The people at Google sometimes don't get the difference but the guys at Yahoo! definitely do. All you have to do is compare Yahoo! Maps to Google Maps to figure this out. Similarly I think the APIs provided by Google don't really enable people to actually build useful apps just cool little toys. MSN Search and Yahoo! seem to be on a better track because they enable more useful scenarios with their breadth of services exposed as APIs* as well as the fact that they provide these results as RSS.

What I've actually found surprising is that neither service puts sponsored search results in the results returned. I assume this is because we are still in the early days of search syndication but once it catches on and regular people are subscribing to search results so they can do things like build their own newspaper/dashboard then I'm sure we'll see the ads creep in.

* RSS is an API. Deal with it.


 

Categories: MSN | XML Web Services
Tracked by:
"modesto estreme" (modesto estreme) [Trackback]
"march 1976 playboy" (march 1976 playboy) [Trackback]
"revlon" (revlon) [Trackback]
"teen" (teen) [Trackback]
"big butt tight jeans" (big butt tight jeans) [Trackback]
"ore boat canadian challenger" (ore boat canadian challenger) [Trackback]
"dirty hip hop images" (dirty hip hop images) [Trackback]
"lesbian fondling teens" (lesbian fondling teens) [Trackback]
"megatettone" (megatettone) [Trackback]
"arredamento negozi abbigliamento" (arredamento negozi abbigliamento) [Trackback]
"posizione fare amore" (posizione fare amore) [Trackback]
"sleeping upskirt" (sleeping upskirt) [Trackback]
"sesso in macchina" (sesso in macchina) [Trackback]
"musica mp3 gratis" (musica mp3 gratis) [Trackback]
"figc cru" (figc cru) [Trackback]
"grandezza orgia a 3" (grandezza orgia a 3) [Trackback]
"rich bitch hates blackmen" (rich bitch hates blackmen) [Trackback]
"lee ryan" (lee ryan) [Trackback]
"boca raton beach hotels" (boca raton beach hotels) [Trackback]
"massage therapy schools in citrus heights" (massage therapy schools in citrus h... [Trackback]
"erotic dating sims" (erotic dating sims) [Trackback]
"prenotazione vacanza" (prenotazione vacanza) [Trackback]
"controllo di gestione" (controllo di gestione) [Trackback]
"legge fallimentare" (legge fallimentare) [Trackback]
"wild college girls flashing" (wild college girls flashing) [Trackback]
"amateur multiple facial shots" (amateur multiple facial shots) [Trackback]
"xxx audition" (xxx audition) [Trackback]
"female athletes posing" (female athletes posing) [Trackback]
"elk grove herald march 6%2c 2006" (elk grove herald march 6%2c 2006) [Trackback]
"hot bitches big asses tits" (hot bitches big asses tits) [Trackback]
"computer grafica" (computer grafica) [Trackback]
"bello fighette fotti" (bello fighette fotti) [Trackback]
"sindrome" (sindrome) [Trackback]
http://www.25hoursaday.com/weblog/PermaLink.aspx?guid=3d499c0b-d7b6-4970-9fa9-92... [Pingback]
http://silauma.info/washington/sitemap1.html [Pingback]
http://fwmwly7.net/florida/sitemap1.html [Pingback]
http://weujmru.net/community/sitemap1.html [Pingback]
http://issl7ti.net/ebay/sitemap1.html [Pingback]
http://eofw1lk.net/alaska/sitemap1.html [Pingback]
http://gator393.hostgator.com/~rocata/sitemap2.html [Pingback]
http://hrtfjhk.net/sitemap1.html [Pingback]
http://kiva.startlogic.com/sitemap1.html [Pingback]
http://host239.hostmonster.com/~blogford/sitemap4.html [Pingback]
http://gator413.hostgator.com/~digital/jewelry/sitemap1.html [Pingback]
http://vahq8px.net/lottery/sitemap1.html [Pingback]
http://gh9kwkn.net/flickr/sitemap1.php [Pingback]
http://hrxc1zr.net/housing/sitemap1.html [Pingback]
http://box432.bluehost.com/~zbloginf/sitemap1.html [Pingback]
http://gator442.hostgator.com/~hockteam/university/sitemap1.html [Pingback]

Thursday, June 23, 2005 6:21:36 PM (GMT Daylight Time, UTC+01:00)
One of the benefits I see for providing RSS capabilities to search is to allow for integrating a pre-defined search into your application. More specifically, think of how nice it would be to highlight a word in VS.NET and then click on SHIFT+F1 and now all of sudden you are searching google for that word on all of the predefined good developer sites. Now take this a step further and imagine you integrate Word to do the same thing.

Essentially, I don't see new applications being created that will use these API's. What I do see is integrating the applications we do already use to take advantage of our new connected universe.
Friday, June 24, 2005 1:56:46 AM (GMT Daylight Time, UTC+01:00)
I'm the author of the Argos search engine API (argos.dev.java.net), so I have some experience dealing with search engine APIs. From the plain Search API POV, RSS/Atom results would appear to be the way to go. Like you note, it enables a lots of useful applications. I can't understand why Yahoo didn't use RSS for their results, since their format and REST based interface is very close anyway. They actually do use RSS for their News search, which is useful.

Some of the other Yahoo APIs are useful, though, and a lot better than anything anyone else provides.
Friday, June 24, 2005 11:24:14 AM (GMT Daylight Time, UTC+01:00)
Wait. Google thinks their API is a failure because nobody has built a real application on top of them?

When you use that API, they cap you to a miserable 1,000 queries per day! Who's going to build a real application on top of an API that has such a low usage limit?

Details: http://www.google.com/apis/api_faq.html#gen7

When they launched their API back in 2002, I thought, "well, that's nice -- free access for toy apps, let's see what they end up charging to remove the cap." Only they never got around to that part. If they're surprised now that nobody has built a large-scale application, they're dumber than I would have thought...
Friday, June 24, 2005 4:29:58 PM (GMT Daylight Time, UTC+01:00)
The only useful Google API application that I've found is Google Alerts (http://www.googlealert.com/). It came out long before Google had its own alert service (http://www.google.com/alerts) and still has some features that the official version doesn't have, like RSS output.

I also heard of someone making an IM client that used the Google API's spellchecker to spellcheck messages (since it deals much better with real world/IM spelling), but I haven't seen it for myself.
Saturday, June 25, 2005 6:30:31 AM (GMT Daylight Time, UTC+01:00)
While I agree with the sentiment, I would like to point out that RSS fortunately is not really an API but a protocol and a dataformat. That makes it much more usable and powerful than an API.
Comments are closed.