Written by Emre Sokullu and edited by Richard MacManus
OpenID, RapLeaf, Amazon S3, Amazon
EC2, Undisposable. All of these services have
one thing in common – they make web development more granular. By granular, we mean they
are making web development componentized and outsourceable. So as a web developer, you
can outsource things like authentication, email check or trust system. This brings 2 main
- Decreased development time and cost
- Power of Masses
Below, we take a closer look at the major granular service providers so that you’ll
better understand the concept and its advantages:
OpenID, as we mentioned in a previous article, is
a distributed authentication mechanism. It allows you to outsource your whole
authentication system to this distributed platform. This results in 2 things:
- You don’t need to develop it on your own (decreased development time and cost)
- Moreover, you decrease the sign up threshold and so attract more users (Power of
Sounds like a good deal, no? But the numbers are not supporting this goal. Even though
OpenID is a widely covered topic in blogosphere, and is backed by big companies like
VeriSign and SixApart and has many fans and the open source community behind, the
adoption and usage rates have remained low.
In the same way, RapLeaf provides a distributed trust system for your site. What made
eBay so successful was the trust system it brought into the online auction arena.
Similarly, many other ventures need trust systems to make transactions between their
visitors more secure. And making your trust portable across all these sites is a very
bright idea – which is yet another advantage of power of masses in granular
In the same way, outsourcing this system to a 3rd party can result in decreased
development time and costs. Although the advantages are obvious, I called the founder Auren Hoffman last week and tried to learn more. He gave
me 2 success examples: SwapThing.com and PetLovers.com. That same night, I had the chance to meet
the SwapThing developers in the Stirr event. They said the biggest advantage of RapLeaf
for them is the portability of the trust system. This is very understandable for a newly
founded site – which wants to stand on the shoulders of other similar startups.
Nevertheless, the usage rate of RapLeaf still seems to be low.
Amazon S3 is all about distributed storage. S3 saves you from the hassles and big
costs of buying and maintaining storage hardware for your site’s needs. YouOS is one example that secures its data on scalable Amazon
S3 servers. You can see if a web site is using S3 or not by keeping track of your status
bar while a page loads; S3 powered sites will frequently fetch data from
http://aws.amazon.com address. R/WW’s MyBlogLog for example uses S3.
Even though the advantages are so obvious and the company that offers it is the well
trusted Amazon, my personal experience tells me that S3 usage is not very large either
Similarly to S3, EC2 (Electronic Cloud 2) is Amazon’s distributed computing power
system. Considering the fact that it’s still a beta service and its implementation is not
that easy, the low usage is understandable. But when it gets ready, Amazon EC2 will
become a compelling service that ends all the hassles of maintaining clusters and
Undisposable.org allows you to outsource email
validation against disposable email addressing (e.g. 10minutemail.com, mailinator.com) and fake accounts spread from the likes
of bugmenot.com. Needless to say, the usage rates are
Why are the implementations are low?
Now it’s time to answer the big question, why are these granular services not being
used as much as they could/should? In my opinion, there are several reasons:
- Transfer of Assets: Actually in all these services, you don’t only transfer
the hassles and extra costs, but also transfer the assets that make your site so
valuable. In the case of RapLeaf for instance, trust has always been the biggest asset of
eBay. That’s why eBay prohibited the use of RapLeaf throughout their site. And that’s why
many web developers may be suspicious about transferring their core assets to RapLeaf,
unless they think trust is not a primary asset of their site.
- Uncertainty of Services: Site owners are uncertain of the granular services we
pointed out above. They can’t trust startups and even big established companies like
Amazon; because it’s obvious that even Amazon may get out of this business if they see no
profits and no opportunities. The availability as well is a big question mark for web
developers. Because relying on granularity means your visitors will be affected by the
outages of the sites you depend on. In the end, in granular dependent sites, even small
outages add up successively.
- Implementation Threshold: Implementation is often not so easy. Web developers
need to learn new patterns and APIs to start experimenting with these services. And
considering that the points above create question marks in your mind, would you neglect
this threshold or would you keep going?
We may have portrayed the granularity of web 2.0 “a little” dark. But in our opinion,
granularity may leap (should we call it web 5.0 after
this and this?) after it
solves the aforementioned problems and creates some trust in developers. One possible
solution would be to clearly state the conditions and availability of these services from
the homepage. Maybe a Creative Commons
style licensing organization may need to arise to support this. Also clearly state that
data will always be open for import/export.
Amazon is working hard on the granular web with their new AWS services and the
seminars they organize. They seem to be the flagship in spreading the Granularity trend.
Please let us know of other web services that may be categorized as Granular.