Will the AppEngine change the landscape?

[Waleed]

Hi everyone,
I'm trying to get feedback for an article I'm writing and also for
my own understanding. I believe that the AppEngine is a huge paradigm
shift that is underrated and not appreciated enough by most people. It
seems to me that this is going to change the industry in many ways and
creates a shift in the balance of power away from big companies and
more to individual developers and small startups just as Open Source
did before it.

I see two areas where the AppEngine will stir things up:

1. Free personal hosting. If you don't need more than 5M pages a
month, then you don't need to pay to host your blog if you can get the
same for free on the AppEngine without the restrictions that the free
hosting options like wordpress.com and blogger.com impose.

2. The disappearance of the Scaling advantage. When it comes to apps
that require massive scaling (ad networks, widgets that are
dynamically generated for each user, analytics, ...etc), it's possible
now for one person to complete with companies that used to have a huge
advantage because of expensive infrastructure.


What companies do you think will be in trouble as new competitors
emerge that innovate faster because they can outsource the scaling to
Google? What are the new business opportunities that you see coming?
And, finally, what products and open-source software are you aware of
that are being developed that you think could have a great impact
going forward?

Thanks,
Waleed

[g-man]

Hi Waleed,

So you might have noticed there hasn't been much interest in your
request, not like when someone asks what your favorite programming
language is! (unless you're getting private messages)

Most of the people here are technically-orientated programmers, who
are grappling with the nitty-gritty details of how to work with this
new technology.

I, on the other hand, am strictly a hobbyist, so I do like to reflect
on the changes possible with the Google App Engine is producing. There
is one other member of this board, Garrett at:

http://gawsh.appspot.com/blog/List.do

...who has also written about the implications of this new technology.

As to which companies will be affected by this change, I wouldn't want
to speculate on any specifics, but I say your initial points were
correct.

In my case, Google has removed all the hassles I was faced with when
contemplating setting up a web app: I would have to find a host,
manage my data (backups, security, etc), create an infrastructure to
run my app (Ruby on Rails, for instance), institute user
authentication and virus screening, then worry about growth and server
capacity, downtime, etc.

The App Engine removes all that headache from me.

This was all covered in the Google I-O videos, and in the introductory
video where Google gave its rationale for opening up the datastore via
the App Engine.

Keep in touch...

My learning app is at:

http://archi-checker.appspot.com

[luismgz]

I believe it's very simple:
When I begun using internet, many years ago, I had to pay a provider
to get my email address.
Then, many free services came up and now it is very common to have one
or more yahoo, hotmail or gmail accounts.

If you wanted a personal blog, you had to pay for hosting and probably
an special software.
Now the common thing is to get a blogger account or any other similar,
so anyone can have a blog in just minutes.

So I don't see why it would be different with paid hostings and
appengine.
Although I believe that the paradigm shift beween RDBMS and BigTable
will be the main concern for many people looking at appengine.

luis

[Tom Kermode]

"It seems to me that this is going to change the industry in many ways and
creates a shift in the balance of power away from big companies"

Instead of moving power away from big companies and towards
individuals, I think it's moving power away from big companies to an
even bigger one - Google.

In the very early days of computing, if you wanted to write a program
you'd have to build the hardware and start from scratch. Eventually,
this moved towards buying a PC with an OS pre-installed and writing
the program in a higher level language. I think GAE is analogous to
an OS for the web. You don't need to worry so much about the lower
level stuff any more.

When writing traditional code, an offline program, you can do it
better if you understand the hardware and OS you're running on, but
you can still manage to write simple programs without much knowledge
of the supporting architecture. This is probably going to be the same
on GAE, where it's easy to write simple apps, but the more time you
put into understanding the depths of the system, the easier it will be
to develop larger applications successfully.

So Google have shallowed the learning curve for developing web
applications, much like Microsoft made it easier to develop desktop
applications when they released Windows. Microsoft became ubiquitous
on the desktop much as Google are becoming on the web.

Cheers,
Tom

2008/7/15 Waleed <waleed....@gmail.com>:

[Yagiz Erkan]

Hi Waleed,

At the moment I'm here as a hobbyist, so I'm not developing a
commercial application. Furthermore, we're not using the GAE in a
project where I work. Needless to say, I'm here to have a feeling of
the GAE from a first hand experience rather than just reading about
it.

Personally, I'm not so paranoid about the uber-controlling-Google
scenario. My first question was "what's in it for me?" and so far I
like what I've seen. If I were managing a start-up developing the next
killer Internet app, I'd seriously consider using the GAE. It may not
be ready for the prime time yet but I don't doubt that Google is
working hard on ironing the few creases out (SSL support, being able
to delete a project, etc.). I'm not sure if Google is going to keep
their pricing strategy the same but as it is, I think it's very
appealing not to pay (much) if you're not making (much) money.

As a developer who had never worked on Python before, I hesitated for
a quick while to jump on it. After all, it's not "that" easy to really
master a new language. But so far so good. I'm still alive :).

Regards,

- Yagiz -
http://blog.decaresystems.ie

[Michael Angerman]

All the points so far are right on target and I thought I would throw my 2 cents in...

From my personal point of view this technology and the concept is incredible..

Amazon started the ball rolling with their back end storage service and Google
is taking the concept to the next level...

The really big technology paradigm shift is the move away from the Relational Database Model
to the Datastore/BigTable...

Another interesting side note here which I think is pretty huge is the rapid adoption
now of Python...  Prior to GAE, programmers really didn't need to learn Python unless
they were working for someone who was already using it...

Now -- because its the only game in town at the moment -- lots and lots of programmers HESITANTLY (myself included) took the dive into Python land...

And I believe we are all being surprised by how nice Python really is --- and liking it too...

With that said -- Yahoo is rolling out their own Python project

The BOSS (Build Your Own Search Service) Mashup Framework is an experimental python library that provides developers with tools for mashing up the BOSS API with other third-party data sources...

So here you have the two largest internet companies in the WORLD both using Python as a way for end users like ourselves to program to their interfaces...

Its actually amazing to see this happen -- and with Guido Van Rossum now a full time Google Employee it makes even more sense...

Michael I Angerman
Albuquerque, New Mexico
http://www.zrato.com

[Sjors]

Waleed wrote:

>  What companies do you think will be in trouble as new competitors
> emerge that innovate faster because they can outsource the scaling to
> Google? What are the new business opportunities that you see coming?
> And, finally, what products and open-source software are you aware of
> that are being developed that you think could have a great impact
> going forward?

Google itself for starters. Although this situation is very far away,
the App Engine is yet another big step to make it easier to start your
own Google.

But there is one obstacles that still stands in the way: the price of
data and raw processing power. The App Engine does not deal with
that.

Google distributes the ballast of storage and CPU power quite
efficiently so they might sell it cheaper than your average hosting
provider, but only by so much. They still need to buy hard disks and
CPU's to run your program and these things need big expensive ware
housing, cooling and maintenance.

There is a place where you can get incredibly cheap hard disk space
and CPU power though: it is all these under used home computers and
cell phones out there. The Berkley Open Infrastructure for Network
Computing, BOINC, (think seti@home) takes advantage of this and it
allows them to get access to very cheap data and CPU power.

So as far as I am concerned, the next real big step will be the App
Engine (or something similar) running in the big cloud. Add the words
"federate" and "open source" and we are looking at something very
powerful, i.e. everyone can start their own Google (well it is not as
simple as that of course).

Sjors

[Sjors]

> There is a place where you can get incredibly cheap hard disk space
> and CPU power though: it is all these under used home computers and
> cell phones out there. The Berkley Open Infrastructure for Network
> Computing, BOINC, (think seti@home) takes advantage of this and it
> allows them to get access to very cheap data and CPU power.

And with "incredibly cheap" I don't mean free; I am thinking about
bandwidth and CPU auctions here, so people have an economic incentive
to install that little plugin. Let the good times start...

Sjors

[ready...@gmail.com]

There are a lot of good posts here, but I think there are a few things
to add:

- Ramp up and scalability for the masses is awesome; it'll open the
market up for a lot of small players, and it'll allow a lot of big
players to focus on development rather than infrastructure. So while
the little guy is the big winner, well, so is the big guy, if they're
smart.

- Private datacenters won't go away. If you're doing work for a bank
or any government, "the cloud" isn't good enough. They want to see
the locations where their data is sitting, if said data carries with
it any security, monetary or legal/privacy weight.

- The big winner will be--big surprise--Google! But not the way most
people think. Even if they never make a dime off the service, they
are setting themselves up for huge gains in two other ways...

1- They make money off of advertising which is placed in their
applications. Most of their applications come out of the 20% "free"
time their engineers get, but this means that they don't have to limit
themselves to just their own engineers. They've got thousands of
people developing new apps on their infrastructure for free! It used
to be that if they saw a good idea somewhere and purchased it from the
author, they had to port it to their infrastructure. Now, not only do
they get first crack at it (they'll see the popular ones and check
them out quickly), but if they do make a deal with the author, there's
no work porting it to their infrastructure. Bada bing, bada boom.

Sure, lots of AppEngine authors are going to put Google Ads on their
apps, and there's a win there, too, but for the really great ones,
this gives Google the first view of thousands of apps that might
potentially be for sale.

2- Recruiting. There's a war for awesome talent out there; a good
engineer can pave the way for earnings that are ten times his or her
salary. A great one can do ten times better, and an awesome one, in
the right environment can earn thousands of times his or her salary
for the company. Just as Google now gets first crack at all the apps
that might be for sale, they have representative work and contact
information for thousands of software engineers. If they are
impressed by someone's work, all it takes is a quick email or phone
call, and they might be hired.

Forget American Idol, people. AppEngine is "reality programming".
Build something great and you might sell it to Google for a few
thousand dollars. Build something awesome with a team of your best
friends, and you might be the next addition to the Google family.

On Jul 15, 3:19 pm, Waleed <waleed.abdu...@gmail.com> wrote:

[Ben the Indefatigable]

I agree that GAE is revolutionary. My blog post on this:
http://benpoint.com/googleappengine.htm