Are web services just wishful thinking?

[Mike Mitchell]

I read in Saturday's "The Independent" that a survey by the Cambridge
Business School has showed that only 4% of mobile phone users intend
to shop on the internet (and, yes, they do spell it with a lower case
i). This figure is down from the 12% recorded just a few months ago.
Also, video on your mobile phone? Forget it! Vodafone reckons they
would need to be the size of bricks to accommodate all of the 3G
technology. Furthermore, although GPRS offers speeds of up to 30 kbps,
apparently not many consumers are interested. The kids prefer the
older GSM phones with SMS, and it's the kids who are supplying the
profits. They like the small size and there's not much to show that
they want either video or on-line shopping. Hardly surprising, when
you see kids in their hundreds of thousands congregating in shopping
malls all over the world, instead of crouched over their PCs at home,
feverishly ordering stuff. So, without effective technology plus the
consumers to use it, where will web services - and .NET - score? Just
how will this "solution in search of a problem" actually make profits?


Except for Microsoft, that is.

(Hint: Actually, the whole "web services" thing is just a big "smoke
and mirrors" idea to get you to upgrade to and use XP, Passport,
Hailstorm, subscription-based charging [the real biggie] and product
activation.)

MM

[Robert Scoble]

> Just
> how will this "solution in search of a problem" actually make profits?

You're late to the party. I wrote about this on August 28 at:

http://scobleizer.manilasites.com/discuss/msgReader$311

Here's what I wrote there:

+++++++++++++++++
Web Services are overhyped.

This is a common theme with our industry. Remember when Java was overhyped?
Remember when Push was? Remember when the Macintosh was?

Whenever something is overhyped, I start looking elsewhere for the real
action.

The real opportunity for the average Joe developer (financially) is not by
building Web services, it's by building business information services behind
the firewall. I think we should call services that run behind the firewall
"HTTP Business Services" since my mom and dad can't see them from the World
Wide Web.

Most of the programmers who use Microsoft stuff (or even Sun's Java) are
building business applications. The kinds you never hear about. They are
ripe candidates to build new kinds of business information systems.

For instance, KGO Radio in San Francisco is being run by a Visual Basic app.
It works. It helps them run their business. It was probably run by a
developer who didn't really care that his app is helping Microsoft sell a
few copies of Windows. The developer probably got paid a good fee to create
the app. He or she is happy.

Those are the kinds of folks that are using .NET. Will the next version of
Quake be written in a .NET language? Of course not. The next version of
RedHat Linux? Of course not. The next version of UserLand Frontier? Of
course not. The great folks who write those kinds of apps, platforms, and
OS's aren't the average Joe developer that Microsoft cares about (if they
did, a stretch limo would appear, and they'd buy you, just like they did to
Anders Hejlsberg when he worked at Borland).

Why do I believe that HTTP Business Services are underhyped?

Jon Rauschenberger, one of the guys at Clarity Consulting, and one of the
world's formost .NET experts, tells me that most of his .NET work is "behind
the corporate firewall." Jon is the guy seen speaking at VSLive! and TechEd
and PDC conferences and you better believe he's excited by the Web service
vision of .NET. He's one of the "Microsoft insiders." He writes SOAP code in
his sleep. If he can't sell a Web service to his clients, no one can (and he
tells me he isn't seeing any Web service business right now).

His company is being kept afloat by building HTTP Business Services inside
corporate firewalls. Makes sense, actually. Businesses have the money to
spend on improving their processes and information systems.

HTTP services will make a great executive information system

Imagine you're a CEO. You have no freaking idea about technology. (Most
don't). It takes all your brain cells working in tandem to have you figure
out how to open a Web browser and type a URL.

Yes, you're a pointy-haired boss. You just run a widget shop. Or you run a
SnapOn Tools division. Or you worked your way through General Motors or Coca
Cola or Proctor and Gamble.

Does the average CEO have any idea what XML-RPC is? **** no.

What CEO's care about is their business. What kinds of cars are selling.
Which ones are blowing up. What salespeople should they fire? Which
salespeople should they give bonuses to.

Trouble is, in any corporation, that data is all over the place. Some of it
will reside in an Excel spreadsheet. Some will reside in a database of some
kind or another (Oracle, SQL Server, Access, FileMaker Pro, I've seen them
all used, even at the same company!) Some of it will reside in custom apps
(mostly in databases there too). Accounting systems. You get the idea.

OK, now, take that CEO again. Does he or she really know where to find all
the data that's important to him every morning he wakes up? **** no. I'd be
suprised if more than 20% of CEOs out there knew how to use their corporate
VPN to dial in and how to traverse the servers in the company.

So, why don't you build him or her a Web page. One that gathers all the
information from disparate sources. How do you do that? HTTP services. Build
something that parses that Excel spreadsheet, emits some
XHTML/HTML/Chart/whatever, and expose it via SOAP or XML-RPC. Then do that
for your databases. Then do that for your other information stores (hey,
build the equivilent of Napster for your corporation! We know you have a lot
of data stored in .PDF documents. Don't lie now, I know you do!)

Wait, what's missing in this equation? Humans, dummy!

Where do we insert humans into this process?

A Weblog. Gasp! A Weblog in a corporation? You've gotta be kidding, right?
No, I'm not.

Disclaimer, I'm the director of marketing for UserLand Software, which sells
Weblogging technology that is controllable via HTTP via XML-RPC or SOAP.

That CEO is gonna be awfully confused in the morning looking at all those
HTTP services. Even if your programmers did a killer job of making those
charts look pretty (and useful!) and they made the numbers the right color,
and all that stuff, the CEO is still gonna need context. Human context.

His or her employees are gonna need to tell him why coffee sales in Japan
tanked yesterday. Why they can't keep a particular SnapOn Tool in stock. Or
why one keeps breaking, which is dragging down their bottom line.

So, imagine a CEO looking at a single URL on his/her intranet. On the left
hand of the screen, your employees are posting their feedback. On the right
side of the screen, your CEO sees his business data. Sales. Profits (or
lately, losses). Trends. Inventory. Support call reports. The right hand of
the screen is built from HTTP Business Services. The left hand of the screen
is built by a content management server, which is also an HTTP Business
Service.

There's gotta be a few companies out there that'd buy such a system, right?

Building it is easy. O'Reilly has a book that'll teach you how to build an
HTTP service with XML-RPC and there's lots of info out here on the Web.

Jon Rauschenberger has already figured it out and he's gonna build these
things with .NET. He, and his employees, still have jobs. See a trend here?
(Hint, it's not cause he bet on .NET, it's cause he's figured out what
businesses need).

Guess what, we can get the business too, and we don't need to wait for
..NET -- which still won't ship until November (or later)! (And, what we do
implement will be easy to integrate into a .NET centric system, if your
pointy-haired bosses force that on you).

Think I'm kidding? I recently read a report -- sorry, it isn't public --
that says that only a small percentage of CEO's have a good business
information system.

That sounds like an opportunity. Are we up for it? HTTP Business Services
are the answer!

[Jay Glynn]

Mike,
I will probably regret this, but here goes anyway....

> they want either video or on-line shopping. Hardly surprising, when
> you see kids in their hundreds of thousands congregating in shopping
> malls all over the world, instead of crouched over their PCs at home,
> feverishly ordering stuff.

Teens are not in the mall to buy stuff (exception: cookies and frozen
yogurt). They are in the mall to be seen and to hang with friends. You
really are seriously out of touch.

> So, without effective technology plus the
> consumers to use it, where will web services - and .NET - score? Just
> how will this "solution in search of a problem" actually make profits?
>

Your basic misunderstanding of the technologies involved are truly exposed
in this post. The first and perhaps largest error is that you have
automatically tied Web Services to retail. I work for a large financial
services company and we are thinking that web services would be a great way
to expose complex calculations for reuse. For example, insurance premium
calcs. Complex calculations that require lots of data and are used by many
application in the org. If we expose this as a web service, we can
centralize the data and the calculations and allow devs in other parts of
the organization to use them. We have developers all over the globe, and now
we don't have to replicate this functionality.


> Except for Microsoft, that is.
>
> (Hint: Actually, the whole "web services" thing is just a big "smoke
> and mirrors" idea to get you to upgrade to and use XP, Passport,
> Hailstorm, subscription-based charging [the real biggie] and product
> activation.)
>

Once again, you really don't understand. Passport, Hailstorm, subscriptions
and XP don't have anything to do directly with .NET (and barely with each
other). Trying to tie all of these together is a serious misunderstanding of
the technologies.

--
Jay Glynn
Introducing .NET
Professional C# Programming

[Jay B. Harlow]

Mike,
I'm not catching the connection between 'Web Services' and Cell/Mobile
Phones.

I'm sure that Robert Scroble is correct in stating that Web Services are
over hyped. Jay Glynn also explained it well!

Where I see the power of Web Services, and I believe Robert was expounding
upon. Is behind Corporate Fire Walls.

For example: our web server is outside the firewall, every one can get to
it. The web application uses a 'web service' to connect to our server inside
the firewall. The web service then connects to our protected databases...
Presto. A reasonable secure way of accepting insurance applications over the
web. Close to real time, 'straight' into an insurance administration system
written in COBOL. We actually wrote the above in ASP & VB6. .Net makes it so
much easier!

I just wise I had .Net & Web Services 4 years ago... Or at least the concept
of them. ;-) I had a different project that could have really used them...

Hope this helps
Jay

"Mike Mitchell" <kylix_is@hotmail.com> wrote in message
news:3b9b5741.8931156@news.devx.com...
> I read in Saturday's "The Independent" that a survey by the Cambridge
> Business School has showed that only 4% of mobile phone users intend
> to shop on the internet (and, yes, they do spell it with a lower case
> i). This figure is down from the 12% recorded just a few months ago.
> Also, video on your mobile phone? Forget it! Vodafone reckons they
> would need to be the size of bricks to accommodate all of the 3G
> technology. Furthermore, although GPRS offers speeds of up to 30 kbps,
> apparently not many consumers are interested. The kids prefer the
> older GSM phones with SMS, and it's the kids who are supplying the
> profits. They like the small size and there's not much to show that
> they want either video or on-line shopping. Hardly surprising, when
> you see kids in their hundreds of thousands congregating in shopping
> malls all over the world, instead of crouched over their PCs at home,
> feverishly ordering stuff. So, without effective technology plus the
> consumers to use it, where will web services - and .NET - score? Just
> how will this "solution in search of a problem" actually make profits?
>
>
> Except for Microsoft, that is.
>
> (Hint: Actually, the whole "web services" thing is just a big "smoke
> and mirrors" idea to get you to upgrade to and use XP, Passport,
> Hailstorm, subscription-based charging [the real biggie] and product
> activation.)
>
> MM

[Paul Mc]

G'day.

>I read in Saturday's "The Independent" that a survey by the Cambridge
>Business School has showed that only 4% of mobile phone users intend
>to shop on the internet

I am not sure what the link to mobile phone services here is, but I have
been looking at webservices very closely and it is *very* easy to find uses
for them

First, you need to understand just what they are. A web service is a "unit
of functionality", ie a method, exposed via standard internet protocols,
most typically soap over http. For example, you may have a function, lets
say:

Function SearchStock(SearchFor as String) as String

that takes a search string, looks through your Stock database, and returns
any stock records that match - (just ignore the String return type for now).
In a typical client/server app, this function would be compiled into the
exe, or possibly in a support dll. Whenever you need to, you just call it,
no problems...

OK, so lets take a look at a case, off the top of my head, where there are
travelling salesmen, carrying laptops, that need this functionality. How
do we provide access to the data? You could do it via an ASP page, and they
perhaps export or cut'n'paste it into excel or something - a bit kludgy,
and error prone. Or, leave them with the client/server app, and have them
dial in - those phone bills will be running up if the sales team travels
widely, plus you need to provide dial-in access, extra admin for it, overheads
etc. Gets expensive.

How about we implement this function as a web service, return the matching
stock records in XML, and call it straight out of the app? The sales team
simply dial in to any ISP in the world, run their rich windows app, and exec
the SearchStock web method. The data goes right through the firewall - no
problem. .Net framework classes will turn an XML string into a dataset with
a single line of code, so that parts easy. Building the service is no harder
that building the non-web-service version, same with the client, so there
is no extra overhead there.

We could give the webservice address (or a dll wrapper) to our business partners,
and they could build the functionality into their own apps!

The webservice is a sound idea - really it is a natural progression. Really,
you only need to think about it for five minutes and you can come up with
10 different situations that they will be eminently useful.

>Hint: Actually, the whole "web services" thing is just a big "smoke
>and mirrors" idea to get you to upgrade to and use XP, Passport,
>Hailstorm, subscription-based charging [the real biggie] and product
>activation.)

We don't use any of those things (well, we are running office XP, which has
activation). Webservices do not imply or require any of these things.

Cheers,
Paul

[Ronald Laeremans [MSFT]]

Hi Michael,

I can't recall seeing a single company that is not a mom and pop operation
where all internal data is open to all on their intranet. Most companies
don't quite like the concept of e.g. human resources data to be freely
accessible to all employees. I am quite sure there are even a few legal
reasons why that isn't a great idea.

As far as security is concerned, multiple layers offer additional defenses.
I think the concept of removing layers that are in practice very successful
like firewalls is never going to happen. In fact I think adding more layers
is probably what is going to be the future direction.

-Ronald-

"Michael D. Kersey" <mdkersey@hal-pc.org> wrote in message
news:3B9BE747.D77D305A@hal-pc.org...
<everything snipped>

[Joe \Nuke Me Xemu\ Foster]

"Mike Mitchell" <kylix_is@hotmail.com> wrote in message <news:3b9b5741.8931156@news.devx.com>...

> (Hint: Actually, the whole "web services" thing is just a big "smoke
> and mirrors" idea to get you to upgrade to and use XP, Passport,
> Hailstorm, subscription-based charging [the real biggie] and product
> activation.)

Yes it is, to the extent that it's being recklessly overhyped. When,
not if, it fails to live up to the marketeers' intergalactic promises,
some good ideas will likely be flushed down the toilet along with the
sewage. The whole thing is merely warmed-over Remote Procedure Call,
but this time it's been relentlessly simplified to the point that it
might even be understandable, deployable, and securable by non-gurus.
How many publicly accessible RPC or DCOM servers do you see out there?
Well, this silver bullet is just forms and "submit" buttons for use by
other computers, not just human beings sitting behind a web browser.
Of course, now that "web services" are supposed to do *everything*,
including sleep on the wet spot, it's all irretrievably doomed.

--
Joe Foster <mailto:jlfoster%40znet.com> On the cans? <http://www.xenu.net/>
WARNING: I cannot be held responsible for the above They're coming to
because my cats have apparently learned to type. take me away, ha ha!

[Ronald Laeremans [MSFT]]

Hi Michael,

Are you serious in stating that you think that there is a single Fortune 100
company has not a single shred of access control on their internal systems?
A company that has no information that is available on a need to know basis
only. Where everyone can read every other employees email, what everyone
else makes, what their personal information is, what their performance
reviews are? Where everyone can read financial data pre-disclosure. Whether
you are serious or not, I am simply not buying that.

In my pre-Microsoft days I have consulted for quite a few very large
companies, and my expericne was that they typically consider the number of
pencils they bought for a department as 'need to know' information that was
protected.

-Ronald-

[Larry Linson]

"Robert Scoble" <scoble@userland.com> wrote:

> Think I'm kidding? I recently read a report --
> sorry, it isn't public -- that says that only a
> small percentage of CEO's have a good business
> information system.
>
> That sounds like an opportunity. Are we up for
> it? HTTP Business Services are the answer!

No, I don't think you're kidding, Robert. That was the case back in the early
1990's when I was involved in pre-release support for IBM's Executive Information
System. There were some good competitors, too. And they all "delivered the
goods" in a manner that many companies thought was cost-effective. That is,
the "goods" that they could deliver -- see below.

The problem was not that there wasn't a good, effective way to deliver the
information. The problem was, and is, that every company's information needs,
and information sources, are different so that each one needs an EIS that
is very much customized. All an "outside organization" -- IBM, the early
or recent competitors, or your company -- can do is to provide the base for
a very extensive customization project. And, when faced with the realities
of the cost of the development project to determine the needs and create
the customized EIS that would benefit them, many companies decided that the
ROI was not there -- that's why there's such a large unserved market, not
that it's something new that they're going to be falling all over themselves
to leap into. It's a far harder sell, once they've rejected a similar project,
already, with good reason.

So, don't get too enthused over the vast market. It's not new a new marketplace
or one that has recently exploded, and HTTP Business Services is just _another_
way to deliver the information to that marketplace. It may be better than
the most-current alternative, but it isn't a "startling breakthrough", just
an incremental improvement.

You shouldn't _avoid_ the market, because there have been and will continue
to be some opportunities there. Just don't look to it as the "killer app"
that will _make_ HTTP Business Services (or any part of .NET).

[Joe \Nuke Me Xemu\ Foster]

"Ronald Laeremans [MSFT]" <ronlaere@microsoft.com> wrote in message <news:3b9c5ea6@news.devx.com>...

> Are you serious in stating that you think that there is a single Fortune 100
> company has not a single shred of access control on their internal systems?
> A company that has no information that is available on a need to know basis
> only. Where everyone can read every other employees email, what everyone
> else makes, what their personal information is, what their performance
> reviews are? Where everyone can read financial data pre-disclosure. Whether
> you are serious or not, I am simply not buying that.

Keep it up, weasel-boy. As we've seen when Microsoft got 0wn3d by
St. Petersburg, it takes more than a single shred of access control
on one's internal systems. In fact, if even a single shred is weak
or missing, you're going to get it right up the *** from a totally
automated attack launched by some snot-nosed skr1pt k1dd13z!

> In my pre-Microsoft days I have consulted for quite a few very large
> companies, and my expericne was that they typically consider the number of
> pencils they bought for a department as 'need to know' information that was
> protected.

Yes -- it's only the employees themselves who don't know! Meanwhile,
I think lots of people in Russia know just how many pencils Microsoft
buys for each department, don't you?

--
Joe Foster <mailto:jlfoster%40znet.com> L. Ron Dullard <http://www.xenu.net/>
WARNING: I cannot be held responsible for the above They're coming to
because my cats have apparently learned to type. take me away, ha ha!

[Zane Thomas]

Joe,

>Keep it up, weasel-boy.

Do I need to start moving your posts to the offramp again?


--
The nice thing about standards is that
there are so many of them to choose from.

[Robert Scoble]

Interesting article on this topic posted today on InfoWorld:

"The term Web services confuses many people, and what was supposed to make
things easier is making things more difficult," says InfoWorld's CTO Chad
Dickerson.

http://www.infoworld.com/articles/op...connection.xml

Robert Scoble
UserLand Software
http://www.userland.com

###

[Ronald Laeremans [MSFT]]

Since I have no clue as to why you want to insult me for no reason, I don't
see how I can ever have any reasonable conversation with you.

-Ronald-

[Dan Fergus]

You are not the first to come to this revelation Ronald :-)


"Ronald Laeremans [MSFT]" <ronlaere@microsoft.com> wrote in message
news:3b9d2eb4$1@news.devx.com...
> Since I have no clue as to why you want to insult me for no reason, I
don't
> see how I can ever have any reasonable conversation with you.
>
> -Ronald-
>
>

[Mike Mitchell]

On Sun, 9 Sep 2001 13:21:47 -0500, "Jay Glynn" <jay_glynn@agla.com>
wrote:

>Mike,
> I will probably regret this, but here goes anyway....

No, you will regret it! <g>

>> they want either video or on-line shopping. Hardly surprising, when
>> you see kids in their hundreds of thousands congregating in shopping
>> malls all over the world, instead of crouched over their PCs at home,
>> feverishly ordering stuff.
>
>Teens are not in the mall to buy stuff (exception: cookies and frozen
>yogurt). They are in the mall to be seen and to hang with friends. You
>really are seriously out of touch.

Er, did I say they were congregating or didn't I? Er, what does
congregating mean in your Webster? Ah, it means "buying stuff", does
it? (No wonder I use The New Shorter Oxford...!)

>> So, without effective technology plus the
>> consumers to use it, where will web services - and .NET - score? Just
>> how will this "solution in search of a problem" actually make profits?
>>
>
>Your basic misunderstanding of the technologies involved are truly exposed
>in this post. The first and perhaps largest error is that you have
>automatically tied Web Services to retail. I work for a large financial
>services company and we are thinking that web services would be a great way
>to expose complex calculations for reuse. For example, insurance premium
>calcs. Complex calculations that require lots of data and are used by many
>application in the org. If we expose this as a web service, we can
>centralize the data and the calculations and allow devs in other parts of
>the organization to use them. We have developers all over the globe, and now
>we don't have to replicate this functionality.

So, suppose I purchase your calculation widget and work it into my
business. Now suppose that six months hence you decide an upgrade is
necessary. But suppose that upgrade changes the way I'm using your
widget. What do I do? Am I expected to re-engineer my business? Now
multiply me by your (hopefully) thousands of other users. Over time,
versions will drift; some users will be on version 1.1.1.1, others on
1.2.1.1, still others on 1.1.1.2, and you will have a horrendous task
- akin to DLL **** - to keep everyone happy.

Meanwhile, at any point in time, given the much more fleeting nature
of a little, almost insignificant web services widget compared with
the old way of buying in and installing a third-party ActiveX control,
I can decide that I no longer want *your* widget, I'll get a very
similar one from the web services vendor "down the road" (where "down
the road" might, of course, be on the other side of the planet as far
as you're concerned).

Suppose you want to add value to your widget. You take every
precaution to ensure backward compatibility, so that my usage won't
suffer. But I'm not interested in the new enhancement, so I only want
to pay for my, now less feature-rich, usage. You then have not only
the different versions to contend wth, but also the differing
featuresets, price points and usage patterns. Now multiply all *this*
by hundreds of other suppliers such as yourself, all supplying me and
thousands of others around the world with their widgets, all of which
are potentially subject to the same upgrade and/or enhancement
requirements. It is going to be a horrendous management task to keep
these web services working efficiently, reliably and profitably.

Also, you will have to remain competitive with all the other widget
suppliers. Because if web services are to be profitable you will have
to either have huge numbers of customers paying micropayments or a
relatively small customer base which you charge a lot. But no matter
which marketing plan you go for, there will be thousands of others all
vying for the same "instant" microprofits across the web.

Finally, consider now the obfuscation you will need, backed up by some
pretty water-tight legalese. Because unlike a third-party OCX or DLL,
compiled to native x86 machine code, you always run the risk of
someone being able to unravel your web services widget at the IL stage
much more easily than they might have done with x86 machine code, and
rip you off. No matter if they are in your country, you can sue their
*** off. But what if they're in a different country under a
jurisdiction that doesn't make it so easy for you to sue? In short,
the whole idea of web services is fine on paper, but the
practicalities are a different story.

>> Except for Microsoft, that is.
>>
>> (Hint: Actually, the whole "web services" thing is just a big "smoke
>> and mirrors" idea to get you to upgrade to and use XP, Passport,
>> Hailstorm, subscription-based charging [the real biggie] and product
>> activation.)
>>
>
>Once again, you really don't understand. Passport, Hailstorm, subscriptions
>and XP don't have anything to do directly with .NET (and barely with each
>other). Trying to tie all of these together is a serious misunderstanding of
>the technologies.

I'm tying them together because they are all appearing now, quite
suddenly - out of the blue, in fact, for many corporates, who want
Microsoft to delay the introduction of some of these things - and
they're all from the one company. Do you suppose that that is a
coincidence, or rather that it is all very much part of a coordinated
plan? I suspect the latter.

MM