What is an interface for a Mashup?

0 views
Skip to first unread message

Sangeeta

unread,
Mar 31, 2008, 7:26:35 AM3/31/08
to EnterpriseMashups
Mashup, in my opinion, is a verb. One mashes up data, imagery, tables,
APIs - whatever, to create something different, and hopefully,
something that is more valuabe than its consituents. In a way, one
mashes up things to create a whole whose sum is greater than its
parts.

And that is the interface for such a mashup? A browser alone? How
about a spreadsheet? How about a desktop interface?

Thoughts, anyone?

-Sangeeta

Sergej van Middendorp

unread,
Mar 31, 2008, 11:24:27 AM3/31/08
to enterpri...@googlegroups.com
Dear Sangeeta,

In our vision any end-user client should be able to consume services to integrate data and functionality and should be able to give the end-user some control over how to configure this. So outlook, groove, onenote, your mobile, mindmanager, excel, visio, powerpoint, browser, rss client, whatever should be mashup ready in the end.

Cheers, Sergej

Kind Regards / Met vriendelijke groet,
Sergej van Middendorp

v-work strategy b.v.
Lassuslaan 23
3723 LG Bilthoven
The Netherlands
Phone: +31 302290242
Mobile: +31 645696035
E-mail: sergej.van...@v-work.com
Website: http://www.v-work.com
Weblog: http://businessjazz.blogspot.com

Sangeeta

unread,
Apr 1, 2008, 12:57:02 AM4/1/08
to EnterpriseMashups
So, in your vision, you see a bunch of folks/vendors providing mash-
ups on the "cloud" for consumption, and a lot of clients on devices
and desktop applications/interfaces ready to consume such mash-ups.
Maybe it does not have to be all in the cloud. Anycase, it is
fascinating!

It sounds very much like web services architecture, does it not?
Except that in case of web services, applications consume web services
- but in cases of mash-ups, end users will consume mash-ups.

Would you then, also envisage a similar set of standards evolve as in
the web services world? Like the UDDI etc?

-Sangeeta

On Mar 31, 8:24 pm, Sergej van Middendorp <Sergej.van.Middend...@v-
work.eu> wrote:
> Dear Sangeeta,
>
> In our vision any end-user client should be able to consume services to integrate data and functionality and should be able to give the end-user some control over how to configure this. So outlook, groove, onenote, your mobile, mindmanager, excel, visio, powerpoint, browser, rss client, whatever should be mashup ready in the end.
>
> Cheers, Sergej
>
> Kind Regards / Met vriendelijke groet,
> Sergej van Middendorp
>
> v-work strategy b.v.
> Lassuslaan 23
> 3723 LG  Bilthoven
> The Netherlands
> Phone: +31 302290242
> Mobile: +31 645696035
> E-mail: sergej.van.middend...@v-work.com
> Website:http://www.v-work.com
> Weblog:http://businessjazz.blogspot.com
>
>
>
> -----Original Message-----
> From: enterpri...@googlegroups.com [mailto:enterpri...@googlegroups.com] On Behalf Of Sangeeta
> Sent: maandag 31 maart 2008 13:27
> To: EnterpriseMashups
> Subject: What is an interface for a Mashup?
>
> Mashup, in my opinion, is a verb. One mashes up data, imagery, tables,
> APIs - whatever, to create something different, and hopefully,
> something that is more valuabe than its consituents. In a way, one
> mashes up things to create a whole whose sum is greater than its
> parts.
>
> And that is the interface for such a mashup? A browser alone? How
> about a spreadsheet? How about a desktop interface?
>
> Thoughts, anyone?
>
> -Sangeeta- Hide quoted text -
>
> - Show quoted text -

Stefan Andreasen

unread,
Apr 1, 2008, 8:22:01 PM4/1/08
to enterpri...@googlegroups.com, Stefan Andreasen
Hi all

Let me throw in my 2 cents here.

The major difference between mashups and other "composite applications", is the usage scenario.

Enterprise mashups are synonymous to agility, being able to create business value when it is urgent and most valuable. The value might be short or long lived, the value might be only for an individual or a smaller team or forum or a whole company.

Here is what I believe has become the consensus definition of an Enterprise Mashup.

An Enterprise Mashup is a combination of 2 or more data sources into an application, dashboard, spreadsheet or data feed, to create new valuable business insight into the data. One could say it is the next generation spreadsheets.

Very importantly an Enterprise Mashup, is an application which can be build without the (or with minimal) need to recode all the underlying data providers, i.e. it has to be able to leverage the IT infrastructure as-is in a very agile and situational business driven scenario.

En Enterprise Mashups typically evolves out of the Line Of Business or from the knowledge workers. They work with data, often in spreadsheets, and need at way to work with data in a more automated and dynamic fashion. The main objective is to get faster and better results and also to be able to handle more of the exponentially increasing amount of data sources on the private and public net.

If you did not know, here is what Kapow Technologies solves for mashups:

Kapow solves the #1 obstacle for businesses to benefit from the value of Enterprise Mashups, which is getting access to the data in a very fast and cost-effective manner. Basically getting access to the right data at the right time.

Today almost all data are available through web application interfaces on the public and private web, and Kapow Technologies gives easy, robust and scalable access to intelligence data through their web front-end.

The advantages are several:

1. There is a one-fits-all interface to all data, no more need for special adaptors for each set of data.
2. Business people work with and know their data through the application interface, so Kapow gives them access to data in their world.
3. There is often valuable business logic build in to applications at the interface level. Accessing data through the user interface, makes it possible to work with the data in the right business context, leveraging all the application business logic.
4. The data that are the most difficult to access often has the highest "intelligence" value. The is because they differentiates and improves the result from the result of competitors who lack these "intelligence" data.

Best

Stefan Andreasen, Founder and CTO, Kapow Technologies, Stefan.A...@kapowtech.com

Sangeeta

unread,
Apr 2, 2008, 6:58:40 AM4/2/08
to EnterpriseMashups
Hi Stefan,

Good to hear your thoughts!

So, in your view, it is a 'Web-at-the-backend-AND-Browser-at-the-
front" that defines the context in which a mash-up would be consumed
by the end-user. Or am I missing something here?

But I wonder if that world view is realistic at this point of time.
Does all enterprise data make itself available on web interfaces?
Should be in the future, but not today - or so I think.

And would users need only the browser to view this data? What about
some other alternative interface? Is there a case at all of browser
not being the interface for a mash-up? Would users throw away all
their powerpoints and visios and what-have-yous?

Further, how about the delivery of mash-ups on the plethora of hand-
held devices that are riding the adoption wave? In the device world,
one seldom sees universally portable application.

If one takes away the browser, which one may have to, to suit the
current world view, the interface of a mash-up could indeed be
anything - it is pretty much anybodys guess. With the devices being
another contender for mash-up delivery, the question becomes even
tricker.

-Sangeeta Patni
Co-Founder, Extensio

Disclosure: I run a company called Extensio Software
(www.extensio.com), that provides information delivery channels for
enterprise data.

On Apr 2, 5:22 am, "Stefan Andreasen" <stefan.andrea...@kapowtech.com>
wrote:
> Hi all
>
> Let me throw in my 2 cents here.
>
> The major difference between mashups and other "composite applications", is the usage scenario.
>
> Enterprise mashups are synonymous to agility, being able to create business value when it is urgent and most valuable. The value might be short or long lived, the value might be only for an individual or a smaller team or forum or a whole company.
>
> Here is what I believe has become the consensus definition of an Enterprise Mashup.
>
> An Enterprise Mashup is a combination of 2 or more data sources into an application, dashboard, spreadsheet or data feed, to create new valuable business insight into the data. One could say it is the next generation spreadsheets.
>
> Very importantly an Enterprise Mashup, is an application which can be build without the (or with minimal) need to recode all the underlying data providers, i.e. it has to be able to leverage the IT infrastructure as-is in a very agile and situational business driven scenario.
>
> En Enterprise Mashups typically evolves out of the Line Of Business or from the knowledge workers. They work with data, often in spreadsheets, and need at way to work with data in a more automated and dynamic fashion. The main objective is to get faster and better results and also to be able to handle more of the exponentially increasing amount of data sources on the private and public net.
>
> If you did not know, here is what Kapow Technologies solves for mashups:
>
> Kapow solves the #1 obstacle for businesses to benefit from the value of Enterprise Mashups, which is getting access to the data in a very fast and cost-effective manner. Basically getting access to the right data at the right time.
>
> Today almost all data are available through web application interfaces on the public and private web, and Kapow Technologies gives easy, robust and scalable access to intelligence data through their web front-end.
>
> The advantages are several:
>
> 1.      There is a one-fits-all interface to all data, no more need for special adaptors for each set of data.
> 2.      Business people work with and know their data through the application interface, so Kapow gives them access to data in their world.
> 3.      There is often valuable business logic build in to applications at the interface level. Accessing data through the user interface, makes it possible to work  with the data in the right business context, leveraging all the application business logic.
> 4.      The data that are the most difficult to access often has the highest "intelligence" value. The is because they differentiates and improves the result from the result of competitors who lack these "intelligence" data.
>
> Best
>
> Stefan Andreasen, Founder and CTO, Kapow Technologies, Stefan.Andrea...@kapowtech.com
> > - Show quoted text -- Hide quoted text -

Sergej van Middendorp

unread,
Apr 2, 2008, 3:18:39 PM4/2/08
to enterpri...@googlegroups.com
Hi Sangeeta,

I think Mashups build further on SOA standards, but are more driven by the user.

I envision a lot of new specific service and data standards to develop to indicate content, identity, activity. Semantics and ontologies for generic knowledge worker activities that enable end users, which is the key word for me, to take control of local task processing and collaboration tied to their specific work rules, context, culture and habits, while still remaining in synch with a legacy content, process and data world.

Personally I use mindjet's mindmanager and microsoft's onenote as a 'mashup' to aggregate content, mix it with activities, reuse processparts, integrate with outlook, crm, teamsites, content sources and the like. Wish they had a server version coming up ;-)

Cheers, Sergej

-----Original Message-----
From: enterpri...@googlegroups.com [mailto:enterpri...@googlegroups.com] On Behalf Of Sangeeta

Sent: dinsdag 1 april 2008 6:57
To: EnterpriseMashups

Sergej van Middendorp

unread,
Apr 2, 2008, 3:27:49 PM4/2/08
to enterpri...@googlegroups.com
Hi Stefan,

Thanks for sharing your vision. You mention the "minimal recoding issue". In my experience minimal support to disclose specific data sources or functionality in legacy applications will be a key issue once we cross the corporate firewall. I see a shift in the skills needed to support the agility of real workplace activity. The people doing the customization would resemble the ERP consultant, who is trained in business processes, but then a 21st century evolution of that character, who understands the ins and outs of 'knowledge work' from the end user's perspective, but who is still able to make the adjustments needed in legacy data and applications through advanced middelware. The thing is though that ERP consultants still had backup developers in the attic to make the bespoke software addons. How will we deal with that?

Another route would be the standardization of generic knowledge work activities, interactions and semantics. If we deliver an interface that adheres to such a standard, real end-users will be able to model their own work and the integration can be put in the background. Are you doing anything like that or am I way off?

Cheers, Sergej

Stefan Andreasen

unread,
Apr 4, 2008, 9:17:19 PM4/4/08
to enterpri...@googlegroups.com
Hi Sergej

I mention the "minimal recoding issue" since if you need to do major
recoding to access the data you want to mash up, you cannot at the same
time be "agile" and thus it's not a mashup.

So how do you solve this. Well there are two ways:

1. Wait (forever) until everything adheres to a set of standard
interfaces to access data

Or

2. Find a way to hook up to the data in the way they already exist.

With Kapow Technologies, we allow our customers to leverage that most
data today are accessible from a web browser, so our product plugs into
a web interface and gives standard access to the underlying data through
say SOAP, REST, ATOM or RSS.

Or did I misunderstand your question?

Best
Stefan

Stefan Andreasen

unread,
Apr 8, 2008, 11:09:07 PM4/8/08
to enterpri...@googlegroups.com
Hi Sangeeta

I guess I am not very good at expressing myself.

A mashup is merely a combination of two or more data sources, creating value that was not in the individual data standalone. The way you access the data or the way you present them does not matter as long as you can get to the data in an agile way, quickly when you need it, and you can build the end-result quickly too.

That typically puts limitation on how to access data sources, to use standard interfaces and accessing the data over the web.

About the front-end, I believe the most broadly used is Microsoft Excel, since most business people have it already, know how to use it, etc. So no it does not have to be web based.

I hope this helps

Stefan

Sangeeta

unread,
Apr 9, 2008, 1:36:09 AM4/9/08
to EnterpriseMashups
Great discussion going there!

Building on the discussion, here is where I see this go. As I see it,
there are three distinct phases to the mashup creation and delivery
part.

1. One is to access the data from the sources- and the more agile and
easy it is, the better it is. If every lay business user can pick the
data she wants it, from the source she wants it, and without any help
of any developers, that would be nirvana. And all of these web
services standards, APIs, web interfaces of enterprise data - are
doing just that. Making the data easier to access. It hasnt gotten to
be as easy as it should be for a lay user, and I wonder if it can ever
be. As Sergij mentioned, it perhaps would be non-trivial to break all
enterprise sources to easy-to-consume web parts, and not a job that a
non-developer can handle yet, but we are getting there.

2. Second is to mashup two or more sources and create something of
greater value - based on what the user currently needs. It can be data
juxtaposed together to create something that is uniquely valuable to a
particular person is a particular situation - and then is junked
forever, or it could be something that is created by someone, found of
value, and shared so that similar value can be derived by others. Here
too, if a lay business user can do it without needing a developer, and
kind of on-the-fly- that would be the way to go.

3. The third is to be able to provide this valuable mash-up at the
place where the user needs it - or the interface of the mash-up, which
is what this discussion started with. It could be used on Excel, as
Stefan pointed out, or could be a Outlook calendar where people use a
map mashup to find the most convenient place to meet other people on
the meeting, or a blackberry device that shows the driving
instructions mash-up that helps the user drive to that meeting.

I think all of these pieces -the data offerers and extractors, the
mash-up creators, and the mash-up deliverers - would be needed for a
full blown mashup ecosystem. It is early days yet, and we are still
finding our feet in this terrain.

For enterprises, the issues of user authentication, identity
management and the others will also be important - and these would
perhaps be another set of supporting services/standards.

-Sangeeta

On Apr 9, 8:09 am, "Stefan Andreasen" <stefan.andrea...@kapowtech.com>

Sergej van Middendorp

unread,
Apr 9, 2008, 6:49:54 AM4/9/08
to enterpri...@googlegroups.com
Hi Stefan,

No it's becoming clear now. I have worked from the portal paradigm for a number of years, there the major hickup is bottom-up data quality and integration. The end user would end up doing his system in a spreadsheet outside of the enterprise processes.

So if I am right, Kapow accepts that this bottom up integration from legacy to user will never really finish and turns it around. Why not honour excel and let users tap into data they see in a webinterface which they know is qualitative (data) and in a context (business logic) and then work with it? Right?

How do you account for business logic end-users may want to add and share? Do they create new rules in kapow?

What happens when the underlying data / rules accessed through the webinterface change?

What support roles do you see close to the user that are needed to leverage the value of mashups?

Reply all
Reply to author
Forward
0 new messages