I was a little more
positive on the last call about this topic,
but thinking about it more, I am tending
more towards Colin’s line of thinking on
this. What is great about the TGF is that
(in my over-simplification of it) it is a
recipe book for using the power of
technology for good. And “good” in our case
is achieved by putting the citizen as our
focal point.
It’s not about the
technology per se – and that’s the point of
the TGF.
Another way to
state it is – how do you orchestrate all the
cloud/web/technology services in a manner
that benefits citizens? In many ways, if I’m
stitching all these services together, I
really don’t care if the service is
delivered by 2 dogs and a cat – so long as
it meets all my requirements – including
security, data protection, availability etc.
Why would you look
beyond the service and into the platform
itself? Well, I would think the main reason
would be that a particular platform might be
used for multiple services – and therefore
many of your requirements could be satisfied
at the platform layer – thus effectively
aggregating your requirements across a
number of services. Like this:
In the picture,
you’ll see a bunch of rectangles
representing platforms. The services run on
these platforms. Some of these platforms
will be in-house, some in the cloud and some
will be multiple layers deep (eg a cloud
service provider ‘white labeling’ other
services, which themselves are delivering
apps and services). Some of these platforms
will therefore be “virtual” (because they
are combinations of other platforms) and
some of these platforms might be “micro”
(because they run on your TV, phone or
device).
As an example, it
may be possible to say a particular platform
meets your requirements for data protection
and availability. That might be the black
box in the picture above. You may therefore
be able to say that the services on this
platform inherit those requirements – and
that might simplify service requirements for
you on that platform.
I think most of
this is consistent with Nig’s perspective.
But I’m not sure the end result is a
“Platform Management Framework” – that
sounds kind of inward, or technology
focused, rather than outward/citizen
focused. I’m also not sure that Colin’s
suggestion of expanding the “Technology
Management” section is right because a lot
of this is more about supplier management
and business processes. Perhaps it should go
under “Service Management” (though with
“platform” we’re talking more about the
inward-looking or supporting services rather
than the citizen-centric perspective on
services – which I think should remain our
focus).
The other thing to
consider is the role of SLA’s (Service Level
Agreements) in this discussion. There is
plenty of work going on in this area (eg
ISO/IEC 19086 – Cloud computing – service
level agreement framework and terminology)
so the intersection of SLA’s with the idea
of platforms could be interesting – although
again I would see “platform” as really just
supporting infrastructure for “services”.
Specifically, what does the “platform”
concept add to SLAs? And with that
perspective – is this work something that
should be done in the TGF?
I look forward to
further discussion.
Geoff Clarke
Thanks
John
And
a special thanks to Nig for the work on
the Platform pattern.
The
extra detail and Nig’s careful attention
to the things being emphasised and the
things it is silent on, shows the effort
put in to get a more balanced result.
I
guess I could live with it, and I can even
offer up a diagram from NZ’s Govt
Enterprise Architecture ..which nicely and
agnostically abstracts everything such
that you can make it fit pretty much any
government related architecture framework
you like, including support for the
platform pattern..
J.
But
……
As
those on the last call may remember, I was
not in favour of the new pattern.
I am
more predisposed to Chris’s suggestion of
extending the Technology Management
section if we do anything at all.
It
was a strategic viewpoint – that we were
heading into dangerous territory by
indicating more technical direction than
is already in the Framework, and it risks
‘tipping the balance’ in favour of
technology over culture and behaviour
and..and.. and being ‘written off’ as
technocrats by our detractors.
Despite
Nig’s most valiant efforts, I’m still of
that view. But I accept I am probably in
the minority, given what a great job he
has done…
J
We
have developed the framework to what I see
as a finely balanced piece of work. We
have given some broad guidance around
technology without over prescription. This
is a classic case of ‘less is more’. By
leaving it the way it is, we allow
deployers to add more of their own
interpretation. And that will change as
the years go on. If ‘platform’ is
introduced there’s an implicit
presupposition about the what a TGF
architecture might entail. And that will
change over time, as it has already done
since the TC was formed. And as I said on
the call, it’s not about the technology.
It’s about the culture, the behaviour, and
the approach to the ever increasing range
of challenges that Public Administration
face. Technology is no doubt an enabler
for whatever the public service wants to
achieve. For example, let’s say that in 10
years, some countries cannot afford the
tax burden to run public services as
bloated as they are now and the public
service is radically cut to just being a
governance/risk/compliance overseer, with
everything else done by the private
sector. In that case, let’s say the toll
road owner has its road sensors, working
with the car, and how many times the
driver crosses his curb into his driveway,
in order to take micro-taxes. Apart from
GRC, there is no public service ‘platform’
as such..not in an architectural sense
anyway. The toll road owner may be
federating with multiple other private
sector parties without any government
involvement or any kind of ‘hub’.
As
the TGF stands today, I think its
‘agnosticsm’ /abstraction allows for that
kind of scenario. With a platform pattern
included I’m less certain…
Cheers
Colin
Attached are the
papers for our TC call on
Thursday (Wednesday for Peter
and Joe). The call details are
contained in the zip file but if
you have any problems please
contact Geoff Clarke.
I would appreciate
some more comments on the draft
Platform Management pattern
before the call please, either
positive or negative.
A note of any
absences would be welcomed in
advance.
Finally just a
reminder that this call is at
the new times of
07.30 UK /
08.30 CEST / 02.30 EST / 23.30
PST (18th) / 18.30 NZ
/ 16.30 AUS.