We're floating the idea of GlassTop, a transparency mechanism for the
open development of standards, particularly in the domain of
e-democracy. We have expressions of interest from 4 projects, as
potential users. ^[1]
* Metagovernment/Metascore
* Telematics-Freedom/Continuous Democracy
* DemocracyLab
* Votorola
An alpha prototype will take roughly 2 man-weeks. The tasks are:
1. XML and RDF design. We need a designer to spec the details of
data storage for the central and client nodes. The designer will
also spec the requirements for the central Wiki.
2. Scripting of the query toolset. We need a programmer to code the
tools for assembling the data records from client nodes, and
caching them for later queries. The programmer will also code
the UI (command line) and the query engine.
3. Management and/or promotion. We may need a project manager to
sheperd the development process, and bring in resources.
Resources will include a few pages on a neutral Wiki (per 1). We
may also need a promotion lift, to get off the ground.
We aleady have one developer (myself). We need another for 1 or 2,
and a manager/promoter for 3. Can anyone commit?
DESIGN SKETCH FOR GLASSTOP ^[2]
==========================
(This is only tentative. We are still looking for ideas and
alternative proposals. Comments are welcome.)
1. Scope
2. Use Cases
3. Structure
4. Function
5. References
6. Glossary
1. SCOPE
========
GlassTop is a transparency mechanism for the open development of
standards. It implements a clearing house for the proposal of new
standards, with facilities to track the levels of commitment,
development and compliance among stakeholders.
GlassTop is not:
* Not a process for standards development.
* Not a collaborative platform.
* Not a standards organization.
2. USE CASES
============
1. A client (C) has an idea for a standard S. It is potentially
useful to other clients. So C posts the standard under
GlassTop, and asserts two facts:
- S is proposed
- C supports S
2. A second client (D) is looking through GlassTop, and learns of
S. S looks like a good idea, so D asserts:
- D supports S
Meanwhile, C and D are talking together, and making improvements
to S.
3. A third client (E) learns of S. S looks like a *bad* idea for
E, so E asserts:
- E does *not* support S
Meanwhile, E is talking with C and D, and explaining why it's a
bad idea.
*. And so on... for all clients, all standards, and all modes of
support.
3. STRUCTURE
============
GlassTop is deployed in subnets. A subnet is the formal model of
a particular domain, as shown in Table 1 ^[3]:
TABLE 1. Modelling.
--------------------------------------------------------------
Formal Models Actual Objects Examples
------------- -------------- --------
subnet domain social software
standards nodes standards standards for social software
client nodes stakeholders projects, business firms and
other stakeholders in the
domain of social software
--------------------------------------------------------------
A subnet is an open database, with indeces and records distributed
across central hubs and peripheral nodes (figure 1). The central
hubs are a standards hub and a client hub, both maintained in a
public Wiki:
SH. Standards hub. Contains an index of standards nodes (S),
and access URLs.
CH. Client hub. Contains an index of client nodes (S), and
access URLs.
S
\ C
\
\ /
/
S ----- SH /
/ CH ----- C
/
/ \
\
S \
C
FIG 1. A subnet, showing central hubs
(SH and CH), and peripheral standards
and client nodes (S and C).
The peripheral nodes are of two types: standards nodes and client
nodes:
S. Standards nodes, one per standard. Each node includes one
or more documents that specify the standard. The standard
is maintained exclusively by its provider, at the
provider's own location.
(Note: S contains no GlassTop-specific content. GlassTop
may therefore cover all standards, regardless of provider,
location, age and format.)
C. Client nodes, one per client. Each node will publish a
short-index of standards of interest. For each standard of
interest, the client will record the levels of commitment,
compliance and other supports.
The index is maintained exclusively by the client, at the
client's own location.
4. FUNCTION
===========
Clients and other users may employ a query toolset. A query toolset
is a set of scripts for executing queries across a subnet. Example
queries are:
* What new standards have been proposed in the last 2 months?
* Which clients are currently agreeing, in principle, to standard S?
* Which clients are actually compliant with S?
* Which new clients have joined, in the last three weeks?
* What standards are supported by client C?
5. REFERENCES
=============
[1] The original statements of interest were posted in
"Inter-project standards - seeding the process":
http://metagovernment.org/pipermail/start_metagovernment.org/2009-February/thread.html#1223
[2] GlassTop was first mooted in the post "GoogleVotes monopoly
vs. JoeVotes ecosystem":
http://metagovernment.org/pipermail/start_metagovernment.org/2009-February/001197.html
[3] Where two domains overlap, their subnets may also overlap. In
that case, aome of the peripheral nodes of one would be included
in the other.
Where two domains nest, their subnets may also nest. In that
case, all of the peripheral nodes of one would be included in
the other.
(In this early proposal, however, we assume a single subnet.)
6. GLOSSARY
===========
central hub
- A node maintained on a central Wiki, and publishing an open
index of periphal nodes.
- The client hub, or the standards hub.
central Wiki
- The Wiki that hosts the central hubs.
client hub
- The central hub that publishes the index of client nodes.
client node
- A peripheral node corresponding to an actual project, business
firm, organization or other stakeholder.
peripheral node
- A node indexed by a central hub.
- A client node, or a standards node.
query toolset
- A scripted toolset for executing queries across a subnet.
standards hub
- The central hub that publishes the index of standards nodes.
standards node
- A peripheral node corresponding to an actual, proposed
standard.
subnet
- An interlinked network of central and peripheral nodes,
corresponding to a particular domain.
support
- (of a client for a standard) Such as agreement in principle
with the standard, participation in its development, or
compliance.
--
Michael Allan
Toronto, 647-436-4521
http://zelea.com/
Knowing you're interested, and willing to help - underwriting it, so
to speak - that may be enough to kick it off.
> We should connect with groups in the U.S. after their TransparencyCamp
> happening in DC this weekend. I'm also hoping to build linkages
> between the US and Canadian gov transparency groups at SXSW, with
> Meghan Warby's help.
I've been talking with some TransparencyCamp'ers on the US side, and a
couple of them are interested - Ed Pastore, Mark Frischmuth.
(Although I'm pitching the alpha trials at e-democracy, it could also
cover other domains, including e-government. So it could make
governmental transparency *itself* transparent, by showing which
governments have adopted which transparency standards, all at a
glance. In theory...)
> Anybody interested in this topic going to SXSWi from the 13th to the
> 17th.
(Can't travel, no budget. I'll buzz around, and see who else can be
hooked... There's almost enough interest, I could just code it
myself. It's not a big app.)
Thanks Jennifer. We'll need a critical mass of initial
users/projects/stakeholders (maybe a half dozen). First, there's the
alpha to code (no problem). Then I'll come back (maybe in a month),
and ask for help - an initial push - if needed.
(Sorry to hear you're sick. Get well soon!)
Yes, it would serve that purpose. Suppose the APIs haven't yet been
recorded, as an existing standard. No problem. You just add an item
to the standards list, on the Wiki (centrally). Anyone can do that.
Tool suppliers then start recording their compliance (locally, at
their own sites). You then query for API compliance, etc.
> It's dumb: Neil Eggsgard (theyworkforyou.ca) proposed a db listing for
> projects ages ago on VG discuss, but I wasn't that into it because of
> all the wikis that were going up, and I thought there would be
> consolidation eventually w/ a master semantic wiki or something.
> Funny how you miss the obvious.
>
> http://groups.google.com/group/visiblegovernment-discuss/browse_thread/thread/581e4559d77b5bd3
Yes, same motivation as Neil's!
My thinking now is, there won't be any custom software at the user
end. The user will have an off-shelf query tool - user's choice.
There may be some scripting at the back end, where the indeces on the
central Wiki are converted to RDF, readable by the tool (or we just
use a semantic Wiki, for that).
I'd like to build this *immediately* - it's a gem - but it's too much
of a give-away to my competitors. So my plan is 1) pound my
competitors a little more firmly into the dirt; 2) give them a brief
rest, while I build GlassTop (if nobody beats me to it); 3) resume
pounding my competitors into the dirt.