The following message was sent a couple of days ago to the axkit-users list, so I believe it's public and it can be republished here. I think it's great, but I leave it to you to comment about.
Hi Matt and all,
>> My company is preparing a huge development project based on AxKit - a
>> virtual document management system. The core will be built on top of
>> AxKit, and we plan to contribute the whole core to the community.
> That's awesome news.
With the risc of beeing of-topic:
Last year I created an AxKit plugin for OpenDocument files based on
Matts earlier work on OpenOffice 1.x files; however, I did use it in
another way than the way I think Matt originally intended. I used it for
a webserver CMS - letting my customers create and edit their webpages
using OpenOffice, and I then styled it and assured a consistent and
professional layout using an XSLT stylesheet. For the customers, all
they had to learn was how to up- and download their
WebDAV, and how to edit them using OpenOffice - I took care of the
Wow, did they like it! That system has a huge potential - especially
when people figure out that they can even cut and paste pictures into
their OpenOffice file, and they'll show up nicely on their webpages. No
more hassles about uploading pictures to the right directory, no more
broken links when people still cannot do it right - everything's clean
On the programming side, I planned to spend around three weeks to get
the system ready to roll: I started Saturday morning, and Sunday morning
I was happily creating the first layout for my sons homepage
By January, I started a new job. The company I work for primarily
develops ideas - real programming is usually outsourced. We have a close
relationship with Ricoh, the Japanese manufacturer of copying machines.
They wanted an idea that could integrate their machines closer in a
typical corporate network.
I brushed off my AxKit/OpenDoc project, worked it over a little, and
the result was VDMS - Virtual DMS. That was February. By March and April
my boss was on a Novell GroupWise tour all over Europe (totalling 15
countries), together with Ricoh and Blackberry. Included was a
description of VDMS - we wanted to see whether it was a product that
could be of interest to the potential customers. At the beginning of the
tour he tried to sell the idea, at the end of the tour he tried to keep
potential customers away - they where crazy about it.
By May, we signed a special high-priority development contract with
Ricoh. By July, we signed another development contract with Blackberry.
We have agreed that we develop the system, and Ricoh will market it
worldwide. That means a salesforce of 9.000 people... So far we have
interest from one of the worlds largest oil companies, one of the worlds
largest car manufacturers, one of the worlds largest shipping companies
(which happens to be Danish, so we're in very close contact on that
one), several very large banks, and the list goes on.
We have signed a pilot contract with a building contractor who will
fund the first development phase; apart from VDMS, we will write
software that allows a Ricoh MFP to be used as a file and directory
server. It is said that the Ricoh machines run Linux - however, I think
it's either Open- or FreeBSD instead, but anyway it's quite interesting.
And we have even more dirty tricks up the sleave that I'm not allowed to
talk about at the moment.
But what is VDMS? In principle, it's a Linux host running Axkit and a
bit of other software. It has a lot of SOAP interfaces to other boxes,
allowing it to handle all documents in a company. We've created a lot of
documentation, but to keep it short, a will give a few examples of the
* All documents handled by VDMS shall be in OpenDoc format. That means
converting MS Office docs to OpenDoc, helping people to migrate their
docs to a better format
* The building contractor can scan a meeting resume at a building site
on a Ricoh MFP. The Ricoh machine has a full Java SDK, so we create a
client that will pass the scan on to VDMS through a SOAP interface. VDMS
has rules that defines where the document - a raw scan - should be
stored. It can be an ordinary file server, allowing access to the
document using ordinary means, or it can be a database system, or
whatever. Anyway, VDMS has another rule that dictates that the doc
should be OCR scanned by a suitable engine. So the doc is forwarded to
the OCR scanner, and stored on the fileserver again, this time as
searchable text as well as the original bitmap. A new rule requires the
scan to be proof read, so it's forwarded using email to a proof reader
and stored again when it's finished.
* Of course you can create all kind of indexing for such a document;
but the interesting part is when you want to retrieve it. You can do
that using a huge range of clients such as a browser, a mobile phone, a
PDA, a Ricoh machine, etc. etc. Every device requires a special format,
and I expect AxKit to handle that flawlessly
interesting special case: We will format the document in XSL:FO, then
convert it to PostScript and print it. But we're in a unique position to
suggest to Ricoh that they create a native XSL:FO engine, completely
bypassing PostScript. That's quite a potential - and Ricoh listens a
* But VDMS can retrieve "documents" from a lot of other sources: As an
example, GroupWise has a SOAP interface, so we can create an XML based
mail client. The idea is to create a VFolder that will allow, say, the
building contractor to create folders for each project containing not
only ordinary files, but also all emails related to the project,
financial data from their accounting system, data from outside partners,
etc. etc. VFolder creates a unified view of all the data, it's XML
based, and handled by AxKit. We intend to create a VFolder as a Windows
Explorer extension, as a Java client for Ricoh and Blackberry, etc.
* The whole core project - the various AxKit plugins, the specification
of the various SOAP interfaces, the (huge amount) of XSLT stylesheets,
etc. will be released as Open Source. The customized clients for Ricoh,
Blackberry, Windows, etc. will probably be closed source and sold.
* I'm particularly thrilled by seeing that Matt has included RDF
support in his Gallery demo. That's another road we want to go with VDMS
as well - creating and handling metadata and making it available in RDF
So - does this sound interesting? At least it's a lot of work, so if
anyone would like to join the party, please contact me the sooner the
better. We already have the customers, we have a global salesforce - we
just need a little coding...