Brian St. Pierre: not quite random: km

Free Evaluation Version

Mon, 13 Jan 2003 22:58:06 GMT

I just uploaded a free demo for those who want to try Prethink before they make a purchase. If you have any questions while you're testing it, please email me (brian at silverback-software.com).

What is Prethink?

Mon, 13 Jan 2003 17:58:06 GMT

Prethink works like a search engine for small offices, home offices, workgroups, or individuals.
Prethink doesn't poke around your network looking for documents. Instead you feed it documents that you want available for search.
You can send Prethink email. It will index your email and make it available for searches.
Prethink can listen to weblogs (RSS feeds). It will index posts and make them part of its search database.
You can import entire directories of files.
Prethink does not run in a browser. It is a regular Windows application.
You can drag and drop files onto Prethink's client window and they will be added to the search database.
Prethink saves you time, energy, and a few brain cells every time you don't have to hunt around your network or your hard drive looking for a document that "you know you have here somewhere".
When you don't have to remember where you saved things, you can remember much more important stuff.
By freeing up room in your head and preserving brain cells, Prethink can make you smarter.¹
If you have Microsoft Word installed on your computer, Prethink can understand Word documents. It will index the contents of these just like any other file.
Prethink consists of a server and clients. A server running on one machine on your network stores all the documents, handles the indexing, and processes search requests from the clients.
Even though Prethink has a server and clients, it is very easy to set up. You can get it running in fifteen minutes or less.
The underlying architecture (called Arsenal) is designed to be open and extensible. We haven't published the SOAP or Python interfaces yet because they haven't been sufficiently frozen. Look for this sometime this spring.
Prethink is not open source, but you will have a significant portion of the source available. It is written in Python and the interface between client and server is SOAP, so many people will be able to customize its behavior without much trouble.
Prethink only costs $50 per user. Volume discounts are available: talk to us.
$50 gets you a year of support and upgrades. Support is through our website. Upgrades will be automatic beginning with v1.1. If, after a year, you no longer wish to get support and upgrades, you can still use the software.
We host our bug database online, out in front of everyone. We've tried to make it easy for you to report bugs and make feature requests.
You aren't locked in. Files are stored (unmangled) on the server. Data is stored unencrypted in a standard database format. Although we don't recommend it, you could extract all of your data with a simple Python program. (We don't recommend it because, while it is trivial to reverse-engineer, the database schema is unpublished and subject to change.)
Backing up your Prethink database is a piece of cake.

¹ Note: this statement has not been evaluated by the FDA and is not a guarantee. Prethink may not actually raise your IQ. But at least people will think you're smarter, and that doesn't hurt, right?

Prethink will make its public debut on Wednesday.

Mon, 06 Jan 2003 17:14:05 GMT

Prethink will make its public debut on ~~Wednesday, 1/15/03~~ Monday, 1/13/03.

Tue, 17 Dec 2002 14:05:57 GMT

Morgan Wilson points to an interesting article on the ethical tight-rope Google tries to walk. I found the following snippet more interesting:

In September, at the height of the China controversy, Google legal eagle Drummond spotted an article about the prospect of a Google IPO, which, the story said, might be the spark to ignite the dormant public offerings market. Drummond forwarded the story with some sardonic comments. In his office, Brin tries to find the email for me but can't. He notes the irony in that, and goes on to paraphrase the note...

Hmm. Google needs a search engine for their email? The full article is here.

Mon, 16 Dec 2002 01:33:42 GMT

Rahul Dave:

One of the questions I want to pose, though, is this: how can we change the way we interact with the operating system, so that we land up creating more implicit and explicit links via our authoring? I suspect the answer there lies with storage allowing multiple characterization by the user and agents and by defining a Radio like shortcut notion which can be used in other documents. But, until we stop using hierarchical storage, this will be hard.

There's an exciting idea: kill the hierarchical file system and instead turn everything into a soup of interlinked documents. It will probably never completely happen, but you can see certainly see a shift coming from CMSs, weblogs, and the like. What would you call it? A "linked file system"? Does a document disappear if there's nothing linking to it? It would effectively become invisible in the absence of a search engine!

Hmm. I guess a weblog implements a sort of "chronological file system": each post is a file, they are sorted, presented, and typically stored chronologically.

I do know that, with a search engine on my desktop (<shameless_plug>e.g. prethink</shameless_plug>), I wouldn't need to remember where all my stuff is saved...

Fri, 13 Dec 2002 22:47:06 GMT

Rahul Dave points to David Galbraith on Google's Command Line Control.

Rahul:

The question of-course is, whats the equivalence to a link and link popularity on the local drive? Google is not just about full text searching, its about the relative importance of documents. So google wont do anything. Whats needed is filesystems to associate more metadata with a file, such as number of reads and writes, to serve as implicit links, and an explicit model for making linking useful in the context of the local system, something like a radio inteface to the whole system!

The strategy that we're using with Prethink is to not attempt to index and make the entire local disk searchable. We'll settle for a subset: those documents, emails, etc. that have been flagged by the user as needing to be searchable. In my experience, it's often only a small-ish subset of your entire local disk that needs to be searched. So we're pursuing the ability to search that subset of documents for an entire workgroup or entire company.

Sat, 07 Dec 2002 16:56:04 GMT

Jennifer Klyse on KM in law firms:

In IT is All About Relationships, Alastair Trower examines the implications of customer relationship management for law firms.
I took a look at the underlying article and was struck, as usual, by this sentence:

CRM is a complex dynamic that relies equally on people, processes and technology. Like a three-legged stool, if more emphasis is placed on technology than people or processes, the solution is bound to fail.

Exactly. I like the three-legged stool analogy. I keep saying that have the KM battle is social. The processes piece is important too. Must remember this.

Thu, 05 Dec 2002 23:07:13 GMT

Ernie the Attorney:

Now, let's see, where did we store our emails? - this was not so funny to the regulators who wanted the banks to turn over the E-mail records. So the banks got fined $8.5 million.

Interesting that Ernie mentions this a couple of days after Will Cox was talking about the headaches of storing all of the nonsense that gets sent around in corporate emails.

Inundated by Email

Tue, 03 Dec 2002 02:41:45 GMT

Will Cox:

I'll hazard a guess that 90% of all corporate e-mail sent is crap, consisting of one-line appendages to massively forwarded posts.

As someone who has assigned co-workers a slot in my spam-filter, I know where you're coming from. And while I agree that a few simple changes to popular MUAs would help to stem the tide, I think that it would be more helpful to attack the root causes. Unfortunately this requires more than just some tweaks to the clients; I'm talking about architectural as well as behavioral shifts.

The symptom is not just the storage required to keep all this email. It is the distraction and the low signal-to-noise ratio that spews into the daily corporate inbox.

First, why use email as a discussion tool? Convenience? Why do all of the "messages" that arrive in your inbox have to be email? If your MUA made it just as easy to post to a discussion group as it was to send an email to your group, you'd reduce the storage requirement from N to 1/N (where N is the size of the group in question). Storage is centralized, where the cost per MB is much lower than on the desktop, and backups are consistently applied. You'd be able to apply a single corporate policy to message retention. The group would gain from being able to apply searches to a larger body of messages.

(Yes, I'm aware that Outlook Express and Outlook provide support for discussion groups. But I've yet to see it implemented in a way that treats discussion groups like email threads, with the same ease of both sending and receiving. Is there a way to make Outlook display the discussion messages in your Inbox? What if someone starts a new discussion that they think you should be in on?)

Second, in all the corporate email systems that I know of, you're storing every message at least twice: once for the sender and once for the receiver. Make a single copy of the message and send a reference to the message to the receiver. The message is retained in one place, and again you can reap the benefits of being centrally stored.

Third, and probably the easiest to solve but the hardest to implement: attachments. One of the more annoying things I've seen is for someone to send out a large attachment (several meg) to everyone in a large-ish group. This is so easy to solve because there must be a hundred ways of allowing the sender to share the file with the group without sending it over email. It's hard to implement because it is so easy from the sender's point of view to just click "send". This last one could probably be solved by a little tweak in the MUA.

Thu, 21 Nov 2002 19:15:35 GMT

Whew. Arsenal is now in beta. Now I've got to nail down what's going into the next release. The biggest part of which is going to be finding a stable database to use with Python. The standard library stuff is just a little too buggy.

Wed, 20 Nov 2002 12:59:56 GMT

I'm currently working through final testing and bug fixes for the next release of Arsenal. This will be the first external release. The main component of this first release is Prethink, a knowledge management tool for small businesses. Prethink includes:

a knowledge base
a search engine for the knowledge base
add articles to Prethink from files
import entire directories to Prethink
email articles to Prethink
import to Prethink by monitoring RSS feeds

There's more in the works for upcoming releases:

full support for drag and drop
publication of new articles as an RSS feed
better searching
interrelations between articles
some new tools in addition to Prethink

(If this sounds like vaporware, it ain't. This release will be available by the end of the week for qualified beta customers. If you want to arrange a demo, please email me: brian at silverback-software.com. Right now I'm shooting for general availability in mid December.)

Tue, 19 Nov 2002 15:25:04 GMT

Ernie points to an article on knowledge management at law firms.

What sort of tools are good for lawyers to use? "The core technology there is a Web server and a browser," says Jeff Rovner, a knowledge manager at Clifford Chance. How about weblogs? Not discussed. Too low on the radar screen, apparently.

Weblogs would help in the generation of some information, but they don't necessarily provide a comprehensive solution. First, you need to be able to find what you need when you want it. That means a good search engine. Second, you probably also want to be able to find information in sources that wouldn't be part of a weblog network: email, memos, other documents (in a law firm, I'm imagining contracts, briefs, correspondence, etc.). Third, you may want to control access to certain information -- especially in a law firm or a health care setting.

Prethink hopes to be more useful by providing more of these tools.

Mon, 18 Nov 2002 13:17:49 GMT

Dave points to Phillip Pearson , who is watching the wizzy front ends for aggregators.

Silverback isn't building an aggregator, but we do have a "mini" aggregator built into Prethink. From our perspective, RSS is just another medium for gathering knowledge.

Sun, 17 Nov 2002 19:09:43 GMT

Phil Wolff: Anil Dash's microcontent client.

Fri, 20 Sep 2002 22:03:42 GMT

Sixteen steps to a renewed corporate intranet [via Jim McGee: k-logs;]

Fri, 20 Sep 2002 18:54:31 GMT

Mathemagenic posts:

10 Best Intranets of 2002 (Alertbox Sept. 2002) with a couple of things highly relevant to KM:

In the long term, we will need better tools to quickly implement major changes in intranet designs. For now, one helpful approach is to structure the intranet's information architecture based on employees' tasks and job goals instead of on the company's org chart. Even major reorgs are likely to leave large parts of a task-based intranet in place, whereas an organizationally structured intranet will require redesign. Indeed, most of this year's winners chose information architectures and navigation schemes that are primarily task-based.

To ideas come to my mind: (1) this is the way to structure any KM application, and (2) I can see come connections with "social capital as a stable part of a company". I wonder if we have to build intranets around people's networks, or this structure will emerge itself with linking, or something blog-like?

Many intranets suffer from a fragmented design and the resulting loss of usability as users are confronted with different rules at every click. The winning intranets had all made great strides toward consistency and were typically successful at overcoming internal politics by the sheer quality of the central design, as opposed to the dubious designs usually produced by individual departments.

Wal-Mart has a particularly fruitful strategy for managing its intranet for consistency: Users own the content and the central team owns the design.

Do we need something similar for corporate blogging?
[Mathemagenic]

Radio Wishlist: Multi-payload klogging

Fri, 20 Sep 2002 13:41:45 GMT

Some links, courtesy of a klog apart:

Radio Wishlist - Multi-payload klogging: a world of content..

I clicked on the event icon and loaded it into my Outlook calendar, later synched with my Palm Pilot.

I prep for the followup meeting by doing my project status flash report.

This is how I want to klog.

This is the fastest, surest way for Radio to take its place in the enterprise architecture. Multiple kinds of payload. Flowing through blogs, metablogging systems, blogspace, and over web service bridges to the apps of our daily work lives.

See also:

How to manually include an event payload in a blog post. To Radio, the event or a vcard are just more html or gems. Radio needs to understand them as data objects outside of the item's "description" blob.

blogging CVs and resumes with Radio.

More payload discussion in

Whither blogs?

Keyless klogging for the rest of us

What is Mediablog literacy?

Radio Wishlist - Audio options.

Chris Janton makes LDAP calls from Radio!

[a klog apart: klogs]
[via a klog apart]

read this later

Tue, 17 Sep 2002 21:15:29 GMT

One more KM blog natureklog.blogspot.com (via Gurteen Knowledge-Log). It's really sad that Ron Donaldson doesn't use something producing RSS feeds - I will be forgetting to check it regularly.
In KNOWLEDGE: FRAMES & CONTEXT (about ‘The Society of Mind’ by Marvin Minsky )

[Mathemagenic]

Want to Come back to this

Tue, 17 Sep 2002 18:41:36 GMT

From Mathemagenic:

I share the point that knowledge has to be addressed as competence. Knowledge is never transferred, but always constructed (=competence has to be developed).

Knowledge sharing consist of (at least) two processes: knowledge is articulated (or expressed, demonstrated) by the one who "shares", and knowledge is developed (learnt) by the one who "receives". "Expressing and developing" process is often mediated (e.g. digitally).

Context is important:

Knowledge comes with "contextual wrapping": knowledge "bit" is "attached" to other bits in our experiences and our mental representations.
Knowledge sharing process has its context as well - who shares, why, when, how. To the great expend it depends on relations between "sharers" and "learners" (and has to do a lot with trust and shared understanding).
I feel that "learning as constructing" has to do with building own "contextual wrapping" for a new knowledge (like adding mortar to get bricks together). I wonder how knowledge sharing context (e.g. relations between "sharer" and "learner") and communication channel influence this process. [Mathemagenic]

Blogging Project Status Reports

Wed, 04 Sep 2002 17:55:58 GMT

Phil Wolff has a good idea:

Blog your Project Status Flash Reports for communicating up..
Reporting a project's status upward shouldn't take more than 5 minutes. The technique is to boil everything down to a well structured, bullet-heavy, one-pager you can forward weekly or monthly.

He provides a couple of templates for project status reports.

Flash reports, if timely and shared electronically, can eliminate a program management round-table meeting every week. Attendees x payrate x meeting length x 52; you do the math. Even with meetings, you get faster, more focused meetings.

I don't know if eliminating meetings should really be the goal. As he says, faster, more focused meetings and allowing better communication and deeper understanding seem like bigger benefits to me.

Flash Reporting Tips:

Create a channel or category for each project's upward communication. This should be an access controlled web site: you'll be reporting personnel issues and bad news on occasion. Not necessarily for the whole team's eyes.

Email a copy of your flash report to your project sponsors, including a permalink.

Use a post title when you blog your report, making it easier to find. "Project Name - Flash Report - Week Starting 7/7/2004" lets you organize different reports

You rarely fit everything on one page, but force yourself. Prioritizing your messages assures sponsor attention to things that matter most to you.
[a klog apart]

A couple of things I'd add or change on that list:

You shouldn't need to email a copy to the PM. He should be using an aggregator to gather the reports.
Set up a process running on the network that harvests from all of the individual channels to produce a master report.
That same process could store and index reports into a searchable database. This makes it easier to find (esp. older) information on a particular topic.

Creating Knowledge is Hard Work

Tue, 27 Aug 2002 22:25:08 GMT

McGee's Musings:
I find that creating knowledge is hard work. And, I've found that keeping a weblog is one absolutely essential tool for helping me catch ideas before they slip away and then working to develop them into something useful.
[via Mathemagenic]

What are the shortcomings of weblogs? How do we solve those problems? What comes next?

There's a Hole In My Bucket

Tue, 27 Aug 2002 19:20:03 GMT

Curiouser and curiouser! in There's a hole in my bucket....
As a klogger, over the past 3 months or so, I have recorded & published tens if not hundreds of thoughts. I doubt if I shared one quarter of output during the last 6 years I worked at various companies. Oh I would probably have emailed here and there, spoken up during meetings. But I wonder just how much knowledge is being lost, second by second, in most companies by each employee. Then multiply up...

But even if they would catch those thoughts, it's going to be very difficult to find something relevant and to understand it our of the context. More or less like forum discussion: you have to follow for some time to make sense of it.

Going through blog archives is not easy... So far I benefit more from the distributed dialog and from the collective filtering. So, blogs is more for sharing, rather than capturing...
[Mathemagenic]

A good point! If klogging is about sharing, then you have to have both parts of sharing: the giving and the receiving! Quite often, the receiving might be just as tough as the sharing -- for a variety of reasons. First, the recipient has to be receptive to learning, receiving knowledge (especially since learning is much more than just receiving knowledge). Second, the recipient has to be able to actually receive the information! If it can't be found, or if it is not in the appropriate format to be able to answer a question, then it is useless and the klogger has wasted his time.

A fabulous idea.

Tue, 27 Aug 2002 19:15:31 GMT

A fabulous idea.

John Udell about on the writeable web, the uses of storytelling, and project weblogging (via Radio Free Blogistan and KMpings):

Nice "sanitized picture" of the projects weblog with a commentary

Time line. In the weblog tradition, recent items appear at the top, and older ones rotate out to archive pages.

Commentary. Entries on the time line refer to, and comment on, landmark documents.

Categorized items. The time line generates narrative flow, but it doesn't categorize items along other important dimensions which are, at the moment, hot issues to resolve, and agreements on how to resolve them. So, these appear in their own columns, and expand on the teasers that appear in the time line.

Directory. Names, e-mail addresses, phone numbers.

Files. These include PDFs, spreadsheets, Word documents, HTML documents, and -- crucially -- selected e-mail messages that I have intercepted and promoted to the status of landmark documents.

It looks like a newspaper and, indeed, serves a similar purpose...
[Mathemagenic]

This could be a big deal

Tue, 27 Aug 2002 18:58:50 GMT

This could be a big deal.

John Robb on using Radio as the next generation desktop [via Mathemagenic]

He says:

His idea, which I think is wonderful, is to put K-Logs at the center of the desktop for these employees and surround them with digital dashboards that connect them to the data they need to do their job.

Indeed. We need to make it easier to gather useful information into a single place; or at least to have the information indexed in a single place. Having a "digital dashboard" through which you can push information into the appropriate bins would be a great step.

Search, very much like a Google on the desktop, would tie it all together.

Google on the desktop would be great. Actually, I think you can do even better. If you're entering the data, you can apply categories and keywords that you trust. You don't have to worry so much about someone spamming the keyword index when you're working within a cooperative organization.

Basically, the K-Log would serve as the routing system for employee data entry.

I can see some problems with this. But there's an opportunity for customized klog input interfaces. Front-ends that are equipped to gather specialized, formatted types of knowledge. Something to think about.

Communities of Scale (2)

Tue, 27 Aug 2002 13:20:06 GMT

Beyond the infrastructure, you have to think about how the content of all those weblogs scales. Sure, you can put 10-20k weblogs on a single static server and fit the 5-10 required servers in a rack. (Note: I don't know a whole lot about the resources required for this; I'm trusting John Robb's numbers here.) But how can you find what you're looking for over 10-20k weblogs?

It's very easy to throw hardware and bandwidth at sites and make them scale. The costs are more or less distributed depending on your architecture. What's difficult is building a scalable community, of finding like-minded souls. And thus we have the Radio Community Server, the Blogging Ecosystem, blogdex and others — or the less-sophisticated GeoCites Member Pages directory.
[The Peanut Gallery]

This point can't be emphasized enough. Without some kind of way of either organizing or being able to search (or filter) the information on a large number of weblogs, you end up with the chaos that is the Web, albeit on a smaller scale (but without Google!). So it is crucial to be able to organize those weblogs into some kind of useful structure.

I'm watching the action on the ecosystem/indexing front to see what happens. (And participating on the blogchalk front, although the organizational scheme and the idea behind it is much less formal or useful compared to, say, the Blogging Ecosystem.