TechFold - Bold tech & web commentary

KnowledgeTree - Open Source Document Management for the Enterprise

This post is for anyone that works at an “enterprise” and has to deal with the horror of big, scattered teams and network drives.

I work with documents on network drives all day, and can attest to the problems that arise: multiple scattered versions, weak search through Windows XP, and folder-based categorization often make workflow awkward, confusing, and frustrating. One company offering a solution is KnowledgeTree - South African developers of an open source, enterprise strength document repository product (here are the KT SourceForge pages).

A “document repository” creates a “super-network drive,” for lack of a better explanation. From a high level, KnowledgeTree can be summarized as the following:

1. Its a hosted storage system running on one or more servers in your organization - a file server.
2. You can save documents to it, and open documents from it with hooks built using WebDAV. (KT includes a Windows Shell browser, MS Office integration, and a Web shell)
3. It has superior searching capabilities, indexing file contents and a whole mess of metadata.
4. It allows you to add metadata to files (tags, custom metadata fields, etc).
5. KT offers a CVS style “check-in / check-out” system and versioning features.
6. If you open the KnowledgeTree server to the Internet, it allows for external collaboration.
7. It allows for logging and reporting on all document activities.
8. It includes user and group based permissions and security features (also does LDAP and Active Directory integration).

KT goes a step beyond file server and search functionality however, building in smart collaboration features. Threaded discussions for files, for instance, or folder RSS feeds. Controlled emailing of files is another one that jumped out at me: emailing a KT URL solves the problems of large attachments and multiple versions sitting in inboxes.

Unfortunately, I’m not going to take the time to try out a free, at-home KT installation (fill out the form to download). I will, however, be recommending that my organization take a look at KnowledgeTree and the document repository space in general - I’m getting tired of mapping network drives, chasing down versions, getting permissions changes, and generally dealing with files the same way I have since 1993.

KT jumped out at me because of its 2.0 features - tagging, RSS feeds, and the like. More important than featureset, however, is the notion that business users can benefit from what has traditionally been an IS-centric tool: with the lines between IS and business blurring as everyone becomes a knowledge worker of one type or another, the notions of workflow management, version management, and searchability need prominence for managers on the business side. KT does this by being a full-featured solution that integrates in business workflow out of the box, and couches the concepts of CVS management in a familiar MS interface. CVS for the business - I love it.

A final note: KT offers a free, unsupported version for download, or paid, fully supported “Enterprise & SMB” versions, starting at $2,200 USD per year for 20 licenses, and switching up to $5,500 per cpu per year for larger businesses. How about hooking a local, KT front end to Amazon S3 for truly scaleable storage? I suppose network latency vagaries would make it untenable for mission-critical settings.

Screenshots below are from KT’s own screenshot & video page. I’ve picked out some of the juicier ones…

cms, cvs, document, foss, knowledgetree, management, opensource techfold

Share This

If you enjoyed this post, make sure you subscribe to my RSS feed!

Where are all the opensource billionaires? At MySQL.

A few days ago, Hugh MacLeod kicked off some industry navel gazing by asking “why aren’t their any opensource billionaires?” Answers rolled in:

1. Jeff Atwood / Coding Horror: its fragmentation, competition, and the very nature of FOSS.
2. Seth laments that corporate aversion to risk and the difference between actual and apparent need holds back FOSS.
3. Dennis Forbes succinctly dissects all of Hugh’s logic, suggesting that Hugh’s analysis was pulled off-track by a poor understanding of sales and service channels.

At the end of the day, I’d agree most strongly with Seth’s assessment, but post script by saying that the era of opensource billionaires is on its way as corporate mindsets begin change on the credibility of FOSS-developed software. Case in point: MySQL’s coming IPO.

billionaires, foss, hugmacleod, mysql opensource

Share This

If you enjoyed this post, make sure you subscribe to my RSS feed!

Emerging Tech: Attention Profiling Mark-up Language

Attention Profiling Markup Language (APML, APML.org) aims to create a stardardized means of collecting, importing, and exporting attention data between web applications (found via: Dallas Freeman).

“Attention Data” is essentially your clickstream - i.e.: what Amazon collects to make recommendations to you based on what you’ve looked at in the store. APML aims to standardize the collection of such info into a portable “attention profile” which can be exported and moved among services much as OPML allows you to move your subscriptions from BlogLines to GReader.

The point implicit in APML is that a user’s attention profile has intrinsic value and utility - and that it would be worthy synchronizing across services (eBay, Pandora, Google Reader, and Amazon, for instance), or exporting when migrating from one service to another. This speaks to the growing importance of discovery services: as the breadth and depth of content availabe on the net in general and on individual mega-sites like Amazon becomes un-parseably huge, automated recommendation/discovery systems will become one of the primary means of finding content and enabling transactions.

Its a very conceptually appealing idea - if you’re a technical person familiar with the concepts of attention data and discovery. Before its relevant for the consuming public, some marketing people will have to wrap a layperson-friendly vocabulary around the concept.

Final note: APML is FOSS.

APML is positioning for user-control in a discovery-centric net future. Someone out there is thinking ahead - good job!

apml, attention, foss google

Share This

If you enjoyed this post, make sure you subscribe to my RSS feed!

Collanos P2P Project Collaboration - Please be better than Groove

A few weeks ago, even after I posted an unflattering list of suggestions for Teqlo, Jeff Nolan pointed me towards beta startup Collanos - a new entrant into the ad-hoc team collaboration space.

CSCW Backgrounder

“Team Collaboration” (or popularly “Computer Supported Collaborative Work,” or CSCW) defines a suite of services that spans:

1. Project management (assigning & monitoring tasks)
2. Communication (instant messaging, discussion threads)
3. Collaborative editing (filesharing, file versioning)
4. Team & role management (inviting members)

…all of which is packaged together in a single application to allow for streamlined, one-place management of all resources/information associated with a particular project (each of which gets its own shared “workspace”). Compared to the ordinary mishmash of disconnected email threads, files scattered across network drives, and so on, CSCW apps like Collanos have a compelling value proposition.

The P2P Angle

Collanos differs from something like SharePoint by maintaining workspaces via P2P, as opposed to a centralized server infrastructure. This means that each Collanos client updates its neighbors, who update their neighbors, and so on - and collectively, the entire team stays in sync. The advantages are:

1. Fast setup (no centralized repository to setup)
2. Inexpensiveness (no infrastructure to pay for)
3. Flexibility (teams can grow or shrink without performance hit or centralized admin overhead)

I had a blank test workspace up and running in about 3 minutes from the time I confirmed my registration online.

The Downside of P2P and CSCW

Scalability. P2P ironically doesn’t scale well in a CSCW setting. Or it didn’t for me - I had the unfortunate experience of using Ray Ozzie’s “Groove” with a team of 60. Every morning, my Groove client would spend hours trying to resynchronize its multi-gigabyte local repository. So would everyone else’s, bringing the office network to a halt and making everyone’s pc’s unstable while Groove ate most system resources. If you took a day off, you were better off to uninstall Groove and start fresh than to try and re-sync.

Conjecture #1: Downloading music and video is not mission critical. Synchronizing a shared workplace is. P2P inherently can’t offer a guaranteed “service level,” making it unsuitable for mission critical apps, however appealing it may be conceptually.

Confusion. Once you actually get sync’d, hope for the best - there’s nothing like finding 15 different versions of a file that different people worked on over night, among which yours is not included because it didn’t synchronize, or whatever &*#@%^$!# (the “Last Modifier Wins” scenario).

Human Nature. Finally, there are many who question the utility of group-targeted applications at all. Given that you’ll be stradling both the groupware world of Collanos and your regular individual desktop-type functions for work outside of Collanos, groupware can create a layer of overhead that results in slow and fragmented adoption by teams - which further decreases utility, leading to user drop off, and so on. This comment is based on experience: with Groove, users and usage declined more or less linearly over a six month period until the beast was finally killed.

Perhaps Collanos and Groove work better for smaller teams (5? 10?) with limited file volume - I don’t know. It begs the question though - would a 5 person team find value in a collaboration tool? My understanding is that collaboration tools are fundamentally intended to streamline work across larger groups, as with a small group its relatively simple to stay organized. I can see CSCW being useful for small, geographically distributed groups - but this is a very small target market on which to base a business.

Conjecture #2: Any group small enough to consistently use a CSCW tool without experiencing user-base fragmentation or overloading the P2P architecture is small enough that the marginal value of a CSCW tool will be very low.

A Harsh Prognosis for a Good Application

My analysis has been colored by a terrible experience with Groove, which Collanos strongly resembles in appearance, functionality, and architecture.

That being said:

1. The Collanos client has a good mix of features, a clean UI, and installs and works nicely (for a Java app).
2. Cross platform capabilities expand its usage scenarios as well.
3. A free version will drive trial usage and adoption.
4. There’s an educated, experienced team behind the app.
5. Collanos is listening and proactively responding to feedback and shortcomings.
6. Collanos uses JXTA for P2P, which with the FOSS community pushing it along will hopefully offer better functionality than Groove.
7. Marketing - say what you will about promotions, but Collanos is working hard to sell itself in the Bay Area.

So - while I’m not bullish on the CSCW space in general, if anyone’s going to dispel my bleak outlook, its going to be Collanos. I’m looking forward to following their ongoing development and seeing how Collanos in the North American market.

collaboration, collanos, eclipse, foss, groove, jeffnolan, jxta, microsoft office

Share This

If you enjoyed this post, make sure you subscribe to my RSS feed!