Saturday, October 09, 2004

I've moved 

Sourcelabs is my new company, the one I've been hinting at over the past few months, and my blog is now over there.

Wednesday, August 11, 2004


I'm looking for a few good people to join me in a new venture. I won't say much here, but send me an email if you're interested. The venture will be focused in a very specific area that is tied to the overall mission of making computing cheap, easy, and ubiquitous. We're looking for people who are passionate about making software work (building it, testing it, supporting it) and know a thing or 3 about open source software.

Wednesday, August 04, 2004

Edit Me 

I have been playing around with editme.com. It's still has a way to go, but it makes basic wiki-style collaboraiton relatively easy and friendly - and it's a hosted service.

My feature list:

easier way to manage users
out of the box templates - calendar would be very useful!
revision history
easy way to create tables of information, with views of that info

There are more companies to come in this space, but they've sworn me to secrecy for now... stay tuned.

Wednesday, July 21, 2004

BPEL Answer: 2 weeks 

Active Endpoints is first to open source a BPEL implementation. The word is that their tools are excellent. I haven't tried them. Has anyone?


Tuesday, June 29, 2004

XML Beans and Apache / How long for BPEL? 

XML Beans is now a full-fledged Apache project, one of what is becoming a pretty comprehensive set of XML projects at the ASF. David Bau led the creation of XML beans, which you can read about here. Basically, they make it relatively easy and painless to deal with XML documents in Java. An intersection with SDO at some point seems likely. This seems to be a good example of a vendor (BEA) working with the open source community to provide open technology that makes developer's lives easier and more fun.

In other news, Oracle bought Collaxa. If you haven't seen Collaxa's product, it is a very slick tool and runtime for BPEL - an admittedly flawed standard, but one that customers seem to like. One wonders how long until a solid open source implementation of BPEL is available...

Monday, June 21, 2004

Mission (Not Yet) Acccomplished 

Taking some time off is a good way to remind oneself of why work matters. In addition to paying the bills I mean .

What I concluded was that 200 years from now, when historians summarize "our" generation, one of the most important points (the end of totalitarianism possibly being the only other contendor) will be "communications and computing became free, easy, and ubiquitous." And thinking of that reminded me why I'm in the software industry - to help be a part of that. Corny, yes - but hey, I'm that kind of guy.

We haven't really accomplished the mission yet. Communications are mostly free, easy, and ubiquitous (email, phone, IM, etc.) and becoming more so everday. Desktop computing is nearly free and ubiquitous. But server software and internet services are, with the notable exceptions of Google and Yahoo, mostly hard, expensive, and not for the faint of heart. Moreover, the intersection between communications and computing - what I've called collaboration in the posts below for the lack of a better word - is still nascent.

I think one of the most interesting challenges of the next 10 years is to change this. And clearly open source and software-as-a-service (or the intersection of the 2) are going to be key ingredients.

Friday, May 21, 2004

Survey results 

Lots of smart people are looking at these four problems right now.

Managing personal information, rights, and preferences. In the Net world, we want to live across different machines - in our own home, when we're traveling, etc. Yahoo has not done much here, and Google's first forays seem rather tentative. Passport - enough said. Amazon Web Services are cool, but they haven't taken it further. So if none of the obvious candidates are going to solve this problem, who will? Paul Maritz recently started PICorp. Worth watching. Here's another. Why isn't Yahoo leading the charge here?

Smart clients. .NET has seen its only real success with people still building thick clients, and interestingly lots of folks still are. In fact if anything the numbers of thick(er) clients are increasing. People tried to fit user experience into a web programming model and in many cases found the results less than acceptable. I won't say much here since you can go to Adam's site and hear it from someone who knows much more about the challenges.

Getting Linux and Windows to play well together. I've spoken with a number of companies trying to tackle this from various perspectives - making Linux work well with Active Directory, building a unified console across both OS's, even running Exchange on Linux. This seems like a fertile area that isn't rocket science, but clearly demands an independent company to build a solution since Microsoft and the Linux guys are too busy fighting.

Post-J2EE web infrastructure. The implosion of Sun, and by inference stewardship of the J2EE community, has got people looking at alternative languages - PHP5 includes a new object model designed to be more friendly to Java programmer, "better" ways of scaling and managing applications (Terracotta amongst others), and Java-based frameworks (Project Beehive from BEA open sources key parts of Workshop.) Will the Java world re-invent itself quickly enough to maintain momentum?

Nobody I've talked to is building the killer app for the web of the next 10 years - a collaboration framework that embraces IM, voice telephony, and webservices/BPEL in addition to email and Office.

This page is powered by Blogger. Isn't yours?