Minding Your Business With Smalltalk (Part 1 of 4)
27 February, 2007
There has been a lot of talk about the future of Smalltalk. There are number of Object Oriented Languages that are candidates for replacing Smalltalk. Why has Smalltalk lasted so long? Why do business software suppliers still choose Smalltalk? Who are the people that still bet on the future of Smalltalk and how do they manage to succeed.
First we talked with Michael Rueger. Michael first came into contact with Smalltalk in 1981. In 1983 he was part of a university group project developing a Smalltalk VM in Modula II(!). He received his computer science degree from the University of Dortmund in 1988 and after the wall fell he moved to Magdeberg in 1994, and received his PhD in 1998. Michael also worked at Disney Online, Squeak Central at Walt Disney Imagineering and then in 2003 Michael joined Andreas Raab and Maic Masuch who were starting a new company: Impara.
Impara is a major source of Open Source Smalltalk development. They have been able to attract some very talented developers and they currently provide a great deal of support and infrastructure to the Open Source Smalltalk community.
Ron: Michael can you give use your view of open source software development? Do you believe that there is a feasible business model behind open source? What are the benefits and drawbacks in your mind?
Michael: If I didn’t believe in a feasible business model I wouldn’t bet a company on it ; -)
Seriously: a few weeks ago I attended an OpenSource business forum organized by the cofounders of SAP AG, Hasso Plattner Venture Fund. The goal of this event was to bring together investors and OpenSource businesses in an effort to explore opportunities in OpenSource. It was interesting to see how both sides are optimistic about the future and viability of OpenSource, but at the same time struggle to come up with truly working business models.
The current consensus for a viable business model is a combination of OpenSource and all that it offers and closed source value added services (e.g. RedHat, XEN). There are different opinions about whether closed source developments based on OpenSource is really a business model based on OpenSource. If someone develops a commercial, closed source application using gcc, or if we develop Plopp using Squeak, is this an OpenSource business model? When closed source projects contribute code and resources back to the OpenSource Community both sides benefit.
If you develop something completely OpenSource, the definite drawback is that it is somewhat hard to make money selling something that is available for free. It is also easier for competing companies to benefit from you ideas. No, I’m not arguing in favor of patents here!
For us at Impara there are two major advantages of working with OpenSource software (Squeak):
– having a community that you can go back to for support. Or, as somebody at the OpenSource business forum put it: most of the best people do not work at your company. IMO an absolute must here is to be fair, that is giving as much back to the community as you can
– being in control. Having the ability to change everything, down to the VM level in your development environment, is, well, priceless.
You always pay a price using and/or developing OpenSource software, but IMO it is worth the effort as long as both sides keep their part of the deal.
Ron: Impara focuses on Education. Do you have any difficulty because you program in Smalltalk?
Michael: Not any more now than in the last 25 years ; -)
The language du jour changes but for some odd reasons Smalltalk stays around. Depending on what you do, people often don’t care what language you are working in. With one very important side condition: what you do has to look and feel, install etc up to professional standards. IMO this is one of the major factors, why Smalltalk applications take a lot of flak, because they are often just not good enough from the graphic and interaction design point of view. If you look at Plopp, I yet have to see a *Smalltalker*, whose initial reaction is not “you did that in Smalltalk?”. Of course they know you can do that, but it is normally just not what Smalltalkers do/like to do. It’s the “boring” stuff that makes an interesting application a usable and successful one.
Ron: Do you still support Smalltalk, or do you have any plans to work in other languages?
Michael: We do almost all of our development in Smalltalk (Squeak). As many people know, I’ve always strongly disliked the not-invented-here syndrome and therefore never had any reservations against using systems written in other languages. That’s why we are using mailman, mantis, drupal etc. For the same reason we have the occasional PHP script that we use for server side stuff and we will actually do some server side Java development in the coming weeks. But also a lot of Seaside stuff as well. Summarizing you could say that the client is 100% Smalltalk (+ plugins and external libs of course), on the server it simply depends on whatever technology does the job at hand.
Ron: What do you believe is the future of Smalltalk?
Michael: That is a very difficult question, especially when trying to take the different existing dialects/systems into account. From where I stand the gap to other languages is getting smaller. Python, Ruby and Ian’s work on Coke (or whatever the current name is ; -) ) show that there will a point in the near future, where even we long time Smalltalkers, especially old fashioned ones like me, will have to admit a better system has arrived. Which one it will be I don’t know, but it won’t be Smalltalk, not this time.
Ron: I know that you have been actively working on PLOPP. Can you share with us more about the direction that Impara is moving? What are your major goals for the company? Is there anything new in terms of services or products that you are offering that you can share with us?
Michael: So many questions about the future while we are still busy inventing it
Impara started out with a strong emphasis on research/prototype oriented work: etoys, croquet, tweak. For a variety of reasons we have moved into a more commercial direction, Plopp as our very first product being the most visible sign of this change. Croquet and Tweak continue to be developed by Andreas Raab’s new company Qwaq , whereas the etoys effort is largely focused on the 100$ laptop (OLPC) project, Bert Freudenberg joining their forces recently.
Impara has been busy working in a larger team of excellent people under the lead of Bob Stein (http://futureofthebook.org) on Sophie. We have new (educational) games in the works and continue to work with the University of Magdeburg and the Design School here on projects related to game and interaction design. Expect to see more fun software (English and French versions of Plopp, language learning and a strategy/puzzle game) and also a first serious release of Sophie soon.
Although we absolutely love what we do it is a constant struggle, as trying to market high quality software for kids is an absolute nightmare. Just take a closer look at the kids’ software section of your local store and you will know what I mean. But, all this wouldn’t be possible without the support of the community and that’s why we will continue to do our best to give back to the community as much as we can.
Ron: Can you give us your opinion of the Smalltalk community? Do you believe there is more that should be done by the community to help move Smalltalk forward? Do you have any opinions about how we might engage private companies like yours and others to participate in the development of the community? Is there more the community should be doing to engage business?
Michael: Now we are entering the true mine field here… Having been part of the transition from benevolent Squeak Central dictatorship to the community managed process we have now, I can’t but come to the conclusion that the community never recovered from Squeak Central ceded being, well, central.
I can see the gazillion entry thread already, but it is my strong opinion that software development by democracy just doesn’t work. (Insert your favorite jokes/anecdotes about what happens when “Michael gets impatient” here ; -) ). While people seriously have given their best to move Smalltalk forward, different forks of Squeak have de facto happened Croquet, Tweak, Spoon, Scratch, and Sophie). Yes, the “f” word…
Working on time and money constrained projects the temptation is to hope the core system doesn’t move at all. Some people succumb to that temptation more than others, but fact is that e.g. Impara just can’t afford to work on the bleeding edge. Companies need stable, but maintained releases. We at Impara supported the 3.8.1 release, but in hindsight it all seems too little to late, to be honest. The community is trying to support everything for everybody, ending up not supporting anything for anybody.
Ron: There are still a number of companies that select Squeak and Smalltalk for their projects. I agree that a focus on stability is very important for business adoption and development. The concept of supporting the stable release for companies makes sense.
On that line a company Agilense contacted Bert and Ian about supporting a critical component of their software. Can you tell us how Impara got involved?
Michael: Impara got involved to help fix the Windows version of the plugin that Agilense wanted to use for their product. It is one of these interesting cases where people grabbed Squeak developed an application in it without anybody knowing about it. So Squeak might be more successful than we think it is.
Ron: Steve’s business is significantly different from your focus on education. Impara though has significant experience with Squeak and Smalltalk. Will you be continuing this relationship and provide support for Agilense and other companies looking to leverage open source software?
Michael: Providing support for Agilense in this case re-started a discussion that pops up occasionally in off-list conversations. The main questions are:
– how many people/companies actually develop commercial applications?
– how many of these would like to have some form of support (contract)?
– who could actually provide reliable support given that the system has so many contributors?
The community is very diverse. It can be very difficult to know where to turn when there is a problem. As we said previously the future of Smalltalk, Squeak and all open source projects may depend on commercial support for new business projects. Stability, support, and the confidence to navigate software obstacles are essential to encourage companies to use openSource software. We at Impara are happy to help everybody getting started and providing support where possible.
Ron: In the next part we will be talking with Steve Hunter, CTO, Agilense. Steve selected Squeak to add an important component to his commercial Smalltalk application. We will be discussing his feelings about software development, open source. We will also look into what Agilense was able to accomplish with Squeak.
Look for the English and French Version of Plopp which should be available in a few days.
Michael Rueger, Phd is the Managing Partner at Impara. He received his PhD in computer science at the University of Magdeburg. He has spent several years in the USA, working with companies like Walt Disney Media Research Group, Alan Kay’s Viewpoints Research Institute and Hewlett-Packard. His original scientific background is discrete event simulation, developing simulation development systems at FhG IML Dortmund and the University of Magdeburg.
Ron Teitelbaum is a long time Smalltalk Programmer. He is President and Principal Software Engineer at US Medical Record Specialists.