Now blogging at diego's weblog. See you over there!

is it symbian-bashing?


My previous two Symbian articles (here and here), while, I think, balanced, elicited some comments that I was being too hard on SymbianOS and its efforts for the developer community. Before I continued writing about Symbian I wanted to clarify what I think, to avoid any further confusion.

First of all, something that probably didn't quite come across (but that I added in one of the comments I made to my first post in mobitopia) was that I consider Symbian's work so far excellent, and that this whole issue is something that in my opinion will be more of an issue in the near future.

I consider these problems a natural outgrowth of the nature of what Symbian does in terms of its multiple target platforms. No one has done anything like this before, and Symbian has done remarkably well so far, and the fact that this is coming out now only means that it's time to move to the next level.

For example, consider my comments on development environments (on the two entries linked above). Then read Russ's entry on the .Net development environment for smartphones, including the comments. I hate to say "I told you so," but there it is.

Information and toolkits exist, but they are not easy to find, or to put together. People with experience (even limited experience) on Symbian had said that I was wrong on this count, but I truly believe that this is a bit of "inbredness" that afflicts the platform at the moment. Once you are familiar with the technology and the documentation, things make sense, but that is true of anything, even Win32, which has to be the most confusing API ever devised by man.

Recently I've been doing a sort of "mini-evangelizing" of Symbian. And people don't get it. They just don't see what the OS can do, the tools that are available. Some of them don't even consider it a development platform. When I start to describe the capabilities of the new Symbian-enabled phones that are coming out, they open their eyes wide "Really?". These are technical people, and they are not hiding in some closet. They have been doing mobile development for some time now, in some cases for years, in many cases starting in research environments at the university and they have been using PocketPC and so on because a) it's easy to "see" what to do with it (an HP iPaq is more "visible" as a development platform than, say, a Nokia 3650) and b) because Microsoft has marketed the hell out of the platform to those early stage developers "See? Just Win32" MS says, and of course that's not true because the Win32 implementation of PocketPC is severely constrained compared to the PC-side of things. But they get started quickly, and that's all that matters. After a while, they are locked in. In my experience, pointing people to DevNet just baffles them completely. They see all these toolkits, pointers to each manufacturer's site, all these IDEs, and they assume (wrongly) that they have to develop for each platform separately.

Some developers take pride in the fact that sometimes information is obscure or difficult to get at, that you need hours of sweat to set up a development environment or to understand the advantages and limitations of a platform ("If you're not willing to suffer for a while, then we don't want you here"). It reminds me strongly of the Linux atmosphere in the mid-90s: "Documentation? Of course you've got gorgeous man pages, and maybe even comments in the source code! What? Graphical environments? Sure, just spend a few hours configuring the vertical scan rate in your XF86Config... And if you can't handle that, well, then you're just not ready for us..."

Linux got over that mostly, but there's one big difference: Symbian is the incumbent in the next-gen mobile phone market. And as such, it can't behave like an upstart. Much less when Microsoft is moving in.

The capabilities offered by Symbian OS and its target devices open up possibilities to create new applications, AND with a massive installed target base. I know people that are not expert programmers, and yet they have good ideas and would like to put them into practice. They are "formal" tinkerers, but there are many like them that don't necessarily have time to set up an environment from different sources, and understand that to use J2ME you have to get Sun's WTK as well as stuff from Nokia, for example. Those tinkerers will come up with the cool applications that will then seep back into the platform to make it vibrant, and within corporations, if someone can easily set up an environment and show a particular idea working easily, they are more likely to get it approved for further development.

Sure, this entails also "widespread evangelism", but simple, easy to use tools out and a single entry point for the information related to those tools (going to the manufacturer's site for device-specific information only) gets you half-way there. (and Javasoft's Java Developer Connection) is a good example of what I'm talking about. I think that when I'm looking for Java information I spend maybe 80% of my time there. While if I'm looking for information on Symbian, I spend 80% outside of Symbian's site. (Yes, I know that if I'm in Nokia's dev site I might be reading relevant information for any Symbian OS phone, but that's not so easy to see, is it, when everything is labeled "Nokia this" and "Nokia that"?). Things like this would go into that site. Even if it doesn't seem like it, The JavaSoft site, and JavaDoc, played a big role in getting Java adopted quickly.

The information is all there, and the only thing that's needed is better organization and centralization, particularly of forums for discussion. Symbian as a development environment has some problems but at this moment I give it the benefit of the doubt. The situation is improving. And Symbian has a great advantage, both in terms of development experience and deployment of its platform. But it has to continue evolving and becoming simpler, and moving into new areas.

So, is any of this Symbian-bashing? No. It's trying pointing out problems that I see, and doing what I can by continuing to write how-tos and such (and sometimes just a bit of frustration seeping out, yes).

I'd like to add: It's one thing for us to just talk about it, it's another thing entirely to actually do all of this while juggling new product releases, partners, company growth, etc., with the big Redmond Machine breathing down your neck.

Credit where it's due: If not for Symbian, MS would have already taken over the mobile phone space as well (or, nothing would have happened in the area at all). Google seems to get a lot of credit these days for keeping MS on its toes, but Symbian deserves a good amount of it for doing the same in the Mobile arena. (Google obviously has a better PR department :-)).

So, hats off to Symbian and its people!

Now, coming up on my end: a short review of J2ME on Symbian OS.

Posted by diego on August 23, 2003 at 1:48 PM

Copyright © Diego Doval 2002-2011.