Pepper Author Maarten Hekkelman

Maarten Hekkelman is the programmer behind Pepper, a Macintosh programmer’s text editor that started life on BeOS as Pe. Two weeks ago, the 36-year-old resident of the Netherlands called it quits, and announced that he was no longer developing or selling Pepper. The Daring Fireball interviewed Mr. Hekkelman via email — who better to interview the author of a Mac text editor than someone who used to work for Bare Bones Software?


John Gruber: Let’s start with the news. According to your web site (http://www.hekkelman.com/), you are no longer taking orders for Pepper, the text editor you’ve been working on for several years. In fact, it is no longer even available for download, except for registered users.

What happened? It seems like an awful lot of work to abandon.

Maarten Hekkelman: It is. It took me some six years to develop what has become Pepper. Pepper started life as Pe on BeOS and was ported to Mac OS [beginning with Pepper 3.0].

Frankly, I was very disappointed by Mac OS X and so [for version 4.0] I decided to do a port to Windows. The code was already quite portable but now I separated Pepper’s code from the more generic code. I had used PowerPlant for the first Mac OS port and now I started replacing PowerPlant with my own framework. I got help from an old friend, Bas Vodde, who helped me get up to speed on Windows (I had never programmed for that OS before). It took me nine months to finish Pepper 4.0 for Mac OS and Windows, and a Linux version was released a couple of months later.

But Pepper 4.0 was not well received. The first releases had some performance issues and people blamed it on going cross platform. However, most of the performance problems were related to fixing bugs internally and were not at all related to the platform switch.

The negative feedback spread around; on VersionTracker there were many complaints, and this influenced sales. And then Pepper’s serial number was cracked, I got my first fraud cases, had a quarrel with Tucows, and Pepper didn’t run on Jaguar and then I gave up.

So Pepper 4.0.6 does not run on Jaguar?

Nope, sorry.

Pepper 3.0 debuted for the Mac in July 2000, only a few months before the first public beta of Mac OS X, and only 8 months before the release of Mac OS X 10.0. Did you ever consider dropping support for the old Mac OS, and making Pepper only for Mac OS X?

No. First of all, Pe was already written in C++ and so the choice for PowerPlant to replace the BeOS API was obvious (in my eyes). And then, although I did have the Mac OS X betas at the time, they were not very useable. And I simply don’t like Cocoa and would never program in it, since it is a Mac-only technology and I don’t like to be locked in.

Do you think most of Pepper’s users are using Mac OS X?

No, I think it is more like 50/50. No hard figures though.

That’s interesting. I really thought Pepper appealed more toward Mac OS X users. For one thing, on Mac OS 9, Pepper’s Perl filters require MPW ToolServer, whereas BBEdit works with the MacPerl application (which is much easier to install than MPW).

I always used it on Mac OS 9 and it is a very good editor on that platform, if I may say so.

I didn’t mean to imply that it wasn’t good on Mac OS 9, only that I thought the market opportunity was on Mac OS X. When people switch to a new OS, they are suddenly in the market for all sorts of new software.

I know you didn’t mean that. And you’re right with your ideas.

What was the argument with Tucows about?

They refused to list Pepper [on the grounds that] it wasn’t good enough. I changed some things and resubmitted. Then I bought an InDepth review in the hope this might help get Pepper listed. But the reviewer didn’t get Pepper to run on the machine he wanted to test it on. This ended up in mud wrestling since eventually I wanted my money back and they refused.

I must admit that I’m very unfamiliar with Tucows (I haven’t used Windows in about six or seven years). Do you mean that to get listed on Tucows you need to pay them to review your software? Seems like it would be hard for them to be objective in such a case.

You can buy bonus points and they will be added to your product’s rating. So yes, you have to buy your listing. But since they have such power [in the Windows community] they can do this.

What effect do you think having the serial number cracked had on sales? I often wonder whether the people who use illicit serial numbers would have bought the software anyway.

No effect at all. But it hurts very much, especially if you work so hard to make something and price it so low everyone should be able to pay for it.

Indeed, I always thought Pepper had a very low price.

Too low, far too low. But I was hoping this would help make it more popular. And on Windows prices are all in this range. It would have been strange to sell Pepper for Mac OS for US$100 and Pepper for Windows for US$30, only to have comparable prices to the competition.

There are several dozen low-cost shareware “text editors” and “HTML editors” for the Mac, many of them built with REALbasic. For the most part, they offer very few features which do not come for “free” with the REALbasic application framework. Did you ever wonder if the existence of all these crappy text editors made it harder for Pepper to gain attention? In other words, people may have seen the announcement for Pepper, and thought “Oh, another piece of crap,” without downloading and trying it out?

No, the tragedy with Pepper was that eventually it was well known and a lot of people were tracking it with VersionTracker. But many decided that this version wasn’t good enough yet to dump BBEdit and start using Pepper.

I received very few negative reactions to my decision to stop developing Pepper, and all came from people who hadn’t yet bought Pepper but were hoping to do so in the future.

Your own description of Pe admits that it was “inspired by MPW and BBEdit — both editors on Mac OS”. Were you a full-time Mac developer before switching to Be? Which was your primary editor on the Mac?

I’d had a couple of full time Mac programming jobs at the time. And when working on Mac OS I used Codewarrior for programming. I did have a license for BBEdit and tried to use it from time to time, but in my eyes it didn’t have enough extra to make it worthwhile to use it as an external editor [for CodeWarrior].

I also used MPW from time to time, same story here. BBEdit did have a worksheet, but it didn’t work as nicely as it could have.

Have you used BBEdit 6.5? Its new worksheet for Mac OS X is very nice.

No, I only use Macs if I have to. And I stopped looking at BBEdit since I found so many of the features I had come up with.

BeOS vs. Mac OS X

The BeOS generally earned positive reviews, but never gained enough users to achieve critical mass. When did you start using the BeOS, and why?

I read about it when they first made their OS public and was immediately hooked. A GUI and a POSIX layer combined in one machine and with multiple anything as well! Great. Several months later, in 1996, I read about it again and decided to take the risk. I went to the bank to get a loan to buy a BeBox. Never regretted it, life was wonderful then.

Be, the company, seemed to have the strategy to market their OS (and their BeBox computers) towards developers. One of their main marketing points was that the BeOS had a very powerful, easy-to-use programming API — and that the API was designed from the ground up to be object-oriented, and therefore wasn’t built on a legacy procedural API, like the Mac OS or Windows. What was your experience programming for BeOS? How did it compare to programming for the Mac OS at the time?

Programming for BeOS was simple when you just started. But it became quite messy quickly. The problem is the multi-threading. As long as you have one window doing the job there’s no problem, but try to display one document in more than one window and you’re in trouble. This was the reason why Pepper was never ported back to BeOS.

The API was nice though. Comparable with PowerPlant on Mac OS. (PowerPlant is better though.)

On the surface, BeOS and Mac OS X sound somewhat similar. They both have underlying Unix/POSIX layers, but both have GUI’s that are designed to be cohesive and intuitive. Both are popular with hackers who like the programming power of a Unix-style command line, but also appreciate a well-designed GUI interface. What did you like better about BeOS compared to Mac OS X?

Almost everything. BeOS was fast, extremely fast. It was new, fresh, etc. I can only think of very few negative points, technically, in BeOS. The ugly GUI might be one of them, the absolute need to use multi-threading might be second.

Mac OS X, however, loses on all fronts. It claims to be a Unix but it doesn’t support much of the more advanced Unix features, since it is using such an old kernel. It claims to be user friendly, but I find it more obscure and difficult to use than my Win2K box. And then it is dog slow.

And since the Mach kernel is a personal project of Avie Tevanian, it doesn’t look like that will ever change.

Yup. Too bad. I had a couple of cool features in Pepper for Linux/FreeBSD but couldn’t port them to Mac OS X since the kernel did not support them.

What features?

Redirecting a pipe to a new window in Pepper. It works in Mac OS X now (well, 10.1.x) but works much better and nicer on a real Unix using SVR3 style RPC calls. There were several other features. One thing I miss dearly on Mac OS X is the /proc file system. I have been thinking about writing a debugger for Mac OS X but this omission made me decide against it.

Did you always have it mind that you might one day port Pe to the Mac, or did you only consider that after the BeOS was discontinued?

No, not when I started it. When I returned to the Mac I started using BBEdit and CodeWarrior first, but I missed so many features I had implemented in Pepper that I decided to simply port it.

Why go back to the Mac instead of Windows with Pepper 3.0?

At the time I was a true believer. I had hope that a Mac OS would appear that would be as great as BeOS. Besides, I was one of those Microsoft bashers at the time.

How are you disappointed by Mac OS X? As a user, as a developer, or both?

As a user, I don’t like it. God knows I tried, really hard, honest.

What counts for me are the details, and they were all wrong. I found so many UI errors in OS X, I couldn’t believe it. A huge amount of work that went into designing the ultimate GUI was thrown away and all we got back was a bag full of candy that was dog slow.

And as a developer I hate it. I had to use Carbon of course, and it sucks, too many changes, too many bugs. Pepper stopped working on 10.2, because in Jaguar your Carbon event handlers have to be reentrant … ridiculous.

It seems like you went back and forth quite a bit on Pepper’s underlying text rendering engine — using ATSUI in the initial release of Pepper 4.0, then making ATSUI optional. Was that where the performance problems were?

ATSUI is slow, really slow. I implemented the QuickDraw option in later 4.0.x releases to illustrate this. But the biggest problem performance-wise in Mac OS X is scrolling. It is really absurd how much time that takes.

However, some of the performance problems were my own fault, I should have profiled more cases like huge files and thousands of changes in such a file. Pepper uses a sophisticated transaction mechanism to store edit changes and I screwed up making Pepper work better with mixed line endings.

Was this problem introduced between 3.6 and 4.0? Is this fixed in version 4.0.6?

It was introduced in 4.0. I wanted Pepper to be more robust with incorrectly formatted UTF8 files, for example. And yes, most of it is fixed. At least in the copy I’m using :-)

ATSUI is a Mac-only technology. What did you use for text rendering on the other systems?

Glad you asked. On other OS’s you simply use the default Text API, on Mac OS you have to create kludges to work around the very poor text support. I really don’t understand why Apple was so stupid. They could have introduced a single new scriptcode to render Unicode using QuickDraw and life would have been great. But no, Apple came up with ATSUI, which is nice for a typographer but a horror for an editor programmer.

And it is slow as well.

And you didn’t uncover these performance problems in beta testing?

No. I didn’t open 100+ MB files very often. Now I do, with my new job.

How many beta testers did you use? How did you recruit testers for the new platforms supported by Pepper 4.0?

I had some 20 beta testers. Far from enough, and although I appreciate very much what they did, they weren’t exactly very useful when it comes to real beta testing. They did provide good ideas though.

I hardly had any beta testers for the other platforms.

Public beta tests seem to have gained a lot of popularity in the last 10 years or so, especially amongst smaller developers. What are your thoughts on public beta testing?

I’m sorry I have to say, but most people asking to become beta testers are only hoping for free software and never give feedback. There are a few testers who are very good and vocal, you have to be good to them. But in the end, they hardly find any serious bugs and that’s not strange. They often use a subset of the features and those are often well-tested. It is the more obscure features and details where the bugs linger. The only way to test software is to use it yourself or to make a test plan and hire a company to do the testing for you.

Why replace PowerPlant? Because you didn’t like it, or because you wanted to go cross-platform?

I wanted to go cross-platform. And I didn’t like some of the PowerPlant implementations, I thought it would be more efficient to use Carbon Events more extensively. My mistake.

How much of Pepper’s source code is shared across the different versions, and how much is platform-specific?

Pepper consists of about 160,000 lines of code. 15,000 is Mac OS specific, 12,000 is Windows specific and 7,100 is X11. The rest is OS neutral.

What’s your opinion on the development tools and API’s for Windows?

I’m using CodeWarrior on Windows, CW5 to be precise. It is a bit outdated but since is still works and CW is so awfully expensive I never upgraded. Microsoft delivers a very, very good set of documentation and tools for developers. For free. It was a real eye opener to have such great documentation so easily available.

I did try some other tools on Windows like BoundsChecker from Numega, a bit like SpotLight on Mac OS but much better. But I didn’t buy it since it was also very expensive.

Windows is like heaven to a programmer. There are so many tools to chose from and the documentation is wonderful. I think the API’s are on average very good. There are some API’s I simply don’t understand, like the one to support inline input, but those are the exceptions.

Porting to Windows and Linux

I have to admit your decision to port Pepper 4.0 to Windows surprised me. I had expected that Pepper 4 would still be only for the Mac (and possibly only for Mac OS X) and that your efforts would have gone toward new features.

I had hoped to find new sources of income. That more platforms would generate more money. How often do you hear BBEdit users hear talking about how great life would be if they could use BBEdit on Windows? I thought to have a market there.

People ask for a Windows port of BBEdit all the time — it is a very frequent request.

I thought so.

But what these people don’t understand is that much of what makes BBEdit special is Mac-specific — like its robust support for AppleScript and Apple events. And also, that there is no magic compiler that allows you to flip a switch and generate a Windows program out of Macintosh source code. The API’s for everything are different, including all the little things like drag-and-drop. It might be relatively easy to create a very basic, generic cross-platform application, but I think it’s exceedingly difficult to do so for powerful, professional software. And I think that’s why the companies that do so are very large.

I know exactly what you mean.

Big companies like Adobe and Macromedia have successfully ported Mac-only software products to Windows, but I can’t think of any small shareware developers who have done so.

Neither can I. :-)

Was Pepper 4 at all successful on Windows?

Relatively, yes. I didn’t get too much attention in the Windows world, it is difficult. Tucows refused to list Pepper, for example. But given that, the number of sales were not disappointing.

For someone who hasn’t spent much time with modern computers, Windows and the Mac may appear somewhat similar — they both use windows, icons, menus, and a mouse pointer. But if you’re an advanced user, there are very big differences in the way Windows and Mac user interfaces are designed.

When I look at screenshots for other Windows text editors, like UltraEdit and TextPad, they look very “Windows-y” to me: toolbars with small, cryptic icons, a tabbed interface for multiple open files in one Window. Pepper, however, looks and feels like a Mac application. Mac users are notorious for rejecting software ported to the Mac that stills retains a Windows-style interface; do you think Windows users feel the same way about software with a Mac-style interface?

Could very well be. But I never bothered too much with that. After all, Pepper was not a Mac program either since it started life on BeOS. And I got lots of complaints about small things from Mac users as well, like the toolbar.

You mean the way that Pepper’s toolbar buttons work —  click-and-hold to get a menu, click-and-release to get a button action? I thought that was a little weird, too.

Yes. I copied that from Outlook Express, so I wasn’t even the first on Mac OS to come up with that. It is weird at first, but very useful when you’re used to it.

Pepper had a Multi-file view [showing multiple open files in a single window], and of course most Windows users were using that mode while most Mac users hated it.

That’s interesting, but not surprising. I think window management is perhaps the biggest difference between Mac and Windows. The MDI Windows interface is so clumsy that the easiest way to work is to use one application window to contain all open documents. I never liked that, and it’s the main reason I find Mac OS X’s Project Builder so distasteful.

It is something you can get used to, but it is very personal. I can work with both but prefer the separate windows. On Win2K and before there was another reason to prefer a single window, since each window would get a tile in the Start bar and that would become messy quickly. WinXP has solved this nicely (by having a look at BeOS).

I agree that many Windows programs look awful, but I see things improving on that side of the fence. I think it would be easier for a Windows user to start using Pepper than it would be for a Mac user to start using UltraEdit.

I agree. I’ve also observed over the years that numerous programs go from the Mac to Windows and are successful there (like Adobe’s entire product line), but that Windows software that gets ported to the Mac almost always fails. Remember Lotus 1-2-3 for Mac?

Yup, that was the shortest lived Mac application I’m aware of. :-)

Corel is hanging in, but I’ve never met anyone who uses their software on the Mac.

My guess is that most serious Mac users care deeply about the quality of the interface of the software they use, especially things like being intuitive and having very good fit and finish. Most Windows users don’t care about such things at all, and instead care more about the number of features.

Most maybe, but not all of them. And there are really good programs on Windows showing up. I’m using The Bat! for email right now which is a very good email program, similar in concept to Mailsmith but much more powerful. The looks may not be optimal, but the feel is really great.

The port to Linux was an even bigger surprise to me. I can’t think of a single other commercial Mac application that’s been ported to Linux.

The Linux/X11 port was easy. Since that platform was already OS neutral it was simply a matter of creating the right bridge implementations.

For text editors in particular, I always figured there was no market whatsoever for commercial software on Linux. For one thing, anyone who runs Linux or BSD on their desktop and has the need for a good text editor is almost certain to be a hardcore, incorrigible user of either Emacs or vi. Of the remainder, there is a deeply ingrained predisposition towards free software, especially in the free-as-in-beer sense.

Did you expect Pepper to be successful on Linux? Was it?

I had hoped that with the new popularity of Linux and FreeBSD there might be more and more users coming from Mac OS or Windows who would like to use a more comfortable editor.

But I was wrong. I sold three copies and one of those three was a fraud. I did have thousands of downloads though.

Wow, that’s unfortunate. But I can’t say I’m surprised — the only commercial Linux software that seems to sell are things like 3D rendering and animation software — packages that cost many thousands of dollars.

I am afraid that will be the future of the entire software industry. Eventually it will only be possible to sell huge software packages with lots of support; all the small apps will be open source in the end.

If you had it to do all over again, would you still have gone cross-platform?

Yes. As I said before, I don’t like to be locked in. I have been an OS refugee for the greater part of my programming life and I’m glad I now have this framework. I can take my editor with me wherever I might go.

Pepper vs. BBEdit

Was it your goal to make a full-time living selling Pepper, or was it only a part-time endeavor?

When I started programming ten years ago it was my dream to ever have a company doing shareware that would allow me to make a living. But that dream has proven to be an illusion for me.

Pe was quite popular on BeOS but the market was far too small to make a living. On Mac OS things were a little better but not much.

That is, until Mac OS X arrived. At the introduction of Mac OS X 10.0, Pepper was sooner to market than BBEdit, and in two weeks I made an awful lot of sales. But when BBEdit [for Mac OS X] was released, this dried up immediately and what remained was not enough to make a living.

But you knew back when you started work on Pepper 3 that BBEdit was the leading editor for Mac OS. In fact, many of Pepper’s features seemed to be geared toward weaknesses in BBEdit 5.1 (which was the current version of BBEdit when Pepper for Mac OS first shipped) — better PHP support, better support for customizable syntax coloring, rectangular selections, and a more powerful regular expression library.

BBEdit 6 and 6.5 addressed many of these shortcomings. How do you think the current versions of Pepper and BBEdit stack up against each other?

BBEdit has a lot of features geared towards HTML coders but technically I think Pepper is superior. BBEdit still uses a Handle to store text internally using the gap array approach. Pepper uses a transaction mechanism. BBEdit did however learn a lot from Pepper and borrowed many features and the difference isn’t as big as it was when Pepper first came out. I still think Pepper is better though :-)

No doubt you prefer Pepper. :-) But I’m curious which features you think BBEdit borrowed from Pepper.

There were many, I compiled a list once and threw it away again. I didn’t care that much, after all I started writing a BBEdit clone in the first place. But it was sour to see my advantage disappear.

How does Pepper’s transaction mechanism benefit the user?

True unlimited undo, even when plug-ins or Replace All have been used; and this undo doesn’t use memory, only disk space. And when Pepper crashes you can reopen your work the next launch and you get the full undo history back of your previous session.

BBEdit’s undo works after Replace All, and for plug-ins which specifically support it. But it is memory-based.

I know how it works, I did that in Pe as well. But the transaction model is way easier to program. I could cut thousands of lines of code after introducing this change.

Pepper can open huge files, try editing a 650 MB file with BBEdit.

You could do it if you had 1 GB of RAM, but admittedly, memory-based editing gets unwieldy for files larger than 500MB or so.

When you decided upon the feature set for Pepper 3.0, how much was based on creating an editor that compared well against BBEdit, and how much reflected your own desires?

I never considered BBEdit at that point. I was more focusing on getting Pe’s features on Mac OS. Pepper 3.0 had almost everything Pe had and not much more. I believe Pepper 3.0 couldn’t even print.

There are not many serious programming editors for Mac OS. Other than BBEdit and Pepper, the only others are CodeWarrior and Project Builder. The CodeWarrior and Project Builder editors, however, are clearly designed to be used as part of their corresponding IDE’s. And with Pepper discontinued, that leaves BBEdit as the only serious dedicated editor for Mac OS. Do you think there’s room in the Mac OS market for another serious text editor?

No, I don’t think so.

Was it your goal from the outset to appeal to BBEdit users? Or did you instead try to appeal to people who didn’t like BBEdit?

:-) I think I wanted to show to the BBEdit users that there was room for improvement. I got a bit tired with people shouting that BBEdit was the best program ever on Mac OS while it simply started sucking in my eyes. I mean, when I had to use other editors on BeOS I discovered many new ways to do things that couldn’t be done with BBEdit. Bare Bones was getting arrogant and I thought I wanted to be the little boy shouting that the emperor didn’t have clothes.

But to answer your question, I was aiming for both.

Scripting

As a long-time BBEdit user, Pepper’s complete lack of AppleScript support would have been a deal-breaker for me, even if I hadn’t been working for Bare Bones Software. Pepper supported basic text filtering with Perl and other shell languages, but had no macro language or internal scripting.

Did you ever consider adding AppleScript support to Pepper? Was it a frequent feature request from your users?

It was an often asked for feature. I never planned to support AppleScript but I did plan to make Pepper scriptable. I even started working on an interpreter and other needed internals. My goal was to come up with something comparable to JED or VIM.

Comparable, or compatible?

Comparable of course. Way too dangerous to mimic an OSS package completely.

One advantage to AppleScript is that it allows you to tie multiple applications together. Custom application macro languages are usually only useful within that one application.

Yes, but AppleScript is Mac only and so it was no option in Pepper. And I thought AppleScript was rather slow and not very useful for more than very simple tasks. I wanted my scripting to take over lots of the logic internally, strip Pepper of all the fat and put it in scripts. You need a more powerful script engine for that.

Did you consider using an existing cross-platform language, such as Perl or Python?

I’ve been thinking about using Java.

Did you have any other ideas for major features that you didn’t have time to implement?

Oh yes, many. As I mentioned before scripting was very high on my to-do list, but there were other ideas as well. My to-do list had some 200 items on it ranging from simple cosmetic bugs to things like scripting, folding, makefile generator, etc.

Your only other product I’m aware of was Sum-It, a spreadsheet for the Mac. What happened with Sum-It? Did you have any other products?

Last year, after releasing Pepper 3.6.6 I wanted to do something else. I had always loved Pascal and Pascal-like languages and I thought let’s try to write a Sum-It in Pascal. That went remarkably well and in no time I had a new version. But unfortunately MW’s Pascal compilers are not really good enough and so I ported it to C/C++. And then I thought why not use a C++ framework and so I started working on this framework I always planned to do and eventually I came up with Pepper 4.0 :-)

Maybe I will finish that Sum-It one day.

The old version of Sum-It supported AppleScript, which led me to wonder if Pepper ever would.

Never had time to do it in Pepper. There’s always so much you still have to do and so little time.

Indeed, that’s the hardest part of software development. If you don’t cut good features from the list, you’ll never finish the job and actually release anything.

Exactly.

Future Plans

What OS are you using now?

I’m writing this on a computer running Win2K. At my new job I installed WinXP while all the others are running Linux. :-)

If you would have told me a year ago I would be using Windows right now I wouldn’t have believed you. But now that I use Win2K/XP daily, I have come to like it. I really believe WinXP is the true successor of the old Mac OS. It is fast, stable and if you pick your software well it is also easy to use. Personally I also prefer that Windows look over Aqua.

Are you really done with Pepper? Do you have any plans to sell the product, or let someone else take over? Whenever software is discontinued, people ask the developer to release the source code. Is that something you’re considering?

If someone shows up and offers me enough I might sell it. But if that doesn’t happen chances are high I will continue working on it for the simple reason I’m using it myself daily. I’m not going to release the code for Pepper right now. I have released other software in the past like Sum-It and Pe. Sum-It never took off and I’m curious to what will happen with Pe.

Maybe I will even try again with Pepper, but if I do I will look for a different market.

So even if you re-introduce Pepper in the future, it won’t be for the Mac?

If I reintroduce Pepper, it will be in a completely different form. I’m thinking more in the line of that hugely expensive Windows text editor.

SlickEdit?

No, CodeWrite. Costs around US$300 a piece.

Never heard of it, I’ll check it out. SlickEdit is around $200, I think, and seems to have a devoted following.

And I wouldn’t market it for consumers but would try to place it vertically as they call it. Selling hundreds of peanuts is lots of work and the result is, peanuts.

I don’t think it’s any coincidence that application software from Microsoft, Adobe, Quark, and other very successful companies costs many hundreds of dollars. A lot of their competitors, who sold software for much less money, are now out of business.

Yup. That’s a lesson I learned the hard way. Well, it was a gamble, I had to choose between a high and a low price. I lost.

What are you working on now? Are you doing Windows development?

I’m no longer developing software. I’m a database manager at the University of Nijmegen at the department for bio informatics. My job is to keep the databases up and running and to make sure they will continue to be up and accessible in the near future. This is a very interesting job since we will have 1.5 Terabyte of data by the end of this year and it is doubling every 10 months or so.

Completely different, but for the first time in ten years I have spare time again. Last week I finally finished reading a novel I started five years ago, and even had time to read another one immediately after. Great.