Looking back, my games that were the most joy to work on were text-based. Even as they were also my least successful. From text adventures through tabletop RPGs and to dungeon crawlers, whenever I was able to blend writing and games, it made me happy.

Being a writer at heart might just have something to do with that.

Oh, the road was bumpy. At one point, having grown disappointed with parser-based interactive fiction, I ran off in search of alternatives. A quest that went well for the most part, except for losing something along the way and not even realizing until recently.

Maybe the problem wasn't with parser games as a medium after all, but as form: the neat little puzzle boxes, polished to death and lined with empty words that every single one was expected to be during a certain era.

Funny how old-school treasure hunts and exploration games came back into fashion once the parser lost its cachet and its remaining fans felt free to experiment again.

And you know... turns out I haven't lost all patience with the parser, as setting up an interpreter on my phone demonstrated. It's just that out of all the games I played and loved in the past, there's just a handful I care to try again. Like The Fire Tower, or The Moonlit Tower: relaxed, meditative pieces with no real stakes that I can jump into and type a few commands now and then simply to pass some time.

Clearly my other attempts in the intervening years failed simply for lacking focus. But that's not the whole story. Interactive fiction moving to the browser wholesale made it more accessible to everyone, but only at a hefty price: now it can only be played on graphical operating systems with gigabytes of memory at their disposal.

Never mind issues like preservation, or waste. This changes how we think about text adventures. Some people now balk if a game as low-tech as they come doesn't have a fancy smooth scrolling effect!

Typography and illustration can greatly enhance the written word. But like in web design, nobody's going to read a gorgeous book whose text is all Lorem Ipsum.

At least it's still possible to open a lot of websites in a text-based browser and pluck the text from among all the pointless glitz. With a game made in Twine, you can't.

Now what would be the interactive fiction equivalent of a text-based browser?

If you answered the Z-Machine, there's still hope for you. The humble, "obsolete" Z-Machine is still around and not going anywhere as long as people still want to play the classics. It can run on toasters, with room to spare, made to look beautiful, adapt to the host platform and last but not least help out players.

That's a problem, you know, since the Z-Machine is fundamentally parser-based. But a whole other problem. One that won't be solved unless we keep working on interpreters and experimenting with the user interface.

Until then, a more urgent question is: if I was to take a trip back to my roots and make a new game for the Z-Machine, what game should it be? Part of the answer can be found above: one I can play to relax. One with a nice world to wander for a bit. As for what to do in it, ever wondered what treasure in games is for? Three words: Easter egg hunts. Not good enough? Have two more: progress indicators.

So is clothing, otherwise an underused mechanic in text adventures; luckily, RPGs know better. And speaking of RPGs, how about stats, combat and magic? They don't even have to be random. Just to allow for self-expression better than puzzles with set solutions ever could. Because, you see, that's what text adventures have lacked badly ever since well-defined protagonists were deemed inherently superior.

It's the difference between telling a story together with the player and talking at them.

All right, so for the sake of the argument let's pretend I decided to make a big open world adventure for the Z-Machine. What tools could be used for that purpose?

Perhaps surprisingly, out of the three big options available in 2020, I'd still go for the good old Inform 6. It even offers me an upgrade path, via Glulx and its multimedia capabilities. They're easy to use. I've done it with no issues for my very first (and so far only) game made with the language. Which, I should point out, also isn't going anywhere while its successor Inform 7 uses it for a foundation.

Wait, wasn't Inform 6 kind of cryptic? My old game's source code certainly is. But then again it was my first, it was 14 years ago as of this writing, and I insisted to do a lot of things in nonstandard ways.

No, a better question is who would play it. And apart from the fact that it would be a personal game, made for me first, there's always bringing it over to another platform if I need extra validation. Making money isn't on the table anyway, so why worry.

Nowadays we fall much too easily into this trap of wanting everything we do to be yet another Big Serious Project. So creative work becomes a chore before we're even started. Then we wonder why burnout follows so quickly.

Even if your goal is to sell games and make a living at it, how are you ever going to progress without prototyping, or the occasional diversion? Relax! Your next project will take years anyway. Set aside a few months for yourself. Consider it preproduction.

Please let art be art. At least sometimes. It matters.