The long-awaited version of Inform 7 finally popped up a few weeks ago (June 12, to be precise), after more than a year of development. It soon appeared that there were Problems. So on July 1, a maintenance update was released. That would be version 6E72.

Tonight I downloaded 6E72 onto my MacBook Pro. Surprise — it doesn’t run at all. The IDE is broken. This has been the state of affairs for almost a week now. I’m sure Andrew Hunter is scrambling around trying to find the problem. Or at least, I hope he’s scrambling around trying to find the problem. According to the bug tracker, it doesn’t show up under 10.6, so if he has upgraded his Mac to the latest OS (as all good computer geeks do), his options for testing and debugging may be somewhat limited.

Here I was, contemplating (with only mild misgivings) the idea of sitting down and really learning I7. Not just the basic authoring code, which I can hack my way through already, but what’s going on under the hood, what are the cool extensions, all that.

Talk about letting the air out of your tires.

A couple of days ago, using my Windows machine and 6E72, I compiled “A Flustered Duck” with the line, “Release along with a website and the “Quixe” interpreter.” In theory, this is a spiffy feature. Quixe is a Glulxe interpreter written in Javascript. It enables large Inform games to run directly in a Web browser. All you have to do is add that line to your game’s source code and the IDE will create a web page (called play.html, but you can rename it) that you can put up on a website to enable anyone to play your game without the trouble of downloading an interpreter program.

It will run slowly in a browser window. The response is very noticeably sluggish. And the game certainly won’t look as nice as it would in Zoom on the Mac. But if those were the only problems, I’d still be pretty happy, because putting a game on a website is a nice way to introduce more people to IF, and to your own work.

Ah, but those aren’t the only problems. “A Flustered Duck” happens to have a longish intro. Not extremely long — it’s eleven short, punchy paragraphs. Trouble is, when the game runs in Quixe, the intro scrolls up off of the top of the window before you get a chance to read it. Yes, you can scroll back, the text is still there, but one doesn’t want the software to blip right past the opening of the story.

Plus, when I run the game in Microsoft Internet Explorer 8, every time I go from one location to another I get a stream of dialog boxes that say, “Stop running this script?” IE thinks the script has been running for too long, so it repeatedly asks the user if the user wants to bail out. This is a known problem, but apparently it’s peculiar to IE; other browsers don’t complain, they’re just sluggish. So even with Quixe, which is designed to let people play IF in a broswer, my game is unplayable in the most popular browser.

Can somebody please explain to me how we’re making great strides here? I’m getting a little mixed up.

I do understand that Inform is free software, and that all of the developers are unpaid volunteers. But even with free software, you’d hope to see a little more testing prior to release.

The Quixe problem in IE is in a slightly different category. Andrew Plotkin is a very bright guy — brighter than me, I’m sure, and also a whole lot more knowledgeable about computer programming. What I’m asking myself is, was it realistic for him to spend weeks of his free volunteer time writing a Javascript interpreter that wouldn’t work in Internet Explorer?

Okay, people can download Opera or Firefox or Chrome. Quixe is genuinely useful (or would be, if it didn’t scroll past the intro). But still … It’s all very well to bash Microsoft. I’m sure there are all sorts of bad, corporate reasons why IE is so lame. But I don’t think that matters to the game player. Millions of people use IE, so how much sense does it make to write a big, complicated app that won’t work in IE?

At the very least, shouldn’t the Javascript be written to detect that the user is attempting to run the game in IE, issue an apologetic error message (“Sorry — Internet Explorer is too slow to run Quixe”), and stop? Why subject game players to such a sucky experience? Yes, individual authors can and should put a cautionary note on their web pages, saying, “Oh, by the way — if you’re using IE, don’t bother clicking on this link. Go get a different browser.” But not all authors will know to do that, because not all authors test their websites in IE.

Yeah, I think I’ll suggest to Andrew that he add an alert message. It would be a courtesy to players.


2 thoughts on “Down the Rabbit Hole

  1. Can somebody please explain to me how we’re making great strides here? I’m getting a little mixed up.

    Isn’t getting the whole thing to even run on a browser a great stride? You’re talking like Quixe has reached its final stages of development cycle and that it will never and cannot ever run in IE. It’s good that you’re active in making comments and suggestions, but I hope you do realize the way you’re presenting them makes you look like an asshole.

    1. I plead guilty to (a) feeling frustrated and (b) airing my feelings in my blog, for the benefit of anyone who happens to share said feelings. If that makes me an asshole, what would be the way to fix it?

      Should I not feel frustrated, or should I feel frustrated and remain silent about my frustration? Which would you recommend? Or would you perhaps be more comfortable if I acknowledged my frustration, but did so in more boring, less colorful phrases?

      To be honest, your comment leaves me rather agreeing with, I think it’s Poster, who on the newsgroup likes to rag on the subculture that venerates the Old Ones of Inform and feels that the Old Ones should never be tasked with any negativity.

      Optimistically, most of these problems should be dealt with in a few days or weeks. Except for the problem with IE, and it’s easy enough to steer people to a different browser.

      I’ve been very ambivalent about Inform 7 from day one. I like the IDE, but the manual is surprisingly bad, and I’m convinced the idea of “natural language” programming is sort of a con game. It convinces (and is designed to convince) non-programmers that writing IF will be easy, when in fact I7 code is more difficult to use than standard coding syntax would be.

      The mad success of I7 is discouraging, in no small part, because it has left TADS 3 in the dust. The tiny size of the T3 community, in turn, means that Mike Roberts has less incentive to, for instance, port Workbench to the MacOS or develop a browser-based interpreter. He’s been working on the latter, but if he had 20 times the number of active authors clamoring for it, I’ll bet he’d be more motivated.

      My primary motivation here is to have an amusing hobby to pursue while sitting in my easy chair with my MacBook on my lap. Thus, my choices are Inform 7 and Inform 6. The former is not functional at all unless I roll back to an earlier version, and the latter, while fully functional, is a bit old-fashioned and requires Terminal to run the compiler.

      Should I paste a happy face on this situation? Or should I allow a little sarcasm to creep in now and then? Let’s have a show of hands.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s