Sometimes I get things wrong. Usually it’s minor errors that come down to a careless moment or something that got wedged between the teeth of my rusting steel trap of a mind. Luckily, you folks who read what I write almost always come through to correct me when I make mistakes or even when I overreach. Something like that happened with the most recent article I’ve written, but it had causes a little bit more complicated than one of my usual attacks of boneheadedness.
Virtually all of the articles published about Melbourne House and The Hobbit — of which, unsurprisingly given the game’s immense popularity, there were quite a few — describe it as largely the work of Philip Mitchell, who wrote it with the aid of Veronika Megler and Stuart Richie. These are the sources which I relied upon to write my story of the game’s development. Shortly after I published my article, however, Veronika Megler contacted me to tell me that the contemporary sources are, simply put, false. She told me that hers was the primary mind behind the game, that Mitchell developed only the parser and handled the porting to the Spectrum and the addition of the pictures after she had left Melbourne House. Richie’s work, meanwhile, was theoretical rather than technical and played little actual role in the finished game.
I was of course quite nonplussed to hear this, but Veronika’s descriptions of the game’s development and the role played by everyone were so precise that I immediately tended toward believing her. That belief only strengthened as I talked to her more. Today I believe that the official story found in the magazines is a distortion (at best) of the facts.
It’s not difficult to understand how this could have happened. The story of The Hobbit‘s development started to be widely disseminated in the computer press during the lead-up to publication of Philip Mitchell and Melbourne House’s next big adventure game, Sherlock. Thus the pieces in question functioned not only as retrospectives, but — more importantly, at least in the eyes of Melbourne House — promotions for what was coming next. It sounds much better to speak of “the next game by the architect of the hit adventure The Hobbit” than “the next game by the guy who assisted the architect of the hit adventure The Hobbit.” Thus Mitchell’s role was vastly overstated, and Megler’s correspondingly reduced; in effect the two swapped roles, with Mitchell becoming the architect and Megler his assistant. As readers like me took those original articles at face value, this version of events passed down into history.
That’s unfortunate, and I understand Veronika’s frustration at having been effectively robbed of credit that is due to her. However, I can also understand how the pressures of promoting the follow-up to such a gargantuan hit could have led Alfred Milgrom and Mitchell down the path they took. I will also just note for the record that Veronika feels strongly that sexism also played a role in the downplaying of her contribution, although I’m not prepared to levy that accusation myself without knowing the people involved better or having more evidence.
Whatever the reasons behind the changing of the record, I’m convinced at this point that Veronika was indeed the major force behind the form The Hobbit took, as well as its major technical architect. I’ve revised the original article accordingly to reflect the true contributions of everyone involved. If you’ve already read it, I’d encourage you to give the new version a quick skim again, or at the least to know that much of what I credited to Philip Mitchell in the original should rightfully have been credited to Veronika Megler. Sometimes, alas, getting to historical truth is a process. I thank Veronika for taking the time to work with me to document what really happened.
I’m actually on holiday as I write this, back in the United States again. So, it will be a couple of weeks before I’ll have more material for you. But keep me in your RSS readers, because we’ll next be rounding the corner into 1983 at last, and things just keep getting more and more interesting.
In the meantime, happy Thanksgiving to my American readers, and to everyone thanks for reading!
November 22, 2012 at 9:15 pm
Wow, I’m not surprised to hear this, but I have to wonder if there are any other corroborating sources you’ve found?
November 22, 2012 at 10:52 pm
Sort of. The author of Wilderland (http://members.aon.at/~ehesch1/wl/wl.htm) has studied The Hobbit’s code in depth, to the extent that he believes he can tell Veronika’s code from Philip’s. His analysis confirms Veronika’s version. This would also account for some other oddities, like the way that Philip talks so much about The Hobbit’s parser and how it works in interviews and so little about the rest of the engine.
I freely admit that this is all far from ironclad, however. In the end it’s a judgment call, and my instincts point me to Veronika’s version of events. If they’re wrong… well, then I guess there may be another mea culpa post in my future. :)
November 30, 2012 at 5:52 pm
I’m a bit confused – is Mitchell claiming to have written the game, not just the parser components? Else you could just ask him.
December 1, 2012 at 11:18 am
To my knowledge Mitchell has not claimed anything publicly in many years. The older sources I referenced in the original article do, however, imply or outright make that claim.
I’d love to ask him directly, but haven’t been able to contact him. If you can help with that, by all means let me know. ;)
January 3, 2013 at 6:46 pm
While developing the ‘Wilderland’ Hobbit emulator I researched the game code thoroughly. Let me first describe what I found, and what this might tell us about the contribution of various authors.
The 40.000 bytes of the game code are composed of three major sections:
*) 25 % are the image data for the various static room graphics
*) The “Technical part” (TP) of the game
*) The actual “Game part” (GP)
The TP consists of the text-engine for input and output, the dictionary, the graphics-engine, and the parser as the heart of the game (which performs lexical, syntactical, and semantical analysis of the user input and therefrom generates an ‘list of actions’ which is passed to the GP. The latter manages databases of all locations and objects (including characters). It effectively generates the physics of Wilderland, with geography, mass, volume and strength of objects; timers for real time events; character profiles for NPCs and rules for what can be done to an object and how it would react upon it.
The code for TP and GP are effectively independent and communicate over two well defined interfaces: the TP presents the decoded user input in a set of simple verb/object1(/object2) sentences to the GP. After processing these the GP tells the TP to produce textual and graphical output and wait for the next input. – Also the coding styles of these two sections are often distinguishable: the TP is very sophisticated, highly effective coded, and very hard to read. While on the other hand the GP is more goal-orientated and sometimes even a little chaotic (clearly aimed at a ‘greater whole’).
And here are some numbers: if we let aside the graphics data (25 %), the character set (3 %), and a number of buffers (8 %), we have 64 % left over. Of these about 22 % belong to the TP (code and data), 36 % to the GP and 6 % could not be identified yet.
In the available online literature it is often stated that Philip Mitchell wrote the parser (the TP in my nomenclature) and Veronika Megler “something else”. This seems plausible. But in my _personal_ opinion the description should be the other way round: Megler wrote a great game which includes a highly sophisticated parser by Mitchell.
As far as I have read, Philip Mitchell never claimed exclusive or major authorship to the game; but neither was Veronika Megler’s part credited adequately. – When I played the game way back in the 80ies, it was the rich physics of the game which fascinated me (and which made me look into the code in the first place, as the message “This room is too full for you to enter” prevented me from going east from the ‘Mountains’). And with English as a foreign language I made little use of the parser’s capabilities and stuck more to the traditional verb-noun commands. So for me personally, Megler’s work was much more relevant for the playing experience; but in the end, both of them generated a great world within the 40 kbyte of the ZX Spectrum memory!
wilderland AT AONdotAT
November 22, 2012 at 9:34 pm
Where are you in the United States?
I hope you are planning to review Sherlock. You said a few months ago that you had not played it yet. Are you planning on playing or reviewing it. Given this new information it be interesting to see how it compares. Also I believe this was the first commercial adventure based on Sherlock Holmes. That might be wrong however. I personally found it a let down. Hope to hear what you think.
November 22, 2012 at 10:53 pm
I grew up in Dallas, and that’s where I’m now visiting family and friends.
Yes, I’m sure I’ll have a look at Sherlock when the time comes. Patience, patience. :)
March 26, 2018 at 6:32 am
So, I realize this is an old post that is mostly there to announce that you corrected the previous post. All the same, I thought I’d point out a typo: “But keep my in your RSS readers” should probably be “But keep me in your RSS readers”
March 27, 2018 at 5:51 am