Z-Machine Invisiclues Project: Timeline

• July 15th, 1998 -- General post to rec.arts.int-fiction requesting help with a project to encode in Z-Machine format all the Invisiclues.
• July 16th -- Steven replies to the message and receives Zork 1's clues.
• July 17th -- Early feedback on that code. Observations:
  1. All formatting was going to have to be done by us. For example, if the clue said something like, "What do _you_ do?", that needed to be edited to "What do YOU do?". This meant that every line of code would need to be at least scanned.
  2. When you selected a hint, it always popped up at least the first hint. In a Beatles'-esque fashion, this was, perhaps, our first creative difference. One side wanted them to be as much like the in-game on-line invisiclues, while the other side pointed out that, if they selected the question, of course they're going to want the first clue (against which the first side couldn't really argue).
  3. The original code didn't remember which hints you'd seen, as a design choice (forcing you to read through hints you'd already seen to see new clues).
  4. Excessively long menu/questions were going to be a bitch-kitten, because "word wrap" isn't really an option if you're trying to have a formatted list of questions, where pressing the down arrow (or NEXT key, or whatever) would always take you to the next question (rather than, say, to the second line of a two-line question).
  5. Point lists were going to be a PitA.
  Later that same day, more Zorks are sent to Steven for editing.
• July 20th -- Obstacles encountered in coding. Word-wrap was going to be nigh-impossible. Debates as to whether or not to edit the original question (to 30 characters or so, to make them fit) ensued. On the one hand, that's how they did the on-line hints in the Invisiclues for the games that had them (neatly side-stepping the whole problem). On the other hand, that seems against the spirit of the project, which is to be faithful to the original while providing the advantages of the electronic format.

  At this time, code is also added to be able to toggle between various styles, such as "classic" Invisiclues (like those found in the Solid Gold Editions), "extended" Invisiclues (which lets you use <gasp> arrow keys), and the like. At this point, discrepencies were noted between the original Invisiclues, the Solid Gold Invisiclues, and the Invisiclues compiled for the Lost Treasures/Masterpieces collections. Flags are dutifully entered into code that will let these different pieces pop up when certain toggles are set.

• July 21st -- The first inkling that the Z-Machine wasn't meant to house 840 pages' worth of information rears its ugly head. Infidel, Enchanter, Beyond Zork, and Hitchhiker's Guide added.

  Also, at this time, certain clues are flagged to be "spoiler" clues -- the ones that actually give everything away. These clues required an extra level of confirmation (which, naturally, could be toggled). The last clue was automatically set to be flagged, but this could be changed manually. This means that all previously done clues need to be re-edited.

  Later that same day, the "spoiler" idea is dropped, since the original Invisiclues were so vague in places that it's often impossible to tell what's an outright spoiler and what isn't.

• July 22nd -- Hitchhiker's Guide to the Galaxy is finished, according to the requirements of the menu library, but isn't working right (illegal op codes from the Z-Machine and the like). It seems like the one game is making the Z-Machine run out of memory. Quite bad.
• July 23rd -- Long menu/question names make Digby cry; there doesn't seem to be any easy solution at this time. Digby takes some time to contemplate the library (which has to be resent from Steven because of massive tinkering).
• August 7th -- The menu system is significantly edited and improved. On the down side, all previously done clues need to be re-edited. On the up side, formatting like boldface, underline, and italics can now be easily used in the clues. Other changes: No more SG/booklet distinction, no more "spoiler" warnings. Steven is happy.
• August 17th -- Digby recontacts Steven after Steven disappears for a couple of weeks. News this time includes a decision to scrap heeding the "original" on-line Invisclues at all (in cases where they differ from the hint booklet), since it's too much trouble.
• October 12th -- Much time passes. Digby yet again contacts Steven, despite the two of them being too busy for their own good. In the interim, Steven has played the Mysteries (Deadline, Suspect, and Witness), removing the only obstacle he was having in editing (since those games, moreso than any other Infocom games, rely so much on not knowing what's going on). A memory management system is released for Inform, and Digby checks it out to see if it's helpful to this particular project.
• October 13th -- It's not.

  This is also the first time when serious questions about where games should end up (in terms of groupings) are asked. Digby promises not to change the menu system such that it requires all previously-done clues to be re-edited.

• January 5th, 1999 -- Digby gets back in contact with Steven, who, yet again, has gone missing. Editing has been done on the mysteries and Hitchhiker's Guide to the Galaxy, which is coming along nicely.
• February 18th -- Digby asks Steven if he is, in fact, actually interested in doing this project. Steven replies in the affirmative; a whirlwind productive schedule begins.
• February 19th -- Steven edits a bunch of the files, and sends them back. Sample transcripts are also added to the hint files. Tres chic.
• February 22nd -- Debate over the Transcript format ensues. Digby wants each line (including the input) to appear at a RETURN; Steven wants the whole thing to appear as a block of text. The idea is tossed out to put up individual games/game groups as they're completed. The idea is also put forth to have the files available on the www.waitingforgo.com/invisiclues web site, rather than sent to ftp.gmd.de. This is done so that they are made available to the public, while avoiding becomng part of the "archive" until the Invisiclues are known to be stable.
• February 23rd --Mystery and SciFi files are edited. Zork Zero Invisiclues are transcribed. The decision to edit out all [sic]s from the original material is made, since the Implementors, in Their infinite wisdom, probably did not intend for typos and grammatical mistakes to litter their works. A comprimise is reached in the transcripts; why not have each RETURN pop up a command -and- the response? The results are quite natural.

  Later that day, it is discovered that the code doesn't work right on ZIP and JZip. Also, the decision to use boldface almost exclusively is reached; it just looks better on all the sample machines tested on. The only exceptions are the "This space intentionally left blank."s, which are left in italics, because Steven insists that's how the Implementors would do it.

• February 25th -- The Fantasy collection is completed.
• February 26th -- The Zorks collection is completed. Credits are re-shuffled so Digby and Steven are nearer the top (since neither one knew what they were getting into when they agreed to this project).

  Later that day, the Zork: Undiscovered Underground hints are found (after some help from the inimitable rec.arts.int-fiction).

• February 27th -- The Adventure collection is finished. There's some debate as to what would be the best thing to do with the Heiroglyphs from Infidel.
• March 1st -- The points are re-edited for all the games that needed it; hopefully they'll display properly (in a pretty table format) on all platforms.

  Also, the web site is begun, although its existance is not announced.

• March 2nd -- Permission is gained from Matthew Murray and Jason Strautman to use Matthew's UHS file of Journey in the collection (since Journey doesn't have proper Invisiclues). Matthew thus becomes the only person to have non-Invisiclue material in the collection.

  ---

  Last updated March 3rd, 1999.