Infocom Test & Development Source Code now available! Thanks and enjoy!
So this is the curated and anonymised USR directory.
The link for the drive is in the video notes, also is a summary of due diligence done which includes reviewing of content from two of the original implementors, Dave Lebling and Steve Meretzky.
No mail files. No personal info. Just ZIL. Youâll need ZILF 0.9 to get the projects to compile.
We now have three public releases of this material (jscott, me, and Adam). Theyâve all been edited to remove personal information, but that happened three times and we made slightly different choices.
(This is not a criticism of anybody; Iâm just pointing it out so that future generations donât say âHey, why are the ZIL compile logs not in all three release collections?â)
Also, catching up, Iâve added Tom Bokâs Hypochondriac to my catalog page (https://eblong.com/infocom/). Also an earlier (1979) version of MDL Zork which Lars Brinkhoff pointed out to me.
Tom mentioned that he made other test games, I think one of them maybe Bath. I started working on that, removed most of the errors, but then time got the better of me.
From what I can see there are at least two other full âtestâ games, and a further 3-4 that may be complete enough to get working.
Iâd quite like to get Trek working, presume itâs a rework of the mainframe Trek game?
The âClassic Text Adventure Masterpieces of Infocomâ collection included a âverylostâ directory containing âa series of files downloaded from the old Infocom UNIX server, which made the journey from Cambridge to Los Angeles but didnât quite make it to our current spiffy world-wide headquartersâ. (Which I guess is whatâs now usually referred to as âThe Infocom Driveâ.)
One of the files is an âOverview of an Infocom Star Trek Gameâ. It doesnât go into any great detail, though. Some of it sound similar to the mainframe Trek game, while other parts donât. For instance, the game was supposed to have multiplayer support (with the caveat that there may not be a market for that), and I donât remember the mainframe Trek game having that.
Of course, itâs possible that the Trek files here bear little or no resemblance to the game envisioned there. It was still an open question if they should go with character graphics (using the technology they already had?), âModify EZIPâ, âReinvent DIPâ (thatâs what Fooblitzky used, I think?), or just âWrite a machine specific gameâ.
I noticed that your version also includes some modified Zork I code. That looks like itâs probably the Zok game Tom mentioned in the other thread. It definitely has a Tiny Turnpike Room, and Evil Kneivel [sic] on a very small motorcycle.
I think âBathâ was another game idea I was screwing around with, involving a miniaturized protagonist trying to survive in a bathtub full of toys (and an enormous kid). All of these fragments eventually turned into adventure stories I told to my kids decades later. I had forgotten that I started actually coding this one.
I wouldnât label any of these âgamesâ â more like âplayable game fragmentsâ. And I donât think I would lump them in with the abandoned work of actual Infocom game developers. I was just a teenage tester, and my âgreatestâ ZIL coding accomplishment while at Infocom was a lowly Sampler (playable mash-up of bits of four previously published games). Iâm not trying to be humble â just stating the obvious. Itâs like if you found a box of previously unseen sketchbooks from Da Vinciâs workshop and some stick figure drawings from his teenage nephew got accidentally mixed in. : )
One other thing to mention â to help those of you who are spelunking in these test/dev directories: many of these test/dev games werenât built from scratch. For ZIL noobs like us, it was much easier to copy the source from an existing game and strip out the parts you didnât need. That meant your code was more likely to compile and already had a lot of the standard vocabulary and verb responses. Even some of the real implementers did this, only more elegantly. So the existence of bits of one game inside another was pretty common. This makes it harder to use just the source code to distinguish one game from another â you had to compile and play it.
Tom, I disagree with your humbleness. I rephrase it ââŠdrawings from one of his apprenticesâ (whose IS your status back then, you should agreeâŠ)
For sure outcast should have evolved into an actual Infocom game. Bathtub is definitively non-commercial grade, but is solid material in the context of the IF community. I estimate that you have (and strongly suspect you still have) potential for entering into the first 10 places of the IFComp.
on the lot, as I have pointed elsewhere, the Wishbringer directory, apparently out of place, seems to be a SG edition of wishbringer, but I havenât started to delve into the actual content, aside the telling hints.zil.
When Hypochondriac first did the rounds the general consensus was (and Iâm quoting @robinjohnson here so he can confirm!) that it was Infocom-level but maybe without the final QC polish ⊠In other words it was viewed as being Imp standard.
Donât knock it, you were 15 at the time, the game is crazy good all things considered!
One thing thatâs been pointed out is that the source code Adam released has a separate Wishbringer directory that contains the Solid Gold version. It does seem to be a more up to date version than whatâs in your archive, because it has the storybook that was included as in-game object (presumably to save on printing costs).
So maybe thereâs still hope that a working copy of that game can be compiled? (From what I remember, the other source code compiled but the game itself was broken. I could be wrong about that.)
Answering my own question, the Solid Gold version of Wishbringer thatâs currently in the Obsessively Complete Infocom Catalog does not compile, since itâs missing several things. In fact, it seems to be just the V3 source with minor modifications and an added hints.zil file.
The version in Test-Dev compiles in Zilf with only minor modifications. I had to change the names of the included files to lower-case, specify that the game was XZIP, and remove a call to <USL>. (The Z-Machine specification notes that the game accidentally uses the V3 show_status opcode.)
So it looks like it is the real thing, in which case it definitely would seem to belong in the Obsessively Complete Infocom Catalog.
A minor point on cleaning the directories: isnât only the privacy issue, but also cleaning from own experimenting.
I noted that in the Boggle, Hypocondriac and Trek dirs the presence of release 0.8 zilf/zapfâs binaries and DLLâs, whose clearly donât came from The Drive image(s)âŠ
@Piergiorgio_d_errico Just to answer the above, you might find some of the ZILF 0.8 files in some of the game directories, thatâs me being a bit untidy Iâm afraid! I had a few initial tries at compiling them - so youâll see some âdebrisâ from that!
âŠor, if Iâm wrong, could you let me know where youâre looking?
Your collection has wishbringer-r69.zip. This roughly corresponds to Test-Dev\Old\wishbringer. It can be compiled into what seems like a working .z3 file with only minimal changes.
Your collection also has wishbringer-invclues.zip. This corresponds to Test-Dev\Old\wishbringer\cheap. It canât be trivially compiled. Itâs missing several things, e.g. the INIT-STATUS-LINE, the H-BOLD constant, etc. Itâs also missing the storybook that was added to the Solid Gold version.
If you compare these two versions, youâll find that there isnât a great deal of changes between them. It looks like someone just took the V3 source code and added a hints.zil file.
Adamâs collection also has a Wishbringer directory at the top level. This contains what appears to be a working Solid Gold version. It has a lot of changes compared to the non-working version: It has the storybook, the parser messages use brackets instead of parentheses, etc. Most importantly, thatâs the version I managed to compile earlier today.
While writing this post I also noticed that Adamâs collection has a Journey directory at the top level. This, too, is different from whatâs in your collection. From a quick look, it might be the source code for journey-r46-s880603.z5. The .z6 versions all have the text âglowing cavern wall. Not ten paces awayâ, but the .z5 version - and this source code - has âglowing cavern wall. Not fifty yards awayâ.
Looking closer, the Journey folder is probably not that particular .z5 version. It mentions illustrations, and the enemy is called âThe Dread Lordâ like in the final game, and not âThe Evil Oneâ. (I donât know when that change was made. In the earliest advertising I saw for the game it was still âThe Evil Oneâ.) Still interesting, though.