What is version r39 of Planetfall? (and why does it have a bug?)

Playing
1

I wanted to try playing some Infocom games! I’ve poked around in Zork 1 and Hitchhiker’s Guide (when the movie came out) before, and that’s been it.

I started with Planetfall, and I was enjoying it. I could generally appreciate certain design decisions–yes, even the hunger timer! Though uh, maybe less appreciation for sometimes dropping stuff when picking up other items, or having to use access cards so many times. I’ve seen lots of references to Floyd before in NPC design discussions, so it was neat finally encountering them. Overall, it holds up well: really good sense of exploration, solid puzzles.

My character dropping stuff all the time did make me notice something early on though, which is that the stuff I dropped wouldn’t get listed in the room description. Hmm I thought, that’s a strange design decision!

I kept playing, and eventually I got stuck on what to do with an item. And looking at Dan’s invisiclues linked from the ifdb, then the Clubfloyd transcript and even a youtube playthrough, I confirmed that in my game, the contents of containers wasn’t getting listed properly. So a cardboard box that contained stuff was displayed as empty, and an opened access panel in another room was also shown as empty. This was making my game rather difficult! I finished the game with the invisiclues guiding me on some stuff I wasn’t seeing.

I was playing version r39. I see afterwards that in the historical source github, someone did submit an issue ticket for the same issue. Two other reviews on IFDB also mention encountering this r39 bug, so I don’t think it’s Gargoyle specific. Clubfloyd played r37 and didn’t have the bug.

So I’m curious where r39 came from, from a historical standpoint.

Historical source github says it’s “final revision,” zarf’s Infocom catalog says r39 is “final-dev” as opposed to r37 which is “Masterpieces version, PC.” Was this a version that Infocom devs worked on, was it an internal dev build, was it actually released to the public, was this something fan modified?

My stunning investigation skills lead me to this “final revision” commit in the historical sources github, which came after “revision r37 (original source)”: https://github.com/historicalsource/planetfall/commit/e85ca899aac575e74a4b3845f44d09a891c1563a. I see “Assembling PLANETFALL.ZAP.3 on Sunday, May 1, 1988 13:37:01” in the file changes, which matches the serial number. Fans wouldn’t change the legalese from “trademark of Infocom” to “registered trademark of Infocom,” would they?

Can anyone shed some light on this?

4 Likes
Trying some commercial era parser IF (first up: Planetfall)
2

As I understand it, it was the final version that was worked on - but it was not released, and most of these final devs, I think, didn’t even receive proper testing. Or, like, any. As I understand it, playing these versions may be “interesting”, but they are not recommended for actualy “playing and enjoying”. They did not go through the proper rounds; were not meant for release in that state.

Approach any “final-dev” version with caution. If it’s your first time playing that game, skip it entirely. As I’m afraid you learned by yourself…

5 Likes
3

The historical repos that wound up at GitHub - historicalsource/planetfall: Planetfall, by Steve Meretzky (Infocom) · GitHub (and later my web site) are from an archive of a work machine.

So you would expect that the “latest revision”, whatever it is, is just “whatever state the source was left in when everybody got laid off.” There’s no particular reason to believe that it works at all. Someone might have been in the middle of fixing a bug.

That’s what “final-dev” means.

8 Likes
4

Oh, no… I think IFDB is linking almost exclusively to “final-dev” versions on https://eblong.com/infocom/. (And that’s my doing.)

EDIT: I’ve updated IFDB links to point to my best guess as to the canonical versions, usually Masterpiece.

Is “final-dev” your term? The fact sheet just uses “F” which it defines as “unreleased final internal versions.”

I think it’s unwise to use the expression “Final-dev” here in the “Games files” list, which will inherently be read out of context of the rest of the document.

Perhaps call it “Unreleased-final-internal” instead?

3 Likes
5

Yes.

The problem is that I don’t know for sure what was released. A “final-dev” version might be the last-released version.

I thought about “dev”, which might more intuitively convey “stay away”. But I don’t know which of the earlier versions were released either. Maybe several of them are “dev”. (I’m pretty sure about the “alpha”, “beta”, and “gamma” labels at least.)

6

I’m fairly sure that the Fact Sheet is working off of the same set of raw data as I am, even though it’s dated twelve years before the historical-source dump.

7

What’s the raw data for what is/isn’t F in the fact sheet’s terms?

8

If it helps… it doesn’t, not to me. In fact, to me it’s more like “director’s cut”: the dev’s intended version.

9

I guess what we actually need is a way to tag one release as ‘this is the version you should play’. It sounds like it might not be possible to deduce this from the flags, but there are a relatively small number of games in total–can we just flag one each as ‘probably best’ by hand?

10

Though I don’t think it’s written down anywhere, the Masterpiece releases are recommended. I recommend them, anyway.

scratch that, there’s this at the Obsessively Complete Infocom Catalog:

The Text Adventure Masterpieces of Infocom CD (1996) is the source of most modern releases and downloads. If you want to play the “official” version of a game, the Masterpieces version is usually the right choice.

Those versions are tagged at OCIC

1 Like
11

Ah okay, thanks for the info, figured as much once I saw it labeled “final-dev” on the Catalog while looking into the version I had.

And Dan fixed some of the ifdb links for other Infocom games so they don’t point to dev builds, great!

1 Like
12

They are tagged “Masterpiece” but the hint that “Masterpieces is what you usually want” is very far away.

Perhaps we could rename final-dev to internal-unreleased?, literally containing a question mark. We don’t know for sure that every final-dev was never released, but there’s good reason to think that none of them were, and we do know that it was scooped up from an internal source drop.

And… we could stand to add the preamble sections to the table of contents. Right now it just starts with “Original Zork,” but if it started with a list of sections like this:

  • The collection
  • Disclaimer
  • Differences from the GitHub release
  • On Z-code and ZIL
  • Canonical versions
  • Some notes on the files

… that would have been helpful.

13

How about internal-recovered to convey “this version was found on a hard drive somewhere but we don’t really know what it is”?

3 Likes
14

That’s true of all of the known Infocom source code!

2 Likes
15

See above. I think both the Fact Sheet and I just surveyed the collected list of source directories and marked the last one “F”.

Good point.

I need to fix my chronological order, too.

16

Can we mark games that are known to have been released and sold, leaving others as “status unknown”? That seems sufficient to direct people to less buggy versions without selecting a single blessed version. If evidence appears that shows a previously unknown game was in fact released, it can be marked as such.

17

I managed to miss this one until now. Question: when it says here that “Versions marked “final-dev” are unreleased final internal versions (according to the Fact Sheet). That is, they had changes in progress when development shut down. These may fix bugs, but they never went through QA, so they should not be considered release-quality.”, does that not mean that final-devs are not, indeed, the latest released? Not everything has a “final-dev”. If every game had a “final-dev” that question might come up more clearly. It seems that the ones which have a “final-dev” are usually preceded by “masterpieces” or “solid gold”.

Unless I misunderstand, a not uncommon occurrence.

18

I think there are two types of people visiting: people doing research and people playing games. The former will read the page and consult the fact sheet, while the latter just need to know which game to play.

Saying somewhere near the top that people looking to play should download the Masterpieces version might be the most impactful change one could make. Perhaps it ought to be repeated, even.

I feel most people will find these games via IFDB, though I guess we can’t really know. I see Dan has been updating the IFDB pages; if there’s more work to do there I’m happy to pitch in. Or perform other tasks related to Infocom’s IFDB pages.

19

As historian, I suspect that is a sitz im leben issue, so perhaps the best course is an interview with an Elder Imp, asking about the internal organization of the company.

and said internal organisation should be compared to The Drive’s internal organisation: I have here, via bitsavers, a good chunk of DECsystem documentation, and I (and Zarf et.al.) should look into it, but what matters is who back then has access to the in-development story files. For example, read access to story files to other users (I use the *nix parlance here) is a pointer toward internal betatesting.

So, what was the practical dev cycle inside the company ? and the file’s permissions (again, *nix parlance) can be related to said dev cycle ?

HTH and
Best regards from Italy,
dott. Piergiorgio.

20

I don’t believe file metadata (ownership, permissions, timestamps) have been preserved at all.