Ruins of Cawdor Development

Release Info | Releases | Release Credits | Development | Technical | Cover Art | Screenshots | Media | Manual | Contents | Soundtrack | Misc

stub page. You can help the Sierra Wiki by expanding it.

 
 

Game Design Document

Concept Art

None available

Developer Notes

Richard Aronson wrote:

When the let me, er, asked me to make the sequel to Fates of Twinion, one of the reasons was my solid reputation for writing solid code. My prior jobs required I knew how to test else the companies (Fortune 100s) could lose a LOT of money. Longbow shipped with only 14 bugs, none fatal, as my first game in 1991, and my others were similarly clean. I also found some very hard to duplicate bugs that had been fatals in 18 straight games when QFGI VGA caused me to see and eventually duplicate it for Larry Scott. Twinion had shipped with almost 50 bugs, many of them fatal. After 3-4 releases with zero bugs fixed, or so I was told, QA refused to work on Twinion any more. I found the primary bug that caused Twinion to fail so often as part of making Cawdor. To ensure that PVP combat had few packets sent between players (at about a nickel per packet, they could really add up) both Shadows of Yserbius and FoT used a large identical index of over 500 semi-random numbers, but the only number that counted was the pointer into the index, only 2 bytes long. It two players got out of synch (I think the next random number is position 8 in the index and you think it's position 11) then you might think I'm dead and I think you're dead. That can lead to a fatal hang where one of us is waiting for a response the other doesn't know to send. I hope the coders enjoy the technicality enough to make up for the glazed eyes I anticipate.

Well, pretty much all the hang fatals originated during PVP combat. If I tried to run away, but failed, my index got one extra pull. We were now out of sync, and sooner or later that would cause a fatal because our random numbers (and health) were different.

We fixed it in Cawdor. I asked for the budget (time) to fix it in The Shadow of Yserbius and The Fates of Twinion, but all these bridge requests had accumulated while I was making Cawdor, and QA was booked, so I never got a chance to retrofit the fixes before getting laid off. I never worked harder in my life than when I was at The Sierra Network. I still regret that I planned solutions that so many modern online worlds could use but nobody thought of them or hired me to fix them. I don't think I can take on a big project ever again unless I get a new kidney. One might arrive in Q4! Then nobody will ask me to do anything and my darling wife would say "See, I told you so!" She cackles a lot and nobody has ever made me happier than her. We were both TSN programmers and should have left TSN but we were unreasonably loyal. AOL *is* scum.

 

Richard Aronson wrote:

Many of the puzzles in Cawdor required multiple players to solve. This was deliberate to build relationships. I could easily have modified them for an offline game, but never did so. AFAIK (and I should) it would take access to the original source code to make it single player. I may have a copy, with permission, and any TSN programmer could have gotten it from the repository, but those puzzles would need source code and the compilers needed to build the game. AFAIK, I was the only TSN employee who knew how to create all the tables.

 

Richard Aronson wrote:

When the[y] let me, er, asked me to make the sequel to Fates of Twinion, one of the reasons was my solid reputation for writing solid code. My prior jobs required I knew how to test else the companies (Fortune 100s) could lose a LOT of money. Longbow shipped with only 14 bugs, none fatal, as my first game in 1991, and my others were similarly clean. I also found some very hard to duplicate bugs that had been fatals in 18 straight games when QfG1 VGA caused me to see and eventually duplicate it for Larry Scott.

Twinion had shipped with almost 50 bugs, many of them fatal. After 3-4 releases with zero bugs fixed, or so I was told, QA refused to work on Twinion any more.

I found the primary bug that caused Twinion to fail so often as part of making Cawdor. To ensure that PVP combat had few packets sent between players (at about a nickel per packet, they could really add up) both Shadows of Yserbius and FoT used a large identical index of over 500 semi-random numbers, but the only number that counted was the pointer into the index, only 2 bytes long. It two players got out of sync (I think the next random number is position 8 in the index and you think it's position 11) then you might think I'm dead and I think you're dead. That can lead to a fatal hang where one of us is waiting for a response the other doesn't know to send. I hope the coders enjoy the technicality enough to make up for the glazed eyes I anticipate. Well, pretty much all the hang fatals originated during PVP combat. If I tried to run away, but failed, my index got one extra pull. We were now out of sync, and sooner or later that would cause a fatal because our random numbers (and health) were different.

We fixed it in Cawdor. I asked for the budget (time) to fix it in The Shadow of Yserbius and The Fates of Twinion, but all these bridge requests had accumulated while I was making Cawdor, and QA was booked, so I never got a chance to retrofit the fixes before getting laid off. I never worked harder in my life than when I was at The Sierra Network. I still regret that I planned solutions that so many modern online worlds could use but nobody thought of them or hired me to fix them. I don't think I can take on a big project ever again unless I get a new kidney. One might arrive in Q4! Then nobody will ask me to do anything and my darling wife would say "See, I told you so!" She cackles a lot and nobody has ever made me happier than her. We were both TSN programmers and should have left TSN but we were unreasonably loyal. AOL *is* scum.

 

Richard Aronson wrote:

Cawdor used the same engine [as Yserbius and Twinion], with lots of redesign so each character class and species had advantages and disadvantage and fixing EVERY f'ing fatal that Ybarra production had failed to find. That also meant the macros (cheats) no longer worked. All I needed was 2-4 PC's on my desks and some hard work; most took me only a couple of days. There were also no cheat codes for Cawdor. That was the #1 complaint from all the guilds during my planning meetings with them; cheating is killing the RPGs. Turns out, most guilds preferred having the cheats. Turns out nobody wanted to admit using cheats. Oh well. I also balanced all the spells, skills, inventory items, monsters. I put lots of folks into the game, mostly in the Rogues Gallery. One Alpha tester with cancer got a special tribute that got me many internal emails for how I identified her so well while still staying true to the language of the game and the English history I was throwing in there. I got more letters about Cedric, but Cawdor was second by a lot.

Related Links

• Yserbius Series
• Game Development