Tweet
I doubt I'll ever reach 10k posts, so for my 1000th I thought I'd do one of these navel-gazing All About Me threads. No real reason, other than that Takkaria gave me commit access to the V codebase earlier this year (to implement the new randart generator), and if I introduce myself a bit it might help put some of my other changes in context, and give people a chance to vent before they happen. (Takkaria is now back in regular contact, btw, so the chances of me doing anything really stupid to the nightlies are reduced.)
I'm about to turn 40, and have been gaming for over 30 years. The first game I can remember was Space Invaders on the Commodore PET, but not far behind was my first roguelike - the Temple of Apshai on the C64 in the early '80s. At the same time I started playing AD&D, but that only lasted a decade or so (I never got over the changes in 2nd Ed.) while the computer gaming went from strength to strength. At university in the late '80s I was introduced to Moria, which I loved but never beat. I just missed Angband, which existed but hadn't made the journey from Warwick to Cambridge by the time I left.
I got my first PC in '92 and didn't go near a roguelike for the rest of the decade, being far too busy with Civ, MoO, Frontier, MoM and tons of other games. In early 2000 I was looking for a new game and a friend recommended Angband, which was at 2.9.0 then I think - RR had just become the maintainer. Over the past decade I've had the occasional break from Angband (though nothing on the scale of Leon), but usually for reasons of being distracted by other games rather than RL. (Diablo II being the main contender - if you're going to do a real-time graphical roguelike, that's the way to do it.)
I love the fact that there are so many Angband variants, but personally I prefer the ones that tweak the gameplay without overlaying a whole bunch of stuff. So I never really got into the complexity of Z or ToME, and instead preferred 'smaller' variants like GSN and O. I'm a huge fan of Leon's work: O and S are the only two variants I keep going back to. I have promised myself to give Un another try one day (I wanted to like it but it felt too complex - to Angband as ADoM is to Crawl), and I'm saving FA and Steam for a day when the whole genre seems stale and I want some refreshment.
IRL I'm a (non-IT) project manager (well, programme manager now), and most RL projects follow a strict 'cathedral' setup where everybody has a very clearly defined set of responsibilities. So the 'bazaar' model of Angband's development fascinates me (read this essay for an explanation of these terms). I started off wanting to improve Greg Wooledge's randart generator, and it so happened that someone else wanted to do the same thing, so we did it together. The new one was actually ready in about 2002, but RR was never really interested in it.
I've not done any serious coding since a gap-year job writing library functions in fortran77. Like many before me, I've learned what I know of C from the V code. Someone on #angband-dev derided the overly simplistic structure of the code (comment line, line or two of code, blank line; repeat) - but it does make it an amazingly effective learning tool.
I've been using Linux since 1995 (Slackware then RedHat then Debian since ~2003), and when 3.1.0 came out I volunteered to become the maintainer of the Debian package. It so happened that the existing maintainer was keen to pass it on, and kindly agreed to 'mentor' me. (Debian maintainers who are not fully fledged "Debian Developers" need a mentor to upload packages for them.)
So my two primary interests in the code are artifacts and packaging. I'm currently working on revising the packaging stuff: the previous maintainer always built it using --with-setgid, but I want to change to --with-private-dirs, which is now Takkaria's preferred default. A straightforward change from the building/installing angle, but trickier for upgrading and purging. I hope to get this done for the release of 3.1.2
Beyond that I have a bunch of other aspirations, mostly around enhancing my own enjoyment of the game:
- sorting out pref files so I can have the right macros and squelch settings and autoinscriptions loaded for each character, and a monster memory that spans save files. IMO the existing pref file system has not coped very well with more recent additions like autoinscriptions and squelching. Could do an artifact memory here as well, though being a randarts man I'm not too fussed about that (worth saying here that I do think it's important that the artifact set is well-balanced and interesting, even though I rarely play with them)
- improving the knowledge menu so that you can browse last-known store and home inventories from the dungeon. Of course you won't know about any stock rotation since you descended, but I share Eddie's view that anything that can be accomplished by note-taking should be available in the game UI.
- packaging Dubtrain's sound pack so that users of Debian-based distros can install them on top of the main game (Dubtrain's choice of licence is considered "non-free" by Debian, so they cannot be part of the main package).
- abstracting elements into a structure so that they can be modified / balanced / added to more easily (there are currently hundreds of lines of duplicated code which could loop over an array of elements, and this would dramatically improve the monster power function)
- more revisions and improvements to monster and object power ratings (including making non-magical armour useful by changing base AC and weights)
- more changes to randarts (more interesting curses, more intelligent combinations of powers, etc. - this will always be my ongoing project) - including new artifact types if Takkaria goes in that direction (artifact staves and rods, for example).
Well there you go. That's what I want to work on after 3.1.2. This is way too long already so I'll save my views on TMJ, sterile statistics and instadeath for another post.
I'm about to turn 40, and have been gaming for over 30 years. The first game I can remember was Space Invaders on the Commodore PET, but not far behind was my first roguelike - the Temple of Apshai on the C64 in the early '80s. At the same time I started playing AD&D, but that only lasted a decade or so (I never got over the changes in 2nd Ed.) while the computer gaming went from strength to strength. At university in the late '80s I was introduced to Moria, which I loved but never beat. I just missed Angband, which existed but hadn't made the journey from Warwick to Cambridge by the time I left.
I got my first PC in '92 and didn't go near a roguelike for the rest of the decade, being far too busy with Civ, MoO, Frontier, MoM and tons of other games. In early 2000 I was looking for a new game and a friend recommended Angband, which was at 2.9.0 then I think - RR had just become the maintainer. Over the past decade I've had the occasional break from Angband (though nothing on the scale of Leon), but usually for reasons of being distracted by other games rather than RL. (Diablo II being the main contender - if you're going to do a real-time graphical roguelike, that's the way to do it.)
I love the fact that there are so many Angband variants, but personally I prefer the ones that tweak the gameplay without overlaying a whole bunch of stuff. So I never really got into the complexity of Z or ToME, and instead preferred 'smaller' variants like GSN and O. I'm a huge fan of Leon's work: O and S are the only two variants I keep going back to. I have promised myself to give Un another try one day (I wanted to like it but it felt too complex - to Angband as ADoM is to Crawl), and I'm saving FA and Steam for a day when the whole genre seems stale and I want some refreshment.
IRL I'm a (non-IT) project manager (well, programme manager now), and most RL projects follow a strict 'cathedral' setup where everybody has a very clearly defined set of responsibilities. So the 'bazaar' model of Angband's development fascinates me (read this essay for an explanation of these terms). I started off wanting to improve Greg Wooledge's randart generator, and it so happened that someone else wanted to do the same thing, so we did it together. The new one was actually ready in about 2002, but RR was never really interested in it.
I've not done any serious coding since a gap-year job writing library functions in fortran77. Like many before me, I've learned what I know of C from the V code. Someone on #angband-dev derided the overly simplistic structure of the code (comment line, line or two of code, blank line; repeat) - but it does make it an amazingly effective learning tool.
I've been using Linux since 1995 (Slackware then RedHat then Debian since ~2003), and when 3.1.0 came out I volunteered to become the maintainer of the Debian package. It so happened that the existing maintainer was keen to pass it on, and kindly agreed to 'mentor' me. (Debian maintainers who are not fully fledged "Debian Developers" need a mentor to upload packages for them.)
So my two primary interests in the code are artifacts and packaging. I'm currently working on revising the packaging stuff: the previous maintainer always built it using --with-setgid, but I want to change to --with-private-dirs, which is now Takkaria's preferred default. A straightforward change from the building/installing angle, but trickier for upgrading and purging. I hope to get this done for the release of 3.1.2
Beyond that I have a bunch of other aspirations, mostly around enhancing my own enjoyment of the game:
- sorting out pref files so I can have the right macros and squelch settings and autoinscriptions loaded for each character, and a monster memory that spans save files. IMO the existing pref file system has not coped very well with more recent additions like autoinscriptions and squelching. Could do an artifact memory here as well, though being a randarts man I'm not too fussed about that (worth saying here that I do think it's important that the artifact set is well-balanced and interesting, even though I rarely play with them)
- improving the knowledge menu so that you can browse last-known store and home inventories from the dungeon. Of course you won't know about any stock rotation since you descended, but I share Eddie's view that anything that can be accomplished by note-taking should be available in the game UI.
- packaging Dubtrain's sound pack so that users of Debian-based distros can install them on top of the main game (Dubtrain's choice of licence is considered "non-free" by Debian, so they cannot be part of the main package).
- abstracting elements into a structure so that they can be modified / balanced / added to more easily (there are currently hundreds of lines of duplicated code which could loop over an array of elements, and this would dramatically improve the monster power function)
- more revisions and improvements to monster and object power ratings (including making non-magical armour useful by changing base AC and weights)
- more changes to randarts (more interesting curses, more intelligent combinations of powers, etc. - this will always be my ongoing project) - including new artifact types if Takkaria goes in that direction (artifact staves and rods, for example).
Well there you go. That's what I want to work on after 3.1.2. This is way too long already so I'll save my views on TMJ, sterile statistics and instadeath for another post.
Comment