Angband 2.9.3L

[Gwarl]

https://github.com/OwenGHB/starband/releases/tag/steamband-0.4.1L

Steamband 0.4.1L

It was surprisingly up to date in some ways for a 2.9.3 variant, e.g. it had already implemented the above fix for 3.0.6

[Gwarl]

Fixed an incredibly minor but incredibly annoying bug where the cursor wouldn't hide on the linux GCU port. It just *looks* nicer now.

Copied the fix from a 3.0.6 fix submitted by Alexander Ulyanov

[Gwarl]

I have compiled the Benbands!

Copying h-system.h from 2.9.0 over the versions present in 2.7.4-2.8.3 allows them to compile and from there just fix the L64 bug and you're set.

I also attempted 2.7.1v2 which is up on AngbandPlus despite rephial's claim that 2.7.4 is the oldest benband with surviving source code. This version still resembled older versions and while I could persuade it to compile it won't run. This version lacks 'gcu' or 'ncurses' support and only provides a frontend for 'Actual Unix Curses'

AngbandPlus also provided me with 2.6.1 (the archive present on rephial will not extract for me). This is a different beast entirely, and while I eventually got it to compile using subtle trickery, as with 2.7.1v2, it will not run.

2.7.4 is the first version which resembles angband in terms of file and code structure. By 2.8.0 it looks like angband as a game as well - previous to these points, angband looks like a moria variant, and not like angband.

[Gwarl]

2.9.3 updates: changed the monster list window colours to conform better to later versions (trivia: Tome2 uses the original, unaltered patch) and added a user interface option "toggle_xp" from Zangband (shows xp to next level rather than total in the character display)

[Gwarl]

I've ported the subwindow handling to the following versions:

2.9.3, 3.2.0, 3.4.1, 3.5.1, 4.0.5, 4.1.3

which is a list still three shorter than the number on the site. I will get around to 3.0.6, 3.1.2v2 and 3.3.0 in due course.

[Gwarl]

just got done porting the customiseable gcu port subwindows code from poschengband into latest V (and 3.2.0 while 4.1.3 was giving me segfaults)

that will be part of the standard to work toward for restoring other variants, I've sent Nick the file so hopefully it'll appear in future updates. the modifications retain support for the much simpler (and less flexible) -n4 style syntax

now I am free to work on adding a site feature to allow the customization for variants which support it so I can customise my kangband subwindow layout

[Gwarl]

Well, see the picture attachment for my motivation. Judging by the code as I found it, Kangband has never been run in a terminal with a main term window larger than 80x24 before, but it's very pretty.

[takkaria]

2.9.4alpha incorporated the reportedly controversial JLE patch so 2.9.3 was arguably the last 'old' angband. Although having played some variants based on 2.8.1, and then playing 2.9.3, I suspect 2.8.1 was also more difficult.. (but won't compile until I pick up more tricks)

I think by the time JLE was incorporated, it wasn't super controversial... it had been around for a few years already by that point. But like any significant change not everyone liked it. The spell list was significantly revised too - once every 17 years seems quite like a good pace to me.

The idea of anything older than 2.8.1 fills me with mild horror

After I removed Lua, the code from 3.0.6 closely resembled the code from 2.9.3. And there were a *lot* of cleanups after that. For example, the #defines around the custom int types were horrendous. I spent ages ripping out code for systems that were dead already 12 years ago and simplifying stuff, so it's funny to see you doing the same thing now, for the same platforms... But all that code was needed to get things working for someone, somewhere, on a some mainframe. It's funny now the dust has settled, desktop computers have peaked, and Windows/Mac/Linux are the only three platforms anyone cares about supporting, but at one point that didn't seem like an obvious outcome. O diversity has reduced a lot.

And it blows my sometimes that bits of Angband have existed in three different languages (probably more than one version of Pascal, and then C and Lua) over four decades. Ben Harrison did a very good job of turning code that still resembled the Moria - that was written in Pascal - into something readable. Even if by modern software engineering practices it's a total nightmare.

[Gwarl]

Ah, this is very cool - 2.9.3 was current when I first encountered Angband. It's funny to see it resurrected.

2.9.4alpha incorporated the reportedly controversial JLE patch so 2.9.3 was arguably the last 'old' angband. Although having played some variants based on 2.8.1, and then playing 2.9.3, I suspect 2.8.1 was also more difficult.. (but won't compile until I pick up more tricks)

The idea of anything older than 2.8.1 fills me with mild horror

[takkaria]

Ah, this is very cool - 2.9.3 was current when I first encountered Angband. It's funny to see it resurrected.

[Gwarl]

I've managed to get zangband 2.4.0 to understand the difference between /pref/ and /user/ so that it functions properly on angband live, and it is a strong candidate for the next bigscreening.

zangband included its own version of the bigscreen patch in 2.5.2 but the original fixed wilderness was removed for 2.5.0

[Gwarl]

Just ported most of this to Kangband 2.9.3r1 (no monster list yet) - possibly the first time it's been playable with a big screen?

https://github.com/OwenGHB/Kangband-293L

Also up on angband.live

[Pete Mack]

There are some things that made it into very late versions of 3.0 that really belong in other variants, notably the knowledge menus from unangband. It is easily portable code, and unangband doesn't use the menu code, so its version of the knowledge menu works independently of it. (But the menu code makes other things easier, too.)

[Nick]

This is wonderful, awesome work.

[Gwarl]

And I almost forgot - thanks to Nick for creating AngbandPlus, which was an invaluable resource for this