I'd say warrior, or paladin. Rogue is surprisingly poor at combat and archery, and has only 15 sided HD. Paladin has Detect Evil, and Detect Animal* rods are easy to come by. Going without detection is a huge pain for certain bosses (Dunlending agent, Azog, Ferny.) I don't know if Azog is even possible without at least detect evil. He runs far and fast.
My recommendation is to make it an open competition.

* There are tons of nasty critters it's best to avoid.

Yeah rogue is a fair bit weaker than in vanilla. The hit dice go up with size & I get the impression race makes a big difference to accuracy. I might try out a half-orc rogue & see if it melees any better.

Next Comp
 

Well, before we think about what race/class combo should be played we should probably think about what the point of this particular comp is.

Just showing off the game in general? Specifically showing off the innovative ideas (such as the new and interesting magic spells) in it? Something else?

Race does make a pretty noticable difference; I know that I've had a much easier time surviving as a dunadan rogue than I have as a man of Dale one.

My suggestion is to hold off for 0.7, when a few glaring balance issues are solved. 0.6.5 was intended as a bug fix release. Any comp now would be reporting a bunch of really obvious problems. Wait until those are fixed, then a comp would actually provide useful feedback.

Sorry folks. Things are a little crazy right now. Will have a comp up sometime in the next week or two.

Suggestion for future PosChengband compo: Consider an Archon Force-Trainer? Any thoughts?

So comp suggestion quick - poschengband iron man high elf ranger. Nature/sorcery easy id&lore, reduced uniques.

Sorry for the delay here, folks. I'll have something up soon.

@all: So, this may be a bit of a weird request... but I'd like to see the top players here try a ToME 2.x (-AH) bard. Yes, bard. (Even if it's the original T2 2.3.4, that'll probably be OK, I haven't changed much wrt. "balance".)

I'm asking purely because it's been known as a ridiculouslty hard class, but I'm seriously considering removing it because it actually takes quite a lot of "if-then-else" code to support. (I think the general conduct for this would be: Only Music-style magic, e.g. no Mindcraft, etc.)

For a *real* challenge, this should be at least a LostSoul Bard for shits and giggles :D

I could make it out with a Death Mold Lost Soul Bard. For a **real** challenge make it a Maia Lost Soul Bard without linked stairs.

Hmm Bard. I never played one, bacause the songs look so weak that they dont matter compared to the other options. For example, you could play it as a simple warrior with magic device and have no problems at all. The restriction "no mindcraft" is rather pointless imo, as there are plenty of other similarly powerfull options. God magic, antimagic, mimicry, symbiosis, you name it.

If you want to create a character so weak that bard songs become a relevant boost, I would suggest no princess quests for starters; other doors to non-bardic power will have to be closed, too, but I am no longer familiar enough with tome 2 to list them all offhand.

Maia lost soul seems off when the intention is to have bard songs explored.

If you make this, I vote for the original Tome (since that is what I remember), and I would play the comp. But, tome being tome, I fear the way to go about this will be to not bother with bard songs much.

Don't Maias have decent melee to compensate? Which version (for the competition) should be the "original ToME"...2.3.5?

Original just means not T4. 2.4 would be my assumption.

I meant 2.3.5. I have no idea about the changes done in 2.4, and with a game where irreversible early decisions affect the character later, you always end up with borked builds for the first toons till you eventually get a grasp of all the details. Or you go to the sandworm lair but that has been rebalanced to now spawn <more_deadly_monster_X> and so on and so forth.

I would prefer the older version because I dont feel like re-learning the game.

FYI to everyone:
2.3.5 was the last 2.x version released by DarkGod. 2.4.11(?) is the last 2.x version released by AH.

Make some savefiles yo. debo is on a soul journey right now.

Jeez you people are conservative -- just try it :). There have been very few game play (or "balance") changes in my fork. The only one I can think of of the top of my head would be the removal of traps and making Thaumaturgy a little less absurdly overpowered. (I've also removed Alchemy + Runecrafting because they've been hopelessly broken. Bard is also on the list because they seem to be almost unplayable and the code support them complicates the spell system quite a bit. I'm not opposed to the class per se, it's just a pain to deal with such a large amount of code that's effectively never used.).

It still plays very much like T2 2.3.5 except it actually runs properly on 64-bit by not instantly giving you 300k piety on save/load, etc.

(I'm also thinking about removing ID and just auto-ID'ing everything.)

EDIT: Of course, just to be absolutely clear T4 is a whole different game.

I looked at all eleven Bard winners on the ladder.

Several were played as monks, including Lord Dimwit's yeek winner. He worshipped Melkor and used Curse to take down tough opponents. The lowest scoring winner was also a yeek played as a monk, but worshipping Tulkas.

There were several that had high Symbiosis, so they could b played as chain summoners (this was my first thought).

There was only one with maxed-out Mindcraft.

So the two common successful approaches are to use Barehand or to go with chain summoning. This makes sense, as Bards start out with decent access to most of the other Loremaster skills.

If you want something relatively easy to see if the class is worth preserving, try a RohanKnight worshiping Tulkas.

If you create a T2 2.4 comp savefile, I'll happily post it for comp 205.

Honestly, I think I probably need to make a proper release for that... (You know with Windows executables and everything :))

I don't know if/when I'll have time for that... but given the offer, I suppose I might at some point :).

There is a compiled version of Tome 2.4 for windows floating around:

https://www.reddit.com/r/roguelikes/...dows_binaries/

It says 2.3.5 but afaict it's 2.4.0 according to in-game help.

Is there any documentation on Tome2 floating around online, with the wiki gone?

Oh, well, better than nothing. Thanks!

There's the skill spoiler and spell spoiler.

https://te4.org/tome2-ldasg

Here's a mirror of Lord Dimwit's Advanced Strategy Guide for ToME2. It's quite comprehensive, including end game spoilers.

http://web.archive.org/web/200803171...5f597850cdc5fa

Another mirror in case the first one doesn't work.

Thanks! Even better. All of this makes me seriously want to try Tome2 again...

Y'all have 9 days to post a TOME2 savefile (whether it's 2.3.x or 2.4.x) that you'd like to see used for comp 205. Otherwise, I may try to make one myself, and I've never played TOME2 before. I promise I'll create said character without reading any documentation first :)

You should actually do that. Just pick race/subrace/class/options that sound interesting and inflict them on everyone. :)

I mean, that's normally what I do anyways!

Just a correction: There's no T2 2.4.0 -- that's the "in-progress" version number of 2.4.1 (I guess). Quite an unfortunate mistake.

Since there seems to be a bit of interest: If anyone wants to give compiling on Windows a go, I suppose I could pretty easily just tag the current HEAD on master and officially make that 2.4.0. Anyone up for that?

(Please use 2c8b8579faf729b9cf21b8b2d827f0e482570bd3 as a starting point. If you have fixes/problems please just post them to the GitHub issue tracker or, even better, make a PR. Failing that, feel free to post in the Variants forum. I'll wait to tag until we get a Windows compile then I can probably attach that as a "Release" on GH.)

EDIT: @debo: I'd recommend choosing the Theme module. For me, at least, it's vastly more enjoyable. There's a "choose random option" key on the character generation screen, but it occasionally causes an infinite loop. I haven't found the exact reason yet, but if you just retry the chargen process a few times, I think it should work. (Or just roll dice or something instead :).). Oh, and we should probably avoid the Bard class -- but then again, maybe it'd be good to see if anyone has the patience to do anything with it.

EDIT#2: Also, just a heads up: This requires a recent compiler; not exactly sure how recent, but this GH issue is probably relevant.

I managed to build 2.3.4 but I'm getting aborts when I try to save+quit, which makes it hard to make a savefile. I'll try again later. Does anyone know why we would have preferred 2.3.4 to 2.3.5 in the previous comp we did for this game? Is 2.3.5 broken or something?

[Edit: I can't save properly in 2.3.5 either. It's generating files in ~/.tome/2.3/, but I can't load them. Can't tell if the savefiles or bad or not because of it.)

Are you building for 64-bit? I know that that is at least very broken for 2.3.4 and 2.3.5 (was there ever a 2.3.5 release?). I also believe there's quite a few overflows which will cause e.g. AppArmor to crash the executable. (I'm not familiar with Windows, they probably have something similar at this point.)

Did you try building the GitHub "master" version? If so, what problems did you encounter? (I think you should probably post a new thread to avoid cluttering up this one.)

How about a V master comp while you're working out the ToME issues? It's in no-more-features final-bugfixing mode before releasing 4.1.

This sounds great to me. Sorry, ToME fans.

I'll set this up sometime this weekend.

Nick: any suggested race/class you'd like to see?

ARGH! I didn't realize it was past the 11th. Well, anxiously awaiting this one. :-)

One of the reasons why 2.3.4 may be preferable over 2.3.5 is because 2.3.5 has the dreaded chardump CTD bug whereas 2.3.4 doesn't: If you have 10(?) or greater completed princess quests more than the completed Lost Sword quests (aka Fumblefinger) quests, the game will crash when you save the chardump. Thus, after that point, you won't be able to use the chardump function until you complete more Lost Swords quests.

I had a vague plan to add in automatic support for competition entry on the webserver (after spectating and recording are done) with the hope I could get it done in time for a 4.1 release/competition.. I'll have my work cut out to get that done in time for this. But I'll give it a go.

(Not responding to you specifically, but...)

Let's be clear: The old versions are broken in numerous ways related to various overflows[1] and don't you even try to compile for 64-bit[2]. Most or all of these problems have been fixed in my fork.

If someone can give me, or link to, basic instructions for getting a Windows compilation setup using e.g. MinGW (or similar), I can try to do the Windows compile myself. (Caveat: I have no Windows licenses at the moment[3] and I'm not going to buy any. Call it "political reasons".). I'm assuming that most Linux players would be capable of compiling the thing themselves.

[1] Which means that if your OS does any kind of hardening, you'll end up with crashes. Frequently.

[2] My favorite was to use Morgoth as a god (piety < 0 at the start, IIRC) and to have a character with several negative skills (Stealth is a good one, but Magic Device or similar would also do). Start game, do nothing, save, reload. You now have 300000 piety and your "negative" skills are now maxed at 50.

[3] Well, I happen to have a volume license for WinXP that's still legal, but I don't think that's of much use to anyone. I do still use it for playing Baldur's Gate once in a while :).

AH--if you are using the Angband build and main--xxx.c system, a windows port is easy. I'd be happy to make one, given a link to the source, and a place to drop the excutable.

The character I was actually going to post as the comp character was a Thunderlord Dark-Priest serving Melkor. I have no idea what gameplay that translates into, but it sounded cool! I named it after the gryphon rider hero in Warcraft 2.

That actually sounds really interesting! It's not a combo I've ever played, but... {fires up a game} Ah, yes, they have Necromancy at 0.800... The Necromancy spells aren't that great but the abilities (press 'N') includes a thing called "Touch of death"...

I'm not sure if it actually works, but if it does, it's pretty amazing, especially if you're going to the dungeons of super-leveled monsters. (At a lesser level of Necromancy you can also get 'Undead Form' to let you avoid death... maybe.)

Priest (hasn't been one of those forever), maybe Dwarf? Standard artifacts - there's already plenty of talk about randarts going on, and the aim here is to find all the issues that need resolution before release. So a very vanilla Vanilla game.

Why not make it a half-orc for variety's sake?

Sure, good plan.

itym "for god's sake"

The source is on GitHub here: https://github.com/tome2/tome2

Do not that I'm using CMake, there is a basic command line in the README.md on how to compile using MinGW. (AFAIK it's been a while since anyone's tried it, but it should basically work. It's not there's loads of changes to the build system...)

If you can manage to build it, feel free to send the executable as an attachment to bardur@scientician.net

Is it possible to specify subdirectories during the installation, or indeed afterward? I'd like to add it to my server but I need score/save files in one place, edit/pref files in another, binaries in a 3rd etc and to be able to specify alternate folders on the command line when calling the executable. Most angbands let me do this but since the make process for ToME is different I'm unsure how I'd go about it.

AFAIK there's only the SYSTEM_INSTALL option for the build (see the README). You can set the prefix for where files are installed via the CMAKE_INSTALL_PREFIX variable.

So, that would be

There are no command line options for changing individual locations (e.g. "save", "pref", or...). I don't there ever were, actually.

You can set an environment variable called TOME_PATH, but I'm not sure if that's very useful to you... or anyone else, for that matter :). Maybe something could be hacked together by creating a "fake" ToME dir with symlinks for "info", "help", etc. to the real directories, leaving only "save" as a real directory... I'm not sure.

I think I should be able to work with that - except ToME is failing to compile at all for me. I got the recommended libs, using Ubuntu.

Let's not clog up this thread. Feel free to post a new thread in Variants, or even better create an issue on GitHub (assuming you have an account) with the detailed error message (etc.). Note also that you may actually be experiencing issue #10 which is non-trivial to fix -- hence not fixed :/ .

The reason I brought it up here was because I'm hoping to be able to offer recorded tournament play on the server by the time of the next competition. Which means being able to support the variant we use.

It does look like issue #10.

Next comp will be poscheng quickband, pick whatever you want. Should be a 2 week comp.

Gwarl: how's the web client going? If it's comp-ready we can do after this next one.

I tried screwing with tome2 a bit more and gave up, I don't have time to be fiddling with typedefs and weird word size bugs. If someone wants to donate a savefile, I'll happily post one, but I'm not going to be making one.

Are you saying you want us to decide the compo combo before it starts? Or this upcoming compo allows us to pick any combination we want??

The latter :)

Gee I wonder what Hugo is going to play if he joins!

It's improving, I'm willing to say I'll have it ready by the time this 2 week competition is over.

In the case of this competition it's possible now - my entry into this comp will be played online.

If you want to use a particular variant for the next one I'll see what I can do about supporting it. Currently the choices are Vanilla, Poschengband, FAAngband and before long Sil.

If 4.1 gets a release before the end of this competition I'd suggest we pick that - this time a low str race red book class.

Very cool! It'd be cool to suggest people try optionally playing on it if they want for this comp if that's possible -- I didn't realize poscheng was already set up. Is there an easy way to get chardumps from the game server to an oook competition ladder?

Char dumps are accessible by URL but there are no instructions so:

If you're logged in as Bilbo and you call your char dump Hobbit.txt your dump would be at

/user/Bilbo/poschengband/Hobbit.txt

Assuming you were playing poschengband.

Providing a list of links to the contents of the user folders is on the agenda. I would also like to add a button to post directly to the ladder but I don't know that all works just yet.

My only issues with the PosChengband on the webserver are the following: Running (Shift+directional key) doesn't work and its colors are off. Other than that, it may be good to go.

Pick whatever we want also means we can choose Monster races too, right? E.g. Possessors and Mimics?

Time for another competition soon.. 4.1.0?

Vanilla 4.1 comp will be going up in the next 24-48hrs. If anyone has a character / birth option setup they'd love to see played, post it here! Otherwise I'll just make something up, probably on the obscenely difficult side.

Half-troll mage, forced-descent no-recall. :)

I'm not a fan of forced descent, but no-recall would be good. I'd be partial to a human warrior. Things like mages and priests have a hard time starting up: if you look at the recent comp, about half the total turn count of the winners was spent just getting to dlvl 30. A warrior (even a human warrior) should be able to move through the initial (boring) levels a lot faster, but has a harder time down deeper. And IMO the inventory problem of no-recall is especially tough on a warrior.

edit: well, maybe a third, not a half. And yes those are Standard turns, not Game turns. But I still claim that casters spend a lot of time on the early, less interesting parts.

@bron: don't know about priests but Mages--especially HE--spend the boring time between around CL 33 and CL 40.
EDIT: *Especially* if the get shortchanged on !DEX and !Restore Mana. I've found 2 of each in my current game!

Isn't a lot of that due to having to rest to restore SP? So if you play a race with innate regeneration, that's not a problem. :D

My preference:
To contrast with the last V comp (High str race green book class).

I mean you could get obscenely difficult with that. Kobold mage anybody?

I was thinking kobold mage myself :)

Next competition is in just over a week. Anyone have requests?

UMoria half-troll warrior.

Quickband?

FA dark-elf thrall rogue in standard wilderness with disconnected stairs.

If we're going thrall rogue dwarf is also a fun option. rblind works against destruction & you can move fast through the rubble.

There's a bit of a tension between race/class combinations which are good overall and those which are good for new players in the early game. My fear is that people lose 20 @'s over the course of an hour, get frustrated, and decide the comp is stupid. I think probably the best way to avoid that is to maximise stealth. There are definitely tactics to deal with less than stellar stealth, but they take time to learn. If you have high stealth then you have a reasonable chance of making it safely to Gondolin just by not waking up too many nasty things. You can instead wake things up and *destruct* them, but you don't have very many scrolls available, so it takes a bit of experience to know when and how to use them.

I probably won't do a thrall but perhaps a high difficulty FA race.

That seems like a bit of a wasted opportunity. The nice thing about comps is the opportunity to experiment with things which are unique to a particular variant.

My request:
PosChengband Tonberry Chaos-Warrior + FULL Wilderness + 100% Uniques (YES, I mean 100%)

Aren't those just the default settings.
Not sure what you meant by this.

Pretty sure the last poschengband comp was so recent that my run is still going...

There's a lot to distinguish FA without going full thrall mode. I'll have a few plays through but I think the harder races might be challenging enough as they are.

More suggestions are still welcome.

Me too haha :)

Thrall isn't really that much more challenging than non-thrall. It's just different. Starting with telepathy goes a long way to compensating for your fragility as a first level character in dangerous territory. The keys to surviving as a beginning thrall are:

- Hounds and vortices sleep at night. Try waiting for dark. You'll be a lot safer. You may need to *destruct* to create a safe space in which to wait.

- Unlight is great. If you're a rogue, assassin or necromancer then it should definitely be your first specialty.

- Choose your first kill wisely. Experience points get divided by character level, so it's best to get as many as possible when you're still level 1.

- The only escapes you start with are five scrolls of *destruction*. Use them when needed, but only when needed.

- Visit (and move to) town fairly quickly. There you can get more escapes. Also, WoR only works once you have a home to recall to.

- Mostly ignore stuff that isn't close to you and far from monsters. It usually isn't worth the risk.

Some of those are more guidelines than absolute rules, but if you follow them then thrall mode isn't really that much harder than standard. Most of your @'s will still die, of course, but they mostly die quickly, and if they don't then you can skip much of the tedium of the early game. The average time required to get a decent midgame character is probably lower than for a non-thrall.

But if the point is to try lots of variants, I don't want to skip the early game..

Those default PosChengband settings + race/class combo of choice (I should have said Tonberry Skillmaster) = highest chance of ME winning any compo (at least one and notwithstanding the 100% uniques)

^ That's what I meant!

;)

Perhaps the next compo should be on PosChengband 7.0.0!!

yeah this is the best idea have more then 1 option a month do like 1 monster class and 2 others maybe at same time and people can keep up on a separate ladder for all 3 so you can keep on with the one your playing.

This should be implemented. Gwarl - you should find a way to make this happen!

So, there's this issue with competition mode on the server:

https://github.com/OwenGHB/angband-webclient/issues/5

Vanilla and FAAngband are afaik the only two variants with this savefile header information, which is what competition mode currently relies on.

Since I've been short of free time lately, I won't be able to solve this by next weekend. I have thought up a way to get around this and still enable competition savefiles to be played on the server, but won't be recording runs. This seems acceptable since there isn't a playback feature to use the recordings with yet. So recording will go on ice, and when I find the time to implement playback all the entries for this competition and the last will be available to watch.

Only at that point will I worry about recording games again.

Also: I would like to change the rules at some point. Getting the same colours of potion and identically named scrolls every run doesn't sit right with me, nor does the ability to get the same first dungeon level every game. That's an opinion I've had for a while. The next competition will follow the same format as is traditional, but for the one after I like the idea of programmatically generating savefiles which have completed the birth process and offering a download link somewhere that provides a new file with every click.

If anyone would like to strongly object, you have plenty of time.

The problem with that is that it theoretically encourages startscumming. I don't think anyone is going to practically do that, though, so it should be fine.

Neither start scumming or paying attention to flavours are things I care to do, I guess unless you force on-line play (something I'd like less then cheat checking) you have to live with something. My preference is to allow someone to just quick start rather than re-download the old save.

I feel like the information leaks from everyone having the same savefile are less important than everyone having the same starting town/stuff/whatever. Changing this feels like a fix to a problem that doesn't really exist. I did not use any of the information leaks in my game this comp either, for example.

The only times I know of that the fixed start has mattered were in Ironband and in FA (ironman, to boot) thrall mode, and those are very much special cases.

I personally can't help but remember object flavours. And once I've realised I've remembered 2 or 3 correctly I get the cognitive struggling of admitting that I know them and not wanting to 'cheat'. I also find it tedious setting the options to my own preferences at the beginning of each run (not everything lives inside .prf's)

Let's remove the ID game from all competition variants! That would fix the problem :)

Perhaps, randomization should be on realtime turns NOT game turns. -- Probably really hard to implement.

Outside of variants with crazy use-id systems, knowing flavors is of negligible importance.

As for setting options, I fail to see how your proposed system avoids that really.

Well the issue is that things like potion flavors are usually randomized when the save is first created, and competitions work by giving everyone a pre-created save file. The point at which the save starts is going to be AFTER the game has already decided to make CCW the gloopy green potion and heroism the speckled brown potion.

Im not sure its worth it to take measures to prevent some perceived cheating. Because I think if players want to cheat, they are going to cheat anyway, by hacking the code to get free identify of these randomized flavors or whatever. I remember a npp comp a few years back where someone had like 50 artifacts by clvl 20, I think I had 2 at that same point in the game. I thought dude had obviously done some hacking, and I cant really imagine how that makes it fun, but I dont think its worth getting upset about either. Theres nothing at stake, Im just here for the enjoyment of it.

I dont think anyone is trying to cheat, its just that you cant avoid noticing that that pink speckled potion is lose memories so next time you play you know not to drink it.