The Tavern

This was offtopic in another thread but somewhat relevant:

twelwe wrote:Too many changes this trunk. Getting impossible to make any meaningful notes comparing to stable because there are so many things flipped upside down. Four new playable races, countless new monsters and mechanics, a complete overhaul of the dungeon itself. Slow down.

The current trunk has got to hold a record for the amount of new stuff and changes that makes it look more like a variant of DCSS than anything. Lite had a multitude of great ideas and changes but most would agree those changes were more deserving to be in a variant than integrated into trunk. Replacing the hunger system with contamination no longer seems like such radical change.

When testing new races, it helps to have a constant with which to compare. Testing something like a regen3/no heal device/guardian spirit package becomes less meaningful the more variables are thrown in the mix, meaning that as more monsters are changes, new hostile mechanics introduced, layouts changed or completely removed and replaced with experimental ideas, the less we are able to make meaningful comparisons to the previous branch.

One thing is clear: there is an incredible motivation to put new and inspired changes into the game. I would suggest to compensate ramping up the development cycle to three or even four releases per year, where the current trend is about two per year. VS and dith could probably alone make a good release, perhaps along with some branch changes like snake. Otherwise I can see people switching from .13 to .14 and having no clue what game they're playing anymore.

I've been away from DCSS for a bit. Reading up on the sheer number of changes in .14 has been overwhelming.

Well, I suspect many of the changes in question won't actually be in .14, part of the reason for trunk is to try out various changes and see which ones work and which ones fit together.

To the extent that an overabundance of changes makes this more difficult I can see your point, but personally, I haven't find it too hard so far to evaluate the changes which have come in (and gone away, forest branch has been axed for example) as things are. But I play pretty much daily and rebuild my local copy roughly weekly so perhaps I'm not a good example

Faster releases <3

I agree, while I LOVE 0.14 and only play trunk, I gotta say, I have no idea what the hell is even in 0.13 any more. 0.14 really should be broken up.

twelwe wrote:One thing is clear: there is an incredible motivation to put new and inspired changes into the game. I would suggest to compensate ramping up the development cycle to three or even four releases per year, where the current trend is about two per year. VS and dith could probably alone make a good release, perhaps along with some branch changes like snake. Otherwise I can see people switching from .13 to .14 and having no clue what game they're playing anymore.

I think it is great that game design for Crawl is this dynamic, and I don't think it is necessarily such a big deal that multiple variables are being tested at once for this particular game. I do agree, however, that the number of releases should correlate upward with the number of changes being made, that's a good point.

TeshiAlair wrote:I agree, while I LOVE 0.14 and only play trunk, I gotta say, I have no idea what the hell is even in 0.13 any more. 0.14 really should be broken up.

I'm not sure what "broken up" means exactly, but you might like to know that just because something is in trunk, it doesn't mean it's going to be in the next release. I can think of a few things which aren't planned for the next version. Someone else can give specifics better than I can, but the basic idea of it is, if a race/branch/monster/whatever isn't finished or isn't working out, it generally doesn't belong in a stable version. And there's a bunch of stuff like this: djinn for example aren't likely to make it in.

The game could benefit from an actual release plan though, just so there could be a proper division between "here is what's planned for 0.14" and "here's what's planned for later releases".

Right. Two of the races added to trunk in 13 didn't make the release.

What percentage of people just play trunk anyway? (I genuinely don't know this.) If it is high then frequent releases might not change that much in terms of how easy it is to keep track of changes..

johlstei wrote:What percentage of people just play trunk anyway? (I genuinely don't know this.) If it is high then frequent releases might not change that much in terms of how easy it is to keep track of changes..

Online the vast majority plays trunk. Offline is hard to say... You could try to derive it from the download statistics, but even then you don't know which percentage actually plays the downloaded version.

I think that recent rapid changes are fun and even more often - meaningful. Like removing minor but nasty stuff, like something not working the right way.

Tedronai wrote:Online the vast majority plays trunk.

I don't think this is really true. At the time of writing there have been 1113065 online games in trunk versions and 1831380 online games in stable versions.

duvessa wrote:I don't think this is really true. At the time of writing there have been 1113065 online games in trunk versions and 1831380 online games in stable versions.

Every time I scroll through the player lists of CSZO, CAO and CLAN at least 90% of the players have a trunk game running. Stable versions peak during the tournaments, but right now very few people actually still play 0.13 (or lower).

Yeah I'd like to see that figure with tournament games not included, it's a good start on that estimate though. Offline would have to rely on an in-game survey or something, I doubt that's worth it.

Code:: <Sequell> 112635 games for * (start>=20140101): 94862x 0.14-a, 17444x 0.13, 238x 0.12, 49x 0.11, 42x 0.10

There actually was a survey. From the draft of the 2012 crawl survey report:

C6. Do you play "trunk" (the beta version of the upcoming release)?
The majority of participants doesn’t bother with trunk and only ever plays officially
released versions.
Count Percent Playing trunk?
2707 48.0% No, I only ever update for release versions.
1029 18.2% Yes. I always check out the stable version, but I switch to
trunk for cool new features.
829 14.7% No. Trunk is too unstable for me, but I like to keep informed
about what is happening in trunk.
438 7.8% Yes. Except during tournaments, I only play trunk.
638 11.3% N/A
The higher number of N/A votes (compared to the previous question) might point to
participants not knowing about trunk to begin with.

The participants are a mix of online and local players:

Overall, players generally prefer playing locally.
Count Percent Variant
2543 44.4% Locally
1615 28.2% Online
1449 25.3% No preference
126 2.2% N/A

It would be interesting to compare the numbers of local and online trunk players. The whole report is pretty interesting, dpeg sent the draft to the crd mailing list. It will most likely end up on the crawl homepage sometime soon.

I suspect the majority of players who play offline play Stable, since I also suspect that a majority of players who play offline download from the download page, which only contains a ready-to-play Stable version.

Playing trunk offline requires you compile your own, which I think is something that the majority of people who aren't interested in the development process and just want to play the game don't want to take the time to do, Online trunk is at least easily accessible.

I also feel like a larger percentage of regular-forum participators play trunk than the general playing population. One of the things about participating in the forum I like is that it gets me interested in the changes that are happening, and I want to try them all.

Actually I don't think I've personally played "stable" outside of release tournaments since .10 or perhaps earlier, maybe I should do some of that just for grins.

There are precompiled trunk builds available: http://crawl.develz.org/trunk/

This is of course not the main download page, though.

Also for non-Windows machines these often are quite out-of-date.

crate wrote:There are precompiled trunk builds available: http://crawl.develz.org/trunk/

This is of course not the main download page, though.

Also for non-Windows machines these often are quite out-of-date.

I'd completely forgotten about that page, I used it back in the day, but haven't in like.... forever.

To get back on topic:

I actually think species and gods (Dith) are the best possible thing to have an excess of development on. This is because these are extremely isolated and can easily be bumped a release/removed/test individually. I would have a bigger problem if there were several large far reaching changes which effect most games (Dungeon structure, permabuffs, new non-optional branch) were occurring simultaneously.

Also of note is that a lot of new stuff is coming from further opening development. tenofswords getting commit access, experimental branches on CSZO, the return of DracoOmega, etc. are all contributing to the faster pace. I tuned out for a couple weeks recently, and I had a great time reading through all the changes when I came back.

A fast pace of development might cause some new problems that haven't occurred before, but I think this is a good problem to have.

I would support an additional release or two per year if the current pace of development keeps up, but I don't think it is necessary. Some more official large updates would be nice, but there aren't any broken systems or critical bugs being introduced which makes me want a release to trigger fixing them.

Spring-Summer-Fall-Winter releases.

Gives the game 20 years to hit 1.0, as opposed to everyone playing being mummies by then.

The Tavern

Trunk: rate of change

Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change

Re: Trunk: rate of change