Forge version 1.3.2

[Chris H.]

Starting a next version thread.

Tentative release target date: Friday November 30.

I thought that a 2 week period of time would be advisable this time around.

[Chris H.]

Starting a next version thread.

Tentative release target date: Friday November 30.

I thought that a 2 week period of time would be advisable this time around.

 
I think that we are moving along quite well and should release on Friday November 30. I agree that it is nice to get out a quick release when we have a number of bug fixes in place.

Not sure if the new Vanguard mode will be merged into the trunk before the release.

Would people like to have another moderate version number change when we add a new gaming mode? When we add this new mode should we go to version 1.4.0?

Seems to me that adding new game modes makes a substantial improvement for the better and would make a large difference for the user base. It does not look like there could be an unlimited amount of new game modes and so the version number would not become to unwieldily.

[Agetian]

Personally I like the idea, Chris. I think that major developments might be a good opportunity for the minor version number change.

- Agetian

[friarsol]

I don't think I would bump the version number for just Vanguard if it does make it into this version, it doesn't seem like that big of a leap from having quests with starting cards. Maybe I'd be more persuaded when the next variant gets added in, but if the progress is relatively quick, it'd seem weird to go from 1.3.2 all the way up to 1.7 in a few months.

[moomarc]

I think I agree with Sol here. My personal opinion is that I'd maybe bump the minor version number again when we have ALL the major variants: Vanguard, Planechase, full Archenemy and Commander. Free-for-all I think could just slot in whenever, while Network play (and Twoheaded giant?) would be another minor version. I think Mini-max AI would probably lead to Forge2.0, but that would be debatable.

[timmermac]

Didn't Braids try to do mini-max AI and wind up throwing her hands up in disgust and inventing several new swear-words?

[friarsol]

Didn't Braids try to do mini-max AI and wind up throwing her hands up in disgust and inventing several new swear-words?

To be fair, we did try to warn Braids that the code wasn't in any state to try to be doing that. Max has patiently been going through and upgrading several key areas to make them more friendly for a potential mini-max AI. It's definitely not around the corner, but it's always good to have an idea of what upgrades we'd like to accomplish down the road so we can plan accordingly.

[Max mtg]

To my mind, a single increment of minor version when any variant of (Commander, PC, ArchEnemy) is implemented with good multiplayer (with several AI players being aware of every opponent and ally) would be most reasonable.

That should be a very massive update to give a major version increase for the only new AI. However, if bundled together with network play it would be more than fair to increment major version.

To be fair, we did try to warn Braids that the code wasn't in any state to try to be doing that. Max has patiently been going through and upgrading several key areas to make them more friendly for a potential mini-max AI. It's definitely not around the corner, but it's always good to have an idea of what upgrades we'd like to accomplish down the road so we can plan accordingly.

That's true. Many people came with burning eyes eager to make some cool stuff like network play or brand new AI. But as they get more familiar to the project, their enthusiasm went away, and suddenly the very person followed his or her enthusiasm (i.e. also left).

PS:
Looks like I am also subject to such resigns, yet it appears to be periodical: make some local revolution, "become much less active", come again next quarter.

Never knew that my changes made minmax implementation any closer, but thanks for appreciating that anyway %)

[Chris H.]

We can continue to stay at the 1.3.x numbering for awhile and wait for something bigger before we go to 1.4.x. I do not think that finding us at 1.7+ with a lot of different variant game modes added would look too weird. Most people would be concentrating on the new modes themselves.

If we do reach a point where we have a min-max, monte-carlo, etc type of AI I would suggest that we move to 2.0.0. Rares released the version 1 source code so the rest of us could continue to improve it while he attempted to re-write the game from the ground up. He planned his version to be version 2 as it would have all of the various missing pieces that people had found in version 1.

At this point we have added to version 1 everything that he had planned except for the min-max and/or monte-carlo type of AI. Oh, he also wanted the additional variants but we are adding them at this time.

[moomarc]

Okay, I've flipped and think that going 1.4.x might be the best.

[Agetian]

I wanted to add my $0.02 to this discussion, as well as provide the overall feedback on the developments in Forge as to how I perceived them being a long-time user and supporter of the project between 2008 and 2012...

On the one hand, it might seem easy to criticize Forge for what it doesn't have. Yes, well, we don't (yet) have an advanced Minimax AI, we don't (yet) have network play, we don't (yet) have some of the cards supported, etc. etc. It's also an "easy way out" to just give up when you see that a particular feature is difficult to implement and then leave (such as the Minimax AI, again, which required very significant modifications to the Forge code to be done).

However, behind trying to nitpick and see what we don't have, it's amazing if we take a look at what we actually do have and what we're trying to do, as well as consider where we've been even a few years ago and how much the project has advanced over the years. I was looking at an old screenshot of Forge, I think it was dated either 2008 or 2009, and I remembered myself playing the earlier versions of Forge in the not-so-distant past, just a few years ago. Forge has come a long way, guys. A very long way. It has come from a 1000+ card, standard Java application interface, half-cards-don't-even-work-completely (and another half either crash or at least cause the AI to go haywire) individual project to a 11600+ card, scriptable, beautiful-looking and themable (well, with some reservations like the top right card detail box, but I'm sure all that will be fleshed out over time as well!), multi-mode MTG simulator with constructed, draft, and sealed modes, with an interesting configurable quest mode, as well as rapidly progressing support for MTG variants, with sound engine and with fun AI. Yes, guys, fun AI - despite the fact that it's not Minimax and not as deeply thinking as some of the other fan Magic projects, it's competitive and fun to play against and can definitely kick your butt if you give it a chance to. I remember that just a year or maybe a year and a half ago, when the Draft mode was still new, I used to nitpick about the draft deck creation AI always combining all five colors into an absolutely unplayable (at least before turn 6 or 7) monstrosity that I always steamrolled even with the simplest draft choices. Now, however, I can't approach the draft mode without strategic thinking applied, because the AI actually builds some pretty good decks out there (and plays with them well enough!) so that I know my butt will be kicked unless I apply some strategy to my drafting.

Anyway, what I want to say is that Forge has improved dramatically over the years, and despite all the things we don't yet have (Magic is a very big game, with a huge rulebook and tons of cards, with more and more being released every year, so it's no surprise), we have a lot, and we strive to add a lot more, and the improvements actually come quickly enough - quicker than in many other projects, by the way.

In the course of time it's sometimes difficult to track the progress objectively, every single improvement taken as a standalone achievement might seem minor enough (unless it's groundbreaking on its own, such as the addition of multiplayer support or networking), but altogether these minor achievements add up to something major quickly enough as well.

As for version numbering, there are indeed several approaches to version numbers in the world. Some projects never get out of the 0.xx numbers, reluctant to even get to the 1.0 level despite already being stable and polished enough (DOSBox is a good example, it's still at 0.74 despite the fact that it plays the absolutely majority of games perfectly already - if 1.0 is the level where it has to just play everything absolutely perfectly, I don't think it'll ever hit 1.0 ). On the other hand, there are projects which are quick enough to advance even the major number of the version, quickly going into tens - take a look at Firefox, for instance, which is already at version 17.0 or something like that, if I'm not mistaken. There is, of course, the middle ground too, with its own gradation.

It's up to us to decide how exactly to go about numbering Forge and when exactly to increase the major and the minor versions. I agree that the major revision (2.0) has to be something groundbreaking, but for minor revisions (e.g. 1.4 instead of 1.3), I think that individual significant achievements, such as adding a new game mode, might as well be enough, but then again, I'm OK with either way of numbering, really, since what really matters is what's in the game.

Thank you all, guys, for staying faithful to the Forge project! It's a blast to help you out with it, I enjoy working on the project and I have faith in the Forge project. Keep up the great work!

- Agetian

[friarsol]

I'm pretty sure cards supported is one thing you can't really criticize Forge for. Aside from MTGO, I'm fairly certain Forge has the highest card supported count out of the Rules Engines out there.

As far as I can tell there are 12760 distinct cards, and we have 11665 of those which is 91% of all cards. This doesn't count Vanguard at all, but is really impressive number.

[Agetian]

Yeah, that's very true!

- Agetian

[Chris H.]

I wanted to add my $0.02 to this discussion, as well as provide the overall feedback on the developments in Forge as to how I perceived them being a long-time user and supporter of the project between 2008 and 2012...

 
Thank you Agetian.

Yeah, forge has changed for the better over the last several years. Most of those early faults have been addressed at this point. And there are still some areas for us to address for the foreseeable future.

I have played several Vanguard matches and they are just awesome. Thank you Hellfish, Max and Marc, your work is appreciated.


And I do appreciate everyone's hard work and I do want to reward them with a version number update but I can hold off for awhile. No problem.

Granted, I am an old timer, have been around for a few years and am a moderator for this forge forum. I do not want to be a dictator for this group. But I do try to find a way to encourage people to have fun and to contribute to the project where/when possible. And as a moderator it is important that I try to encourage some discussion at times.

[moomarc]

One additional bonus of releasing as 1.4 is that minor version increases seem to encourage more people to update to the new version than sub-versions do. I guess some people just find that current bugs don't affect cards and/or modes they use so there's no reason to update, but a new feature version changes that.

And as a moderator it is important that I try to encourage some discussion at times.

Nobody told me I'm supposed to do that I'm not discussioney enough! Will have to fix that