[TOPIC SPLIT] Development progress and direction

[bort]

Moderator Edit: Original thread, http://forum.metalforge.net/viewtopic.php?t=1538

Wraiths are very hard to play without being screwed ingame.

Halforcs, haflings, gnomes, trolls, and northmen should come with bigger bonuses. Think of ADOM http://adom.de.
It makes playing all of the races the same. The same gamble, I mean.

I say CF should open up to incorporating elements from ADOM, nethack, Diablo, etc.

The CF core is pretty stable. Unless the game engine will change soon, a full scale wiki, rebalancing, and features should be implemented.
I've been monitoring the devel list for a while...
Nothing much is occuring. Currently, the topic is fixing a map.
Personally, there should be more emphasis on new features and active development.

[Rednaxela]

Wraiths are very hard to play without being screwed ingame.

I'm not sure what you mean. I use a wraith for my secondary character and it's quite capable of surviving (and killing many things that most characters couldn't at it's level, though I doubt that has to do with it being a wraith)

The CF core is pretty stable. Unless the game engine will change soon, a full scale wiki, rebalancing, and features should be implemented.
I've been monitoring the devel list for a while...
Nothing much is occuring. Currently, the topic is fixing a map.
Personally, there should be more emphasis on new features and active development.

The core is reasonably stable, but there are some plans for significant refactoring in the fairly near future to make some things more maintainable, of course that doesn't prevent features and rebalancing either. In terms of rebalancing and major new features, that's one of the reasons for the 2.0 branch in SVN, and there has been discussion on various types of balance issues and new features mostly on the wiki and on IRC. The issue is most of the developers have been fairly busy lately, so ideas and proposals are being made significantly faster than they can be implemented.

[Leaf]

Nothing much is occuring. Currently, the topic is fixing a map.
Personally, there should be more emphasis on new features and active development.

Hmm.. Just a couple of weeks ago there was strong sentiment from a player(s?) to fix open & existing bugs before allowing the implementation of new features.

http://sourceforge.net/tracker/?group_i ... tid=113833

[bort]

Nothing much is occuring. Currently, the topic is fixing a map.
Personally, there should be more emphasis on new features and active development.

Hmm.. Just a couple of weeks ago there was strong sentiment from a player(s?) to fix open & existing bugs before allowing the implementation of new features.

http://sourceforge.net/tracker/?group_i ... tid=113833

Maybe I should have added an "As of" date to that post.
At the time of posting, that was the current disccusion.
My point , though, is that every new release of crossfire server seems more like a bugfix.
Perhaps it seems that way because there is no real release cycle. It either occurs or it doesn't. The world of CF doesn't seem as dynamic as it used to be. I really don't think there is enough community involvement needed to make it live.
The same graphics and basic gameplay are still in place when I fire up the client now in comparison to two years ago.. Well, excepting the sudden overpowing of monsters....

I find in CF that even speed becomes useless because the human player has trouble with the movement interface. What good is 5.30 speed if you can react with your arrow keys and lag time to a 2.5-like speed?

Something doesn't seem to be right... I just can't seem to place my finger on it...

[cavesomething]

Maybe I should have added an "As of" date to that post.
At the time of posting, that was the current disccusion.
My point , though, is that every new release of crossfire server seems more like a bugfix.
Perhaps it seems that way because there is no real release cycle. It either occurs or it doesn't. The world of CF doesn't seem as dynamic as it used to be.

Whilst I must confess to some degree of responsibility for that (I have not been as active in cf development over the last six months, for a variety of reasons I won't go into now) I think nonetheless a large part of that is that there have been a great number of changes in the last few months which have been occuring at a fairly low level, which aren't really user-visible - things like key-value pairs and extra map layers, even transports, since most of the game-world support for them isn't complete yet.

These sorts of changes are neccessary, even if you don't see any tangible results from them immediatly. - The same goes for the map editor, which ragnor has been doing a great deal of work on, and which, in future should make map creation faster and easier.

I really don't think there is enough community involvement needed to make it live.

I think this phrasing is an artifact of Non-Free software, the idea that there is somehow this divide between 'the community' and 'the developers' is true of things like World of Warcraft, but not crossfire, anyone can submit a patch and make the case for its inclusion.

There is an issue that some aspects of that process are somewhat tricky, more than they should be. Map creation is an obvious example of this (which goes back to Ragnor's work on Gridarta)

The same graphics and basic gameplay are still in place when I fire up the client now in comparison to two years ago.. Well, excepting
the sudden overpowing of monsters....

Well, yes, crossfire is still a multiplayer real-time RPG, it isn't going to suddenly mutate into a platform game - that's what jumpnbump is for.

If you mean that the client interface needs help, well that is a different issue, and mostly seems to be being dealt with in the gtk2 client - again something you should hopefully see suddenly 'change' when gcfclient2 becomes the default.

I find in CF that even speed becomes useless because the human player has trouble with the movement interface. What good is 5.30 speed if you can react with your arrow keys and lag time to a 2.5-like speed?

Agreed, but in principle that is a fairly trivial change, it just requires a /lot/ of playtesting to get right, something you may see come in fairly quickly once someone is prepared to babysit a testing server.

Something doesn't seem to be right... I just can't seem to place my finger on it...

I suspect that another issue is that it has been summer in the Northern Hemisphere, and daylight presents more options regarding activities that don't involve coding. As the nights close in, probably several developers will have more time to spend on coding - hopefully I will be one of them.

[Ryo]

As cavesomething says, people don't always work on the game

Lately i've been doing more fixes than new stuff (though there are some improvements here and there). Also i changed my priorities around, so CF isn't as high as it used to be.

The important thing to remember is that we're all volunteers (including players ^_-), so well, it's only natural we sometimes feel like not working on Crossfire. When not many work, the game seems to be evolving slowly. But, as opposed to many other projects/games, it does go on living

I know there are many things I'd like to see in the game, but either I don't take the time or want to write them myself....

But I do plan on finding time to develop, that's for sure


Of course, we can always use new people (artists, coders, map makers, ...). We just need to find'em

I just hope we don't have more forks, that's always painful

[Aaron]

well i think the game core and characters are pretty stable, its pretty much tweaking things until the 2.0 rewrite. one thing that seems to be neglected though is the client. i think native clients all adhering to the same design layout would be best. ie, get gtk out of windows and osx, and use something that people can install. unless of course someone who can get gaim, the gimp, and crossfire to work on the same windows system can take over with the windows client. and currently we are quite lacking in mac clients. a big hit for lots of players on non-easy-gtk systems.

[Rednaxela]

well i think the game core and characters are pretty stable, its pretty much tweaking things until the 2.0 rewrite.

One quick note about 2.0, in SVN, the 'trunk' branch is already for work on 2.0, and the 1.x branch is for 1.x which will probably have a number of additional releases before 2.0. Not much 2.0-specific has been done yet though, but in the near future some 2.0 code refactoring will start. Also, crossfire 2.0 won't be a rewrite, but it will have major refactoring of code to improve ease of development, new/changed game features, and many tweaks in many areas. There is no real set list of what is targetted for 2.0, but some ideas/planning around is here

one thing that seems to be neglected though is the client. i think native clients all adhering to the same design layout would be best. ie, get gtk out of windows and osx, and use something that people can install. unless of course someone who can get gaim, the gimp, and crossfire to work on the same windows system can take over with the windows client. and currently we are quite lacking in mac clients. a big hit for lots of players on non-easy-gtk systems.

Two points about that:
First, we need to decided on a design layout in the first place. It seems to be agreed by most people that the current designs are far from ideal. One issue with this that has been brought up and IMHO is a large part of the issue, is that currently the crossfire clients ride the border between a mouse-oriented application and a keyboard-oriented application, and that it would greatly enhance the UI if we were to make up our mind on which is the PRIMARY input method. Due to the nature of crossfire, I don't believe a mouse-oriented UI is well suited, and so far as I can see, the main thing to do to make it properly keyboard-oriented, would be to have an efficent interface for inventory management.
Secondly, I believe that if we want to natively support a much wider set of toolkits, we should think about either making some "XUL"-type system like used in mozilla, or as a simpler alternative, have some sort of "toolkit abstraction layer" in the code, that allows the same code to define how the UI is layed out and rendered. Personally I belive the later is better to go with, due to the first being over-elaborate and probably not worth the complexity for crossfire. One note, at first the idea of a "toolkit abstraction layer" might sound daunting, but I'm thinking it might not be too bad considering that we'll only need to have an interface for a small subset of toolkit capability. That said, it could be a little annoying to emulate spacing methods on UI toolkits that don't natively support them, however I feel that it would still be the best way to go if we do want to support a greater variety of toolkits natively.

[Aaron]

abstraction is fine, that would probably be the easiest to code too, and leave out the server-interfacing part. the main problem is lack of native clients on anything but linux. it was fine when i was only using linux, and the windows version can be hacked into a machine with minimal problems, but osx has nothing, because its just assuming you have a standard linux distro backend running.

on the mouse/keyboard thing, i think keyboard is the best way to navigate in a 2d world, but inventory is a hell of a lot easier with a mouse and drag and drop. all that really needs to be done is to clearly seperate the two interfaces, ie, autopickup needs a gui, and moving should be entirely keyboard based (sorry if that nerfs the sexy spellbook we just got.)

[Rednaxela]

abstraction is fine, that would probably be the easiest to code too, and leave out the server-interfacing part.

All current clients (not counting ones that arn't official ones) use the same code for the server-interfacing already. What I suggest is going a step further and having an abstraction layer on the UI as well, not just the low level protocol side. This way the code that defines the UI layout, is identical while using any GUI toolkit, making sure the UI layout is the same in all cases.

the main problem is lack of native clients on anything but linux.

Well, before making those statements, remember that linux isn't the only *nix that natively uses X11 and is easy to install GTK on.
Yes, I agree that is an issue though.

on the mouse/keyboard thing, i think keyboard is the best way to navigate in a 2d world, but inventory is a hell of a lot easier with a mouse and drag and drop. all that really needs to be done is to clearly seperate the two interfaces, ie, autopickup needs a gui, and moving should be entirely keyboard based (sorry if that nerfs the sexy spellbook we just got.)

Well, I'm not convinced inventory is a hell of alot easier with a mouse drag and drop. For example, nethack has a very good keyboard interface for inventory that I would say is much more efficent than our current mouse inventory interface. IMHO one of the bigger problems in crossfire is that inventory access is hell with the keyboard currently, and that making it at least as easy to use with the keyboard as it is with the mouse would greatly enhance playability. In order to make an ideally usable keyboard inventory interface though, I think one may need to move to an interface that is not as mouse friendly as currently, but IMHO, if the keyboard interface is good enough, that would be worth it.