Post by lamarr on Jun 7, 2010 4:57:30 GMT -5
I've recently discovered Portralis after having been burnt out on roguelikes for a while now after playing too much Crawl Stone Soup and I'm enjoying it a lot, on a conceptual level. For one it's nice to actually see a roguelike with a plot to follow and for the other Portralis has a really nice set of innovative, unique features that I haven't seen in any other roguelike (or many other games, really).
The multiclassing system is really fun to play around with, the classes are varied and have a lot of interesting abilities. I like the flexible crafting and spell systems, which give you a lot of freedom to create your own gear. The large amount of skills is very nice, and the feats you can get at certain skill levels make putting points into them more meaningful than simply getting another +10 to magic rolls or whatever.
I like the Monster Mage class and the Monster race option in particular. The Monster Mage gives you a lot of interesting tactical options, while the Monster race is awesome for its uniqueness alone. I created a Lesser Ghost character and was amazed to find that I could actually walk through walls! Really cool, most roguelikes that offer similar options really only give you a different representation on screen and maybe a few inherent spells that are made obsolete by the time you hit level 3 or so. It reminded be of the Possessor class in ToME, which I liked for similar reasons.
There's a lot more I could talk about, the game is really chock-full of interesting little things, but that'd take forever, so I'll get to the stuff that's bad instead. And it really IS bad. No offense, Variaz, but on the technical side your game is a huge, badly documented mess. Would I be right to guess that you're not a programmer by trade? Because it kind of shows.
I've been looking at the source, because I like to compile my own software and maybe make a few personal modifications and bugfixes here and there. But I can't even figure out how! The makefiles are horribly out of date and obviously from way different versions of the game, there is no documentation at all, I can't even figure out what dependencies (if any) I need. Any attempts at actually compiling the source (either under Linux via Make or Windows via Visual Studio) fail horribly, with literally HUNDREDS of warnings about mismatched type conversions, missing return values, unsafe functions, variables that are declared but never used... Seriously, it's horrible.
And the code isn't much better. The only properly documented sections of the code are those that look like they're left over from some earlier version of Angband. Huge sections of some classes are simply commented out for no apparent reason. Everything is completely jumbled, with tons of unrelated game elements sitting in the same files instead of being split up into a more manageable format. Honestly, I'm not surprised progress is so slow on this project, it looks completely unmaintainable.
If I could make a recommendation, I'd say to set this up as a Sourceforge project or something a.s.a.p. and implement some version control first thing. Tons and tons of pointless stuff seems to only exist anymore because Variaz can't remember what it's for or why he did it, which is exactly what version control is meant to prevent. Second, set up a real bug tracker instead of using a confusing forum thread full of chatter. Third, document and comment everything. Seriously. Once you've done that, you could try getting some other people to join the team, if only so you'll have more people for testing and bugfixing (and maybe spellchecking -- "shaftses" sounds cute, but it isn't good English).
If you want to look at a roguelike project done really well, look at Crawl Stone Soup. It's positively boring and spartan in comparison to Portalis as far as the feature set goes, but the degree of organization is amazing and the code is a joy to look at. I've taken the whole thing apart and put it back together backwards without having to google for anything, simply because it's so well structured and documented.
Anyway, that's what I have to say about this. I hope nobody is offended by this post, because I said a few pretty unflattering things, but I really like this variant so far and I'd be sad to see it die or become even more obscure because of some bad project management choices. I'd be glad to help with some of this stuff as well, since I've currently got a bit of free time going until college starts up again in the fall.
Oh, and if someone could tell me how to get this frickin' thing to compile, I'd be deeply appreciative.
The multiclassing system is really fun to play around with, the classes are varied and have a lot of interesting abilities. I like the flexible crafting and spell systems, which give you a lot of freedom to create your own gear. The large amount of skills is very nice, and the feats you can get at certain skill levels make putting points into them more meaningful than simply getting another +10 to magic rolls or whatever.
I like the Monster Mage class and the Monster race option in particular. The Monster Mage gives you a lot of interesting tactical options, while the Monster race is awesome for its uniqueness alone. I created a Lesser Ghost character and was amazed to find that I could actually walk through walls! Really cool, most roguelikes that offer similar options really only give you a different representation on screen and maybe a few inherent spells that are made obsolete by the time you hit level 3 or so. It reminded be of the Possessor class in ToME, which I liked for similar reasons.
There's a lot more I could talk about, the game is really chock-full of interesting little things, but that'd take forever, so I'll get to the stuff that's bad instead. And it really IS bad. No offense, Variaz, but on the technical side your game is a huge, badly documented mess. Would I be right to guess that you're not a programmer by trade? Because it kind of shows.
I've been looking at the source, because I like to compile my own software and maybe make a few personal modifications and bugfixes here and there. But I can't even figure out how! The makefiles are horribly out of date and obviously from way different versions of the game, there is no documentation at all, I can't even figure out what dependencies (if any) I need. Any attempts at actually compiling the source (either under Linux via Make or Windows via Visual Studio) fail horribly, with literally HUNDREDS of warnings about mismatched type conversions, missing return values, unsafe functions, variables that are declared but never used... Seriously, it's horrible.
And the code isn't much better. The only properly documented sections of the code are those that look like they're left over from some earlier version of Angband. Huge sections of some classes are simply commented out for no apparent reason. Everything is completely jumbled, with tons of unrelated game elements sitting in the same files instead of being split up into a more manageable format. Honestly, I'm not surprised progress is so slow on this project, it looks completely unmaintainable.
If I could make a recommendation, I'd say to set this up as a Sourceforge project or something a.s.a.p. and implement some version control first thing. Tons and tons of pointless stuff seems to only exist anymore because Variaz can't remember what it's for or why he did it, which is exactly what version control is meant to prevent. Second, set up a real bug tracker instead of using a confusing forum thread full of chatter. Third, document and comment everything. Seriously. Once you've done that, you could try getting some other people to join the team, if only so you'll have more people for testing and bugfixing (and maybe spellchecking -- "shaftses" sounds cute, but it isn't good English).
If you want to look at a roguelike project done really well, look at Crawl Stone Soup. It's positively boring and spartan in comparison to Portalis as far as the feature set goes, but the degree of organization is amazing and the code is a joy to look at. I've taken the whole thing apart and put it back together backwards without having to google for anything, simply because it's so well structured and documented.
Anyway, that's what I have to say about this. I hope nobody is offended by this post, because I said a few pretty unflattering things, but I really like this variant so far and I'd be sad to see it die or become even more obscure because of some bad project management choices. I'd be glad to help with some of this stuff as well, since I've currently got a bit of free time going until college starts up again in the fall.
Oh, and if someone could tell me how to get this frickin' thing to compile, I'd be deeply appreciative.
). It's about getting the feature implemented in the least time possible.
