Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Robsoie]

Ah ok, i wasn't sure if it was just uneeded additional micromanagement, good to see that there's then a purpose.

[Frumple]

Hoi! Requested crash replications: For summoning in chasm rooms -- just keep summoning at the critters on the island until the error pops up. Continuing will crash the game.

For firing such that the ammo can't land -- just fire directly south. This is a straight CtD, no in-game error message.

Both the save file and a crash dump are included in the .rars.

I'll see about getting these on the tracker, but wanted to make sure it was out ASAP. Because yeah, if you get those two fixed, that will correct probably 70+% of Inc's instability issues as of H4. Those two were literally what was causing the majority of crashes.

[chooseusername]

For firing such that the ammo can't land -- just fire directly south. This is a straight CtD, no in-game error message.

When you fire a missile, and it collides with an obstacle like the wall or whatever in this case, it is placed at the end of it's path before that wall.  There are no terrain checks whether it should be placed there.  Next there's an attempt at "placing it near" to that placement position - I assume this could be considered a bounce off the obstacle.  This does do terrain checks, and if no nearby tile can be found which is not marked as TF_WARN, then the missile is destroyed.  The code assumed if the placement succeeded, the bounce would also, hence the crash.

I've checked in a fix for this.  But there are a decent number of TF_WARN terrain types, and it is questionable whether a missile cannot be placed on all of them.  Deep water, magma, brimstone.. perhaps.  But there are ones like acid where the placement might succeed if the type of missile is not the type of item the acid degrades.  Anyway, that's a problem for another day.

[teoleo]

the best game i have played

[chooseusername]

Hoi! Requested crash replications: For summoning in chasm rooms -- just keep summoning at the critters on the island until the error pops up. Continuing will crash the game.

Fix checked in.  This was because the logic which decided that the monster would get falling damage (for landing over the chasm and falling down it) did not check if this damage killed the monster, and the logic then went on to crash.

Both these fixes will be in the next release.

[Frumple]

Cheers, choose. That's going to be a major benefit to the game's stability.

For other folks trying out the new stuff... anyone else having what I'd guess you'd call performance issues? Game's running fairly smoothly (well, except for character generation -- rolling stats and whatnot are taking considerably longer than they did previously), but it's also pretty consistently chewing up the entirety of one of my CPU cores. Fairly sure this is some kind of video card/etc. issue on my end, but I'd like to know if anyone else is having the same trouble. Like to get some feedback on the issue before I submit anything on it to the tracker.

Similarly, is anyone else having rendering (I guess that's what it is, anyway) stop after a window focus change? I can swap over to another program or minimize, and when I come back to Incursion the screen will be blank until I enter some kind of action -- move, pull up the character sheet/inventory, etc. Again, fairly sure this is some kind of video/renderer issue, probably on my side more than libtcod or whatev', but it's more than a little annoying, ha.

[BishopX]

Two, actually, if my memory's not failing me. The big and definite one is Asherath, who rewards favor when you kill stuff that's higher CL than your level. Bigger difference, bigger reward. The lesser one, and one I'm less sure of, is the fact that, if I'm not misremembering, you gain more XP when you fight greater challenges (challenges being determined by difference between level and enemy CL.). If you hold off leveling for a bit, you can actually get more XP, overall.

But it's mostly Ash. Double checking the manual, at least Khasrach also has a CL component to sacrifices -- she only accepts (or, at least, is only supposed to.) kills that are equal or greater power than you. You can also tick Aiswin off if you level up before killing something that badly wounded you. So there's a few god-related reasons not to, basically, and one pragmatic one.

Google breakpad or whatever it is called, is integrated.  It might dump a minidump file when the game crashes, which should allow debugging of whatever caused the crash.  But a save game at the point where it can be reproduced with a series of attempts would be most cromulent.

Sure, sure. If no one beats me to it, I should be able to produce save files and reproduction steps sometime later tonight or tomorrow. Either one of them is very easy to reproduce in wizard mode.

Also, I believe that monsters are scaled by a function of your level and dungeon level when they're generated.

[Robsoie]

Similarly, is anyone else having rendering (I guess that's what it is, anyway) stop after a window focus change? I can swap over to another program or minimize, and when I come back to Incursion the screen will be blank until I enter some kind of action -- move, pull up the character sheet/inventory, etc. Again, fairly sure this is some kind of video/renderer issue, probably on my side more than libtcod or whatev', but it's more than a little annoying, ha.

Yes i see something similar, if i alt tab to another program and come back to Incursion, the rendering is not catching up, i have to move/press spacebar/etc... to actual have the window (i play windowed 800x600 in case it makes a difference) refreshing and displaying the actual game again

For the CPU usage, i have not noticed a problem, and looking at my task manager , it seems to take +/- 55% of both my dual core when running (though i have firefox opened with a few tabs, that usually adds to the cpu usage) with a spike from time to time (using the latest version from bitbucket).
But i have not adventured to more than depth 2 so far

[Frumple]

Also, I believe that monsters are scaled by a function of your level and dungeon level when they're generated.

That... is something I'd have to defer to a code diver to check. I'm fairly sure critter generation is primarily dungeon level based, secondarily room based (vaults and floating rocks, ferex, have higher CL stuff), and tertiarily randomized somewhat if Out of Depth monsters is active. If player level is heavily involved I'd be a lil'surprised.

I usually play with OoD off (and challenge mode on -- it evens out the difficulty a bit and lets things get crazy in the end, without being nightmare level crazy), so...

[chooseusername]

For other folks trying out the new stuff... anyone else having what I'd guess you'd call performance issues? Game's running fairly smoothly (well, except for character generation -- rolling stats and whatnot are taking considerably longer than they did previously), but it's also pretty consistently chewing up the entirety of one of my CPU cores. Fairly sure this is some kind of video card/etc. issue on my end, but I'd like to know if anyone else is having the same trouble. Like to get some feedback on the issue before I submit anything on it to the tracker.

This is fixed already, and will also be in the next release.

Similarly, is anyone else having rendering (I guess that's what it is, anyway) stop after a window focus change? I can swap over to another program or minimize, and when I come back to Incursion the screen will be blank until I enter some kind of action -- move, pull up the character sheet/inventory, etc. Again, fairly sure this is some kind of video/renderer issue, probably on my side more than libtcod or whatev', but it's more than a little annoying, ha.

Not happening for me, and heard no tell of others with this problem.

[Rolan7]

It's happening for me on Win7, but I don't consider it a problem particularly.  I guess technically.

[chooseusername]

It's happening for me on Win7, but I don't consider it a problem particularly.  I guess technically.

It sounds like there are at least three people are experiencing it.  It's likely just an SDL2 windowing event that libtcod needs to better respond to.

I also use Windows 7, but don't see it.  Does it happen at any particular stage of the game? Is it certain types of windows?  Can someone give a really simple and well described reproduction step?

I can't really fix this unless I can reproduce it, as otherwise it will be stabs in the dark.

[Frumple]

The reproduction step for me is literally for the Incursion window to lose focus for any reason. Alt-tab, minimize, whatever. Upon returning to the screen, some sort of action has to be performed to cause the window to update. It happens in any window*, any game state, etc. It happens full screen just as readily as windowed, as well.

I would agree it's probably something SDL related, though -- I've had similar issues with ToME4, which is SDL based, and whatever was causing it wasn't consistent across users. Is there anything beyond reproduction steps that might help?

*E: Ah, though if you mean different versions of the windows OS, I'm currently on Vista. I'll actually be able to test on a few different computers next week (local school's campus) and see if the behavior is the same there, but that'll be two or three days from now.

[beorn080]

The reproduction step for me is literally for the Incursion window to lose focus for any reason. Alt-tab, minimize, whatever. Upon returning to the screen, some sort of action has to be performed to cause the window to update. It happens in any window*, any game state, etc. It happens full screen just as readily as windowed, as well.

I would agree it's probably something SDL related, though -- I've had similar issues with ToME4, which is SDL based, and whatever was causing it wasn't consistent across users. Is there anything beyond reproduction steps that might help?

*E: Ah, though if you mean different versions of the windows OS, I'm currently on Vista. I'll actually be able to test on a few different computers next week (local school's campus) and see if the behavior is the same there, but that'll be two or three days from now.

I think it has to do with the display code not issuing an update all too often. Or rather, the display only updates when something happens on the screen. When you cover the screen or switch to a different window and back, the screen is blank because there is nothing to draw.

If you want to test it for them, add a screen refresh key, and have them press it. Not rest, but a key that forces a redraw to occur.

[chooseusername]

*E: Ah, though if you mean different versions of the windows OS, I'm currently on Vista. I'll actually be able to test on a few different computers next week (local school's campus) and see if the behavior is the same there, but that'll be two or three days from now.

I wouldn't worry about it.  It's not so much whether it happens, or which OS, but why it's not happening for me.  And which event should be listened to, in order to properly handle it, for people who are experiencing it.