Re: Brain Party for Linux (and now for N900)
 

I had a quick look to the source of brainparty and didn't guess why this problem is happening. If it is indeed writing gigs of log files it would cause similar symptoms.

Btw, call SDL_GLES_Quit() (~ equivalent to eglTerminate()) before SDL_Quit(). The N8x0 leaks _a lot_ without such a call, but the N900 EGL driver seems to be more robust in this regard (so this won't improve the N900 situation).

Re: Brain Party for Linux (and now for N900)
 

First, nice game.

Second, as it has been said, it slows down. First test is fine, then it starts glitching, at first only noticeable via back animation. After the first test, it's noticeable. After the second, getting worse. I can't finish the third, it's so slow UI registers touches in the next game.

However, there is no need to remove battery. First, with patience the game exits, even if each menu takes 30 seconds. Secondly, leave it be and it gets killed after a few minutes (or crashes).

Also, a feature like best score per game would rule. Also, best score ever could also be cool. Finally, what score should I get in baloons? I pop quite a bit and get failed. I only got it once and with score in the hundred thousand I got a C- or so. What am I missing? Is it luck, too? I have A* in all other games minus the sheep (it's unplayable by the time I get to it)

Also, excluse typos, N900.

Re: Brain Party for Linux (and now for N900)
 

strange...
I've never experienced any problems with the game. No slowdowns, no crashes, no leaks...

Re: Brain Party for Linux (and now for N900)
 

Thanks for the hint; I've added this call now and BrainParty 0.5-3 is on its way to the autobuilder.

I don't know if I can do anything about the test scores and settings being written out to the file (/home/user/.brainparty). The code is in BPGame.cpp in the LoadSettings and SaveSettings methods of the BPGame class if anybody wants to have a look.

Re: Brain Party for Linux (and now for N900)
 

Try not to pop the last remaining one on the screen right away, but wait until the next one appears before popping it (so that all the time at least one balloon stays on the screen). After balloons stop appearing, pop the last one and you won't get failed.

No problem - typos exclused :p (SCNR)

Re: Brain Party for Linux (and now for N900)
 

Hey!

<< Finally, what score should I get in baloons? I pop quite a bit and get failed. >>

A good player can score over 500,000 in a game. I can get 50,000 in about 20 seconds. Remember, it's a brain game: the key isn't just to tap the colours blindly. Instead, you need to think, "well, if I pop two red balloons there and three green balloons there, then I end up with 16 purple balloons all lined up" - and then you score big points.

Your score doubles for every extra balloon you add to a chain, so it pays to think.

Andrew_85: I just released 36 minigames and you're saying you want even more? :)

lma: I'm sorry; I can't help you with that bug, because I don't have a device. It's strange, though, because it works fine on other platforms. Perhaps C++'s ofstream is slightly different on the N900? What exactly is making the file so big - is it repeating itself? Is it whitespace?


Paul

Re: Brain Party for Linux (and now for N900)
 

Attaching a copy. It seems to repeat sequences of scores over and over, and there's an interesting pattern emerging when counting them up:

Re: Brain Party for Linux
 

Ah, thats what we did wrong! Thanks!

Re: Brain Party for Linux (and now for N900)
 

lma: I reckon I can see what's causing that, and it's an easily fixed bug. Go to the LoadSettings() function in BPGame.cpp and add this line at the top:

AllTestScores.Clear();

Now save and re-run make. It should fix the problem.

Re: Brain Party for Linux (and now for N900)
 

I don't think this bug is part of the port, but the "Next in line" puzzle contains some errors. On some puzzles there might be more than one mathematically correct answers but only one of them is acceptable. This is not a big flaw, but there exists some real errors as well. Atleast I've found one or two flaws where you had to factor the given values with a number, but one of the calculated values in the line was incorrect. Therefore I assume that the puzzles are just saved strings and not actual calculations?