I know I am annoying, but the rotation direction is wrong again.

Ha ha, I'm sorry... I forgot it... Do you know? what happens is that I'm used to game on right rotation (right but not correct)... I make another version on the correct rotation.

Regards

pd: Sorry, I was wrong and put a new thread. I have deleted it already...

The next test report:

1. The bottom of the play field is not marked in any way and the landing of the figures "in the air" is little bit unexpected. The same about the right border. Sometimes I simply miss to fill the right most column, simply because I am not aware it is there (the field is wider than usual and this makes the problem even worse).
2. The game crashes on 4 rows removing.

1. Yes, I guess that should paint borders
2. But not ever on 4th rows removing... at least for me. Many times I can ending the game (I have reached 9th level and over 3 hundred points ). Yes I will revise the code.

Thank youk, John

It crashes on $401f5b and not only on 4 rows.

too much bugs man, share your code please to see where are all these bugs.

Really?



I will revise it. The debugging task will be hard, I presume.



Well, some changes applied:

1) Now, there's no flickering in main window
2) Delimited borders with red lines

There're two versions of the program:
TetrisWF2.exe -> shapes rotates on the right
TetrisWF2b.exe -> shapes rotates on the left

I have noticed that sometimes the programs hangs up yet. I have revised the program and pusha/popa in everywhere I could to avoiding hidden changes on registers values, specially on windows internal functions callings. Hence I will do this two things:

1) Revise variables' type casting.
2) Get non-initialized variables into reserved memory

If none of them works, the debugging task would be very hard, because most of times the program works well, so it is difficult to find the characteristic of the error.

Regards

Sorry, but it crashes instantly on the start (inside user32.dll). Debugging shows the call from $401125 to the function UpdateWindow. The reason is probably bad arguments.

Yes, your are right, it fails at the first glance, but only in WXP, in W7 works well (sometimes hanged up, as usual)... Well, I have W7 and I was able to prove it many times with this bug, I don't know how could do it W7 for not take it in account... I feel that I have to give an explanation: let's say that for implement double buffering in main windows I had to recover an extra parameter in a central procedure but, because I recover parameters by hand, I forgot to add it in its RET sentence. I cannot get out of my surprise thinking how W7 gives the correct offset on return to EIP.

Anyway, time and tests lost... Sorry for inconveniences. This error is now corrected in both versions.

It works now and it seems to be more stable. At least I was able to play 7..8 games without crash.

Regards.

Thank you a lot, JohnFound. It is possible that some registers changed its value "hiddenly" when I called system functions, because them not only changes eax, but sometimes any others. So, it is possible that I fixed this with those pusha/popa... This would be great.

Regards.

avcaballero: All Windows API calls use either standard-call or c-call convention. Both of these calling standard conventions state that EFLAGS, EAX, ECX and EDX can be clobbered at any time. EBX, ESI, EDI and EBP are all preserved at all times.

crashed when i held down the left arrow key to get one block to "get under" another one, kind of spammed the left key not to make any space unused.
bottom row if it makes any difference.

bubach, if I didn't understand you wrong, the bug appears when you hold the left key down. Do you have the lastest version? I have fixed in one version that when hold down any key (left or right) to get the figure under other, the first one put over the second before reaching the bottom of the second. This would be fine in the latest version. Also, in a previous version a hidden shape drops hiddenly and the next shape held static apparently before the end because it is above the previous hidden shape... this is also fixed and the latest version should works fine.

The latest version works well almost every time, but from time to time keeps on crashing. I have to review the code when I have time.

I don't know if this is what you mean.

Thank you