Age | Commit message (Collapse) | Author |
|
svn-id: r19124
|
|
Now you may clearly see rotated numbers as well as hint when your code
is wrong.
svn-id: r19088
|
|
shows numbers, not just rotation of them. Still no screen updated during the
process.
svn-id: r19067
|
|
which fixes bug #1324814 "GOB1 ega: lock up when game is over"
svn-id: r19056
|
|
svn-id: r19027
|
|
turning on the screen auto-dirtying - some as Simon uses. It's not perfect
by any means, but it may be good enough for 0.8.0 at least.
svn-id: r18649
|
|
svn-id: r18639
|
|
svn-id: r18604
|
|
Improved debug output.
svn-id: r18556
|
|
svn-id: r18529
|
|
svn-id: r18444
|
|
typ of char regardless of any compiler flags. Now looks very nice on the Symbian platfor.
svn-id: r18435
|
|
svn-id: r18421
|
|
on all targets. On symbian a char is unsigned char but Gob is assuming int in this case.
len = *(int8*)inter_execPtr++; // must be a signed char typ and char is not default signed on all platforms.
svn-id: r18393
|
|
svn-id: r18392
|
|
for symbian build
svn-id: r18391
|
|
function is used.
svn-id: r18390
|
|
svn-id: r18360
|
|
svn-id: r18268
|
|
it will be useless at least for some months :)
svn-id: r18233
|
|
verified by letting ScummVM compute the checksum (note: ScummVM runs into a deep freeze running the gob1 target for the woordruff game data, very bad)
svn-id: r18157
|
|
svn-id: r18137
|
|
svn-id: r18106
|
|
svn-id: r18084
|
|
other game detection-related changes, but they're mostly cosmetical at this
stage.
svn-id: r18082
|
|
svn-id: r18040
|
|
because we already have many classes with that name)
svn-id: r18039
|
|
svn-id: r18037
|
|
may not be correct after all, so I've removed it.
svn-id: r18002
|
|
svn-id: r17993
|
|
svn-id: r17975
|
|
svn-id: r17944
|
|
demos I have, except for the Woodruff game which I don't have on this
particular computer. (I'll add that one later, if no one beats me to it.)
svn-id: r17936
|
|
svn-id: r17935
|
|
svn-id: r17921
|
|
don't know anyone who can verify that.
svn-id: r17914
|
|
svn-id: r17913
|
|
svn-id: r17908
|
|
I've changed it to something that doesn't trigger assertions in the
Gobliiins demo intro.
svn-id: r17907
|
|
svn-id: r17906
|
|
svn-id: r17900
|
|
to not be played, and probably fixes wrong frequency on some that were
played. It's all guesswork, though. I don't know if this is what the
original did.
svn-id: r17882
|
|
svn-id: r17872
|
|
svn-id: r17865
|
|
pretty unnecessary function - we could call RandomSource directly - but
let's take this one thing at a time.
svn-id: r17863
|
|
assumed, given their tiny size, that the sound samples are 8-bit mono.
Looping is not implemented yet.
Some sounds are skipped since the engine asks that they be played at a
negative sample rate. (I think there is more to this than just a simple
signed/unsigned issue, because they sound wrong even if I treat the
frequency as unsigned.)
svn-id: r17862
|
|
close to each other on the ground. (Happened to me on the first level after
destroying the voodoo doll, where I'd drop the banana, the soap and the
false nose close to each other on the ground after using them.)
Reasoning behind the change:
From what I understand, map_itemsMap[] contains information for each "cell"
of the map about which objects are there. Each cell can contain two objects
which are stored in the upper and lower byte of a 16-bit word.
When dropping an object, it is written into map_itemsMap[], but not just to
the indicated cell but also to a few of the surrounding ones. Presumably to
make it easier to pick it up afterwards.
When writing an object to a cell, we check if one of the bytes is already
occupied. If it is, write to the other byte. Otherwise, write to that byte.
(If both bytes are occupied, one will be overwritten.)
The old code assumed that if one byte was free at position (x,y) the same
byte would automatically be the free one in the surrounding cells. This
could cause bad values in the array, since the item was added to an
existing value, rather than replacing it.
This new code makes the check for each cell that is modified. (It also gets
rid of some code duplication.)
svn-id: r17851
|
|
e.g. in the German cutscenes.
svn-id: r17805
|
|
support for other languages appears to work at least fairly well, and this
makes it easier to test it.
svn-id: r17804
|
|
we'll end up trying to free invalid pointers. (Happened to me at the end of
the level where you get the voodoo doll.)
svn-id: r17802
|