|
Post by Mairtrus on Feb 20, 2010 10:34:26 GMT -5
There is a terrible bug in the command LOADTILES, as well as the command LOADSCD (when used to load tiles). The [Tile Source Offset from Label] argument isn't coded properly (I think it was since the first version of BEX).
The problematic code is:
move.l #Tile Source Offset from Label,d0 lsl.[b]w[/b] #5,d0 move.l #__DATA_PTile Data Label,a0 add.l d0,a0 But it should be:
move.l #Tile Source Offset from Label,d0 lsl.[b]l[/b] #5,d0 move.l #__DATA_PTile Data Label,a0 add.l d0,a0
|
|
|
Post by TheMVRules on Feb 25, 2010 12:00:07 GMT -5
A new bug in the FAKE command: "Phase error"
|
|
|
Post by Tom Maneiro on Mar 12, 2010 11:25:11 GMT -5
|
|
oompa loompa
I AM THE GOVERNATOR
"Git 'Er Dun!"
Posts: 1,301
|
Post by oompa loompa on Mar 12, 2010 12:29:27 GMT -5
Thanks for adding the bugs I indeed only look at this thread when I'm about to make a release. Makes me feel confident that more people are using BEX (these bugs should have been fixed a long time ago!). If the bug is not in here, then I pretty much don't fix it because I know I'll forget about it (yea short term memory effects, but who cares ) I'll add some more too
|
|
|
Post by TheMVRules on Mar 26, 2010 13:01:23 GMT -5
I said on the Some ideas for a update thread: SRAM broken. Only strings work.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 28, 2010 19:40:58 GMT -5
Found another bug / undocumented-limitation concerning operators. Bit shift operations that are larger then 8 fail to compile ( regardless of type integer / long ).
|
|
|
Post by Tom Maneiro on Mar 31, 2010 0:52:21 GMT -5
In my rudimentary attempts at doing my own assembler code, i've found a few interesting bugs:
1) regmove.x &h0100, a0 fails to compile with an ASMX error:
basic.s:213: *** Error: Missing operand *** basic.s:213: *** Error: "," expected *** 00340C move.l #&h0100,a0 Workaround:
addr& = &h0100 regmove.l addr&, a0
2) I don't know if this is a bug or my lack of knowledge about passing memory addresses to the address registers, but, this fails:
addr& = &h0100 x = TestAX(addr&) print "Result: "; chr$(x) 'should print "E", but does nothing declare asm function TestAX(a0 as long) MOVE.W (A0), d0 end function
Workaround: abuse regmove.l
addr& = &h0100 regmove.l addr&, a0 x = TestAX() print "Result: "; chr$(x) 'works, prints "E" declare asm function TestAX() MOVE.W (A0), d0 end function
|
|
|
Post by TheMVRules on Apr 8, 2010 7:28:51 GMT -5
For error 1, try instead: regmove.l 128, a0
|
|
|
Post by Tom Maneiro on Apr 20, 2010 23:12:31 GMT -5
A couple of IDE-crashing bugs just surfaced today: - Try to open or save a assembly file (.s): KRASH! (Error 424: An object is required) - An EXPLICIT division by zero in a print statement ("print 1/0" will do) will crash the IDE when compiling (Error 5: Invalid argument or procedure call). Not that forcing a divide by zero is a great example of "programming"... but surely this can qualify as a DoS
|
|
|
Post by TheMVRules on Apr 21, 2010 11:02:39 GMT -5
The BEX IDE is a little buggy and also have compatibility issues with Vista+ (i suspect DirectX 9 problems). Gimme command line!!!! In the next version plz!!!!
|
|
oompa loompa
I AM THE GOVERNATOR
"Git 'Er Dun!"
Posts: 1,301
|
Post by oompa loompa on May 3, 2010 12:33:30 GMT -5
Hi all, Good news. I found the sources for BasiEgaXorz again (after the update from LAKABAJO), so post your bugs now
|
|
|
Post by TheMVRules on May 3, 2010 13:16:33 GMT -5
Good. Read the above posts maybe. We've got a load of bugs reported since last release. But I repeat - FAKE is broken, ASMX says "Phase Error". It hasn't happened in previous versions. And then please fix SRAM, both I and theelf needs it for our projects.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 10, 2010 18:58:45 GMT -5
Small ( but annoying ) compiler issues.
- Arithmatic operations ( Addition / Subtraction ) don't compile when separated from their target variable.
a+= 1 ' Works a += 1 ' Doesn't work
- Immediate operators don't compile ( + crash the IDE ) when separated from their target variable.
a++ ' Works a ++ ' Doesn't work
- Arithmatic operations don't follow standard conventions. Division should be applied first, multiplication second, addition / subtraction last.
a = 1*2+1 ' Output is 3 ( Correct ) a = 1+1*2 ' Output is 4 ( Incorrect )
All tiny issues that can be easily avoided ( thus low-priority ).
|
|
|
Post by TheMVRules on Jun 14, 2010 9:22:45 GMT -5
And here's another critical bug..
Commented IF's don't work.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 7, 2011 1:50:23 GMT -5
It seems INCLUDE demands that the library be in the main BasiEgaXorz folder. Usual behavior has been that BasiEgaXorz looks for tiles or fonts in the same dir as the source .BEX file. Why the change for BEX BASIC libraries? Maybe libraries should be expected to be in a C:\BASIEGAXORZ\LIB folder, yeah? Oh, and BasiEgaXorz poogies itself and exits when it cant find the library Finally, any code or data in the library doesn't seem to be accessible even when the library is found. When you hit F5 to compile it throws up a message like the libaries functions and DATA statements aren't there. I am compiling a common community BASIC library with handy code plus needed workarounds (real hardware save support). It would be awesome to get this INCLUDE function working For now, I'm wondering to what extent the INCLUDE function IS working. Is it usable in any way in this release of BasiEgaXorz? Not being flippant here - I'd really like to use it!
|
|