The official release of fasm 1.54 is now available in the Download section.

The biggest changes are in the Win32 GUI version - the interface contains rewritten AsmEdit control, support for CHM help files, updated set of macros for Win32 programming and some other small improvements.

All versions are now able to automatically generate the name for output file, so the "output" parameter in the command line versions is no longer obligatory.

awesome.awesome.awesome. I Like.

;fasm1.54 for win98 key chinese for good ,
;(1)fasmw.asm
section '.data' data readable writeable
zzz dd 0 ;(1)fasmw.asm row 36 <===add it
_caption db 'flat assembler ',VERSION_STRING,0

;(2) Asmedit.inc
convert_to_ascii:
mov ax,word [lparam+2]
and eax,7Fh
lea ebx,[kbstate]
lea edx,[char]
invoke ToAscii,[wparam],eax,ebx,edx,FALSE
or eax,eax
jz ignore
mov [zzz],1 ; (2) Asmedit.inc 628<== add it
mov al,byte [char]
cmp al,20h
jae character
cmp al,0Dh
je character
cmp al,9
je character
jmp ignore

wm_char: ;(3)search wm_char
; test [lparam],1 shl 31 ;<===del it
; jz ignore ;<===del it
cmp [zzz],1 ;<===add it
jne .wm_Ascii ;<===add it
mov [zzz],0 ;<===add it
jmp ignore ;<===add it
.wm_Ascii: ;<===add it
cmp [wparam],8 ;<===add it
je ignore ;<===add it
mov eax,[wparam]
jmp character

I must ask, shouldn't you fix the perimissions for the output on the linux version?
when diffing I see this: (among else)
(diffing 1.44 against vanilla sources + permission fix, I'll diff for the libc version too later)


btw, tsunghsien_lin, please put the code in a [ code ] block [ /code ] (without the spaces between [ and ] ).

These permission are as were intended.

ok, why do you want the output to have -rw-r------? it should (IMO) be (if the umaks so allows) -rwxrwx-r-x. It seems gcc does set x for at least the owner - I think fasm should too - else I'll ahve to chmod after compile :/ (or patch the sources )

The "chmod after compile" is exactly what was my intention. fasm sets only the minimum, "safe" permissions.

i thought it was because users seldom see the FASM source code.

I don't see the problem with setting rwx for the current user; if I just ran a command to assemble a program into runnable object code, I want to be able to run it without more steps

ok, I see your point. But most other compilers are "unfase" in this point of view. I don't really see the danger in setting to my suggested permissions, as:
1) to excute a file you must type "./file", unless a) it's in any of the dirs in $PATH b) you got "." in $PATH - neither a or b are likley to be done by the average coder (unless (s)he's not practising safe-hex)
2) Usually you build you sources in some dir that's only avalible to you(owner) and those you want to have access (group), and when you release the files, you want the world (others) to beavle to read and exec it. Having o=rx (other = r-x) is safe, as long as the dir you keep the files in don't have o=x (other = --x ) - as the ousiders can't enter the dir.

btw here are three patches:
http://w1.364.telia.com/~u36404088/fasm/

fasm-patch-1.53-sci0-to-1.54-sci0.patch.bz2 - probably useless to you
fasm-patch-1.54-to-1.5-sci04.patch.bz2 - patch for the vanilla 1.54 to set the permissions -rwxrwxr-x on output
fasm-vanilla-patch-1.53-to-1.54.patch.bz2 - patch the vanilla 1.53 to the vanilla 1.54. (no permission fix here)

apply with "bzcat file | patch -p1" when in the fasm dir.

It's only one of the fasm's output formats that can be executed directly in Linux, and even this formatting can be used to make programs for other systems (like BSD or BeOS). So I thought it's better to leave it to the programmer to set up the permissions (when he knows what he is doing) - fasm is not able to guess it.

well, it doesn't harm to have eg PEs runnable, and really how often does one compile PEsunderlinux? But ok, I'll keep bugging the forum with patch(es) to set the permission on every major release (if it's ok with you)

Of course it's OK.

What do you think about developing the fasm IDE for Linux? I'm already planning making IDEs for DOS and maybe Menuet - I'm trying to make my editor core independent from the interface in the same way as fasm's core, you can see the progress of it in the current fasmw release.

I'd love to have the FASM IDE for Linux! If you could do it as a console app, that'd be great, but I would also use an X11 version.

good idea, maybe something similar to the source/[os] dirs (maybe (source/[os]/ide ), sort of abstracting gui calls in some similar way to the system.inc files.

tip: as new FASMW version can handle CHM help files, you can set freshguide.chm file from fresh package as your help file. This will give you context help on all FASM directives and all supported instructions.

why don't people use VIM under linux?
I use it under Windoze and it works, so under Linux it should be native (my opnion of course, i never worked under linux).
I think tings like parser-stage skipping not-taken "if" blocks would be better, but maybe linux users would like FASM native editor.

Should it be copy of AsmEdit or something new?

Yes, I suggested this because I'm making Assembly Editor core system-independent so it could (hopefully) be used as a base for fasm IDEs with "common look and feel" for different systems.