Sorry for the delay...
Rugxulo wrote:
> Hey, I noticed (when periodically checking) that
0.99.05 was released.
Right. When we make a release, the very last step is
"send email
notice?", and a count on how many people are being
notified. I have no
idea how to *get* on that list...
> But there's already a DOS 32-bit binary. So I
downloaded it. Well, it
> seems to be the same build as the daily snapshots
Yes. I just downloaded 'em from zytor and uploaded 'em to
sf. A "release
build" - different from the "snapshots" in
name, but the same build (AFAIK).
> (which didn't work
> for me on Vista although my DJGPP builds do despite its
current 32 MB
> DPMI limit).
That's a problem!
> However, upon further testing, I can run "dos32a
nasm
> myfile.asm -o myfile.com" and it works.
"-hf" and "-r" don't work,
> though, but "-v" does. Kinda weird if you ask
me.
Well, we've been "advertising" for some time that
"-r" was going away -
gotta keep our word! "-hf" should work... I
noticed some kinda weirdness
with the dos4gw build under dosemu - I forget exactly what -
I figured
it was a "dosemu issue" and just went back to the
last "good" djgpp
build I had handy...
> Plus, it's a lot
> bloatier than a DJGPP / UPX build would be (thanks to
DOS4GW). For the
> record, I dunno what compiler is used for these DOS
builds (old
> Watcom, perhaps?),
Watcom, I think - running on Linux. I don't think much
effort is made to
keep 'em small...
> but DOS32A.EXE is a < 30k .EXE, so it's highly
> preferable to DOS4GW. If you want swapping, just go
with CWSTUB
> (Causeway).
I always used cwsdpmi when I was doing real dos - seemed to
work okay.
I'm not familiar with these options.
> But as I said, the included DOS4GW didn't work for me
on Vista Home
> Premium, nor did using WDOSX's STUBIT on it (sadly).
Maybe I should
> test to see if it ran on my other cpus, but I figure,
why bother? I
> know DJGPP stuff works fine because I've tried my
previous builds on
> various real DOSes (plus Windows XP).
>
> Here's a summary of the cons for the current 0.99.05
DOS build:
>
> - doesn't work on Vista (except sorta w/ DOS32A, not
included):
> http://dos32a.narechk.net
> - bloated (not UPX'd)
> - no license text nor documentation included
>
> So, for your downloading pleasure, lemme rebuild it w/
DJGPP 2.04: 
>
> EDIT: Sorry for the dopey log below, I just felt like
being really
> detailed! )
>
> 11:11pm (already download, so I need to unpack and
build it)
> 11:13pm (okay, quick diff of NASMDOC.TXT explains -Z
change and
> implies -r is removed)
> 11:16pm (okay, I see the OpenWatcom 1.7 makefile, which
I should
> probaby test since I have it installed [ 1.7a, actually
] )
> 11:19pm (okay, dos and win32 build easily and
"-v" and "-hf" work
> okay; sorry, don't have the OS/2 part installed right
now, but the
> installer still has it)
> 11:22pm (each assembles a simple example okay, ndisasms
work too)
> 11:25pm (started "./configure" under DJGPP's
bash 2.05)
> 11:27pm (started "make" with default
options)
> 11:28pm (bombs out in FLOAT.C:760 and FLOAT.H:21
>
> "float.c:760: error: conflicting types for
'float_const'"
> "float.h:21: error: previous declaration of
'float_const' was here"
>
> 11:32pm (trying again w/ "int" instead of
"int32_t" and "-Os" w/o
"-pedantic")
> 11:33pm ("./nasm -v" and "./ndisasm
-v" report "0.99.05" correctly;
> ndisasm "-v" and "-h" plus
"nasm -hf" also work)
> 11:35pm (COPYING still says "59 Temple
Place", wrong snail mail address!)
> 11:42pm (putting everything together in its own dir,
renaming, UPXing .EXEs)
> 11:59pm (finished typing README.TXT and DESCRIPT.ION)
> 12:01am (okay, 16-bit DOS compile w/ tweaked makefile
won't work,
> NASMLIB.C's 64k limit to struct / union size is
reached, no huge
> surprise, just checking anyways)
> 12:05am (okay, time to .ZIP it all up, feel free to
tweak / remove whatever!)
>
> Okay, here it is (DJGPP/DOS, OpenWatcom DOS and Win32):
3 copies of
> NASM and NDISASM, plus COPYING and NASMDOC.TXT, feel
free to adjust
> accordingly (or whatever)! --> 648k .ZIP
>
> ht
tp://www.geocities.com/snoopimeanie/nasm9905.zip
Part of the delay (the other part's "lazy") is
because I've been
thinking about exactly what I want to do with all these. I
guess if the
Watcom build's not working everywhere (haven't had many
complaints), we
should replace 'em (new release?). For now, I'm thinking of
maybe
repackaging just the djgpp builds, and adding 'em alongside
what's
there(?). Renamed to "nasm.exe" and
"ndisasm.exe", I guess(?). That
sound okay?
Thanks for the binaries and the feedback!
Best,
Frank
------------------------------------------------------------
-------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and
a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Nasm-devel mailing list
Nasm-devel lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nasm-devel
|
Frank Kotler wrote:
>
>> Plus, it's a lot
>> bloatier than a DJGPP / UPX build would be (thanks
to DOS4GW). For the
>> record, I dunno what compiler is used for these DOS
builds (old
>> Watcom, perhaps?),
>
> Watcom, I think - running on Linux. I don't think much
effort is made to
> keep 'em small...
>
Right now, OpenWatcom 1.6 beta 2 for Linux -- since that's
the only
version I could find prepackaged.
I wouldn't object to switching to DJGPP if someone would be
willing to
roll rpm packages for a DJGPP-on-Linux cross-compiler. I
don't have
time to do it myself, and the build robot obviously runs on
Linux(*).
-hpa
(*) Well, there is always the option of running a virtual
machine under
KVM. That's rather complex to get to work in automation,
though.
------------------------------------------------------------
-------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and
a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Nasm-devel mailing list
Nasm-devellists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nasm-devel
|
