> Okay, let's talk about Nasm!
As some of you might have already noticed, during the past two days I've
gone through all of NASM's outstanding bug reports, support issues, and
feature requests. I have taken the liberty to close anything that looked
resolved or like a duplicate, and I have modified the priority of those
items for which I have working code. If I have been to aggressive, please
feel free to re-open items, and/or drop me an e-mail.
> There are several other contributions - to preproc.c and others - in the
> same shape. This may be a worse problem than the rudderless list, and
> maybe something we can do something about.
The real problem was that the preprocessor was getting drowned by a list
of fixes, workarounds, hacks, proposed changes, etc. -- but that nobody
took the time to take a glimpse at the bigger picture. Let alone analyze
it for flaws, shortfalls, missing functionality, etc.
By now I have taken that look, and performed that analysis. The result is
a preprocessor that is way more powerful, but at the same time has fewer
bugs and quirks. Overall its feature set feels rich and balanced, whereas
its code is based on NASM 0.98, not a complete re-write.
In addition I have managed to improve performance, and add 64-bit support.
That said, the 64-bit support also marks the major showstopper for why my
code can't just be included in NASM 0.98.x. I don't yet know whether I'll
get the opportunity to release NASM64 as a whole, because that is not my
decision to make. However, I may be able to come up with an interim step:
a standalone preprocessor based on my code.
In the meantime I'd like to offer the attached file, which gives a short
overview about my improved preprocessor, with references to those bugs or
feature requests it addresses.
Let me know if you have questions.
Now, as for the immediate future, my recommendation is to have an IEEE754
expert review the proposed float.c/float.h code, because that does mark an
area where existing versions of NASM emit bogus data. In addition a number
of small issus should be fixed here and there, and then we should go for
the next release. I don't care about its version number.
My general feeling is that NASM 0.98.x has pretty much settled -- radical
development should, and most likely will, occur elsewhere.
Which IMO is good.
I haven't had a chance yet to study the proposed ndisasm.c changes.
As for #822614 -- see my latest update to it.
Last but not least, I'm planning to look at the proposed EXTERNDEF soon.
Do you Yahoo!?
Yahoo! Finance: Get your refund fast by filing online.