Re: Segmentation in real mode
From: Betov (betov_at_free.fr)
Date: 03/23/05
- Next message: Betov: "Re: AoA is now available in Chinese"
- Previous message: Annie: "Re: International Peace March (19th / 20th March 2005)"
- In reply to: Herbert Kleebauer: "Re: Segmentation in real mode"
- Next in thread: Ro : "Re: Segmentation in real mode"
- Reply: Ro : "Re: Segmentation in real mode"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: 23 Mar 2005 11:23:39 GMT
Herbert Kleebauer <klee@unibwm.de> écrivait news:42413717.7E72EEC4
@unibwm.de:
> That depends on your history. I started with FORTRAN IV and
> punchcards. There where 72 columns (the first five for labels,
> one for continuation and the rest for the code). When I later
> saw C code, my feeling was maybe similar to that of Betov,
> but if you try it, you get very fast used to it and will like
> it. But for assembler code I still prefer the simple form:
> one instruction per line, an optional label, a TAB, the instruction
> (all starting in the same column) and an optional comment.
> But this is just a matter of personal preference.
I am absolutely convinced that some details are _not_
a matter of personal preferences. For example, having
a Comment after an Instruction is an evident error.
1) This is simply _un-maintainable_.
2) There is no reason why an Asmer would want to comment
an Instruction of the Language he uses. We do not need
Infos about what the Instructions are doing, but about
what the Instructions Chunks (the _Program) are (is)
doing.
So, the only possible room for a Comment is above a Chunk
of Instructions, in order to recall, several week after
the first write, what Code Flow is doing.
I sometimes also put some Comment(s) after an Instructions
Chunk, in order to explain the result, but, in all cases,
_always_, stand alone Comments.
All of this depends, indead, on the personal history. But,
not changing, after the personal history has pushes to bad
practices is nothing but intellectual laziness. I also
begun writing the old traditional fashion way, but, all i
got convinced with during the developements progresses, was
simply, that this way was utterly wrong, and that the new
writing style i introduced with Multi-Instructions Lines,
isolated Comments, Multi-Lines Instructions, HLL Indents
combined to A86-Like True Local Labels, Macros for HLLs
Constructs, and so on,... was the definitve way to go,
dispiting the resistances:
The old style is simply not appropriated for writing real
Life Applications in full Assembly.
You, herbert, are in a difficult position, where, on one
hand, you say that Asm is not for writing real life Apps,
and the the right Tool for the job, is C, whereas, on the
other hand, you do not make use of the Asm writing style
that would prove the reverse to yourself.
Betov.
- Next message: Betov: "Re: AoA is now available in Chinese"
- Previous message: Annie: "Re: International Peace March (19th / 20th March 2005)"
- In reply to: Herbert Kleebauer: "Re: Segmentation in real mode"
- Next in thread: Ro : "Re: Segmentation in real mode"
- Reply: Ro : "Re: Segmentation in real mode"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|