[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gm2] Rant against ISO Libraries
From: |
Gaius Mulley |
Subject: |
Re: [Gm2] Rant against ISO Libraries |
Date: |
08 Feb 2005 09:28:48 +0000 |
User-agent: |
Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 |
john o goyo <address@hidden> writes:
> Greetings:
>
> As an exercise in using the ISO library, I wrote a wee hex-dumper. It
> works with the ISO-compliant compilers at my disposal: GPM on Linux
> and Solaris, and XDS on Linux. I thought that I would test it with
> GM2 by defining those bits of the ISO library needed.
sounds interesting..
> Herein lay my tale of woe.
> <rant>
> Trying to define the ISO parts for a bit of simple I/O turned into a
> headache ("The gasman cometh", if you recogize the reference).
>
> Let me compare this with the saga of WG14. They chose to break as
> little existing code as possible. They certainly made controversial
> choices by deciding on the Unix library but at least they did not
> break a lot of code. In contrast, WG13 chose to create completely new
> libraries so as to ensure that all existing code broke.
> </rant>
yes the libraries in Modula-2 are its Achilles heel. No sockets.
I'd have preferred a clean interface to the Posix 1003 (?) standard.
I'm not sure which came first though. However I like FIO and
DynamicStrings (although neither are PIM or ISO :-)
I also wonder how complex types found their way into the core
language. Not that this is a problem for gm2 (or wont be) as the back
end gcc supports this nicely. I would have thought that complex types
would be better served by a hidden type and functional programming -
and introducing the __INLINE__ keyword to make functions very
inexpensive. It would keep the language smaller. Does anyone use
complex types in M2 ?
regards,
Gaius