scotartt on 16 Jan 2001 17:07:30 -0000


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: <nettime> Code as (literary) text


Soeren, before this rant below, a snippet of info for you; source as
literature, see e.g. perl poetry  and perl itself might be a very good
source code language to start with as it has linguistic origins. also lisp.

but, onward;

From: Eric Berthelette <eric.berthelette@colorado.edu>

> Soeren, perhaps this takes your point too far afield, but I wonder if we
> could usefully take the literary metaphor a step further. You distinguish
> between a literary approach and (I think) a commodity production approach
to
> software development.  Presumably, this involves significant changes in
the
> way software is made and how it may be used.  For example, open source is

Just to digress completely and first talk about this purely from the
software engineer's perspective, of how code is designed and used. i have
to say that first the *purpose* of the code would have to change. business
now dominates the idea and economics of engineering and mostly probably
always has. modern software programming and design systems are a long way
from the raw, made with emacs, with a dog-eared copy of the K&R book, a
bunch of text files with odd names, and a Makefile or two all compiled into
a functional bit-blipping machine controlled from a Bourne Shell command
line environment on a Unix operating system. Sure all true programmers love
open source and the True Path of the Unixen, but for most of us, i think,
the real deal in the daily technological treadmill tends to be IBM,
Microsoft or Oracle and their innumerable lesser daemons. Two of these
unholy trinity (if you like) support linux as a platform reasonably well
and it's become an 'enabler' for their e-commerce strategies.

But as has been said before, IBM and those guys have accepted the open
source / linux and they are now busying layering their 'patterns for
e-business', DCOM, EJB, ASP, WAP, JNDI, etc, architecture over the top.

open source needs to get beyond the program, not just extend the idea of
what the program is. if you look at 'open source' its only at one or a few
layer(s) of the overall system design. open source needs to or could be
repeated at the level of *design* in each of the following;

open source mind-body interface (cartesians beware!)
open source information sources
open source information structures
open source artifical/intelligent/action/proxy/agents
open source object models
open source object code
open source object containment and runtime infrastructure
open source operating system and networking services
open source firmware
open source hardware systems
open source chip design + mfg
...
open source atomic structural integration (nanotechnology)

we need open source architectures for the long term storage and
classification of email lists! we need open source indexing and
categorisation **designs**. databases can be copyrighted; formats can be
patented.

open source should extend to information structures, data and
transactional, for exchange between software agents operating on behalf of
humans.

especially important is that security, authorisation and identification
have to be designed in from the starts based on open-source designs and
methodologies.

***even those design methodologies themselves should, or could, be
"open-source"***

however to adequately model or structure these types of problems are
requiring less and less reliance on ideas of 'code' as in execution
instructions or rules and logics but also non-linguistically based systems
like UML (software design representation) or data representatioal
structures like XML which is only partly 'linguistic' in function. just to
program Java now with an advanced tool like IBMs VisualAge for Java is to
see where its heading. There, despite it being a code-driven compiled
system using Java, there's a lot that the tool allows you to achieve
precisely *because* you leave the idea of the flat-text-file "source code"
behind, while still retaining the luxury of working in that medium when you
want/have to.

But, fuck, I like the price, and the attitude of Linux. and I can *run*
this on Linux.

Linking all this back to language?? How about the French anthropologist
Andre LeRoi-Gourhan and 'Parole et Geste' (speech and gesture) and the idea
that language, gestural impulse, and tool making are closely interrleated
anatomically and neurophysiologically. Each develops alongside and enforces
feedback upon the other. storytelling, artworks, 'engineering' if you like

> praxis, and expression.  If literature and software cannot transcend
social
> relations, perhaps we should cease to think of the open source movement
as

I would argue there's never a possibility for that. 'social relations' will
always operate (and be operated on by) objects of human activity. one
affects the other by neccessity. one needs reforming, so does the other.

regards
scot

#  distributed via <nettime>: no commercial use without permission
#  <nettime> is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo@bbs.thing.net and "info nettime-l" in the msg body
#  archive: http://www.nettime.org contact: nettime@bbs.thing.net