Group items tagged

'60s corporate comp[uting was IBM or single purpose appliances

PCs come out of a hobbyist branch, not a corporate branch---they were "solutions waiting for problems".

Watch and calculator are the consumer equivalent of the information appliance --- and at heart this has got to do with having someone to blame when it doesn't work.

PCs were always expected to run s/w written by authors other than the manufacturer ... but the PC, until the early 80s, was still considered a hobbyist toy.

Corporate computing moved to the mini-computer - still a dumb-terminal system that ran entreprise specific tasks and some general buisness functions.

Un-networked, hobbyist nature of PC made it particularly lax about security.

The PC started winning through its killer business apps - wordprocessor, spreadsheet and relational database, while the hobbyist uses led to the development of games.

Business adopted the PC for cost reasons, despite its basic unsuitability to group work; homes bought the PC for one thing and discovered many things they could do with it.

Microsoft genuinely wanted an open PC running many companies' code, not a s/w monopoly for itself.

A vast network of PCs ready to run new s/w offered a market for new s/w, a market that only grew as the Internet eliminated the cost of the transaction.

The core of Apple's and MS's business was in making attractive Operating Systems rather than in creating applications themselves, pace the MS antitrust cases of the 1990s that were precisely about such moves.

The backwater of hobbyist computing environemnt was essential to the experimentation & risk-taking that made the PC.

But the PC revolution went beyond the tinkerers to the non-technical user because of the ease with which PC h/w can be re-used for other purposes.

Here is the typical pattern: creative amateur chaos -> successful product -> market take-over -> risk to interests from original chaos --> temptation to lock it all down.

Actually, this product rarely encouraged developing new products from scratch. While it was possible to do with difficulty, by that skill level it would be much easier to use a protoboard and discrete components. Any level of structural control will limit the user to some degree from the freedom to create. This is actually a very good example, and a survey of the various products and their creative possibilities would make for an interesting way to explore this question of flexibility and generativity.

Abraham's new Lego-style electronic project kit is much more "controlled" than my old Radio Shack one -- and he will get less out of it, though it may be easier and more satisfying in some ways, he won't understand discrete components. Compare Ted Kilpin's e-mail about PIC controllers.

The fundamental threshold difference was the ease of transporting functionality with the floppy drive. This gave the Apple II a first opportunity to allow this kind of conspiring. Other platforms (Imsai, Altair, Compucolor, etc.) all had, or eventually supported, floppy drives, and there was some conspiracy, but the Apple's was the most plug-and-play of them all.

worm shows danger of generativity: ease of reprogramming could be dangerous.

Situation is worse today than in 1988 - network bigger and faster, processors more powerful and users less skilled.

Public reaction to worm: other Interent engineers thought it was a challenge; govt set up an enquiry and a program at Carnegie that still today is a clearing-house for info on network security, viruses, etc.

Fairy story ending to the fate of Morris, the first worm-maker, dot com millionaire, and now tenured professor at MIT.

The Internet Engineering Task Force, the de facto UNIX gods of the Internet, published a report ("The Helminthiasis of the Internet" :) describing the technology of the worm, and recommending some ethics training for new comers on the Net. ..

Naivety of Internet's response has to be understood in relation to logic of computer architectures.

Original ATT crack at 2600Hz : ATT, because it controlled a centralised network, could change the architecture to close-down the backdoor.

Compuserve similarly adopted post Cap'n Crunch rule: "do not let the paths that carry data also carry code."

the networks evolved slowly and with few surprises either good or bad. This made them both secure and sterile

Security = slow to change, sterile

1988 internet had no ways to control spread of morris worm

endpoint machines were disparate and open - no single way to secure them against morris attack

Insecurity of Internet - allowing code to be run on end-poitns, eg, - is critical to its creative potential

Post Morris, no one really offered to solve the problem, because there was no non-transforming solution ,,,

generative is good: great products, and also great opportunities for self-expression.

generative systems require trust; trust gets abused; if it is abused enough, people choose security over generativity.

Remarkable scalability of Internet, especially given new demographic of users in 1990s

hacker non-malevolence has slowed the issue of bad code from getting out of hand.

graffiti vs. drugs trade - where is the pay-off to bad code ... we can now see one

organised crime apparently controls 1 in 10 home pcs through botnets.

spam, ddos ($50k per day), gaming passwords, private infrmation -- all are biz model for malware

malware has become very sophisticated, with recovery essentially impossible against the worst infections

generativity is put at risk - either from death by 1000 cuts, or by watershed security moment.

Wipe the world's drives - the watershed scenario.

Badware subterfuge examples

Badware or not is subjective, in the eye of the installer - see VNC

Frustrated consumers, their machines getting infected and mis-performing, will go for appliances

consumers will increasingly to use the XBOX that works over the PC that crashes.

appliances clip the wings of invention

Vital function of PC is to keep the appliance honest - it is general purpose, and so can be reconfigured into the appliance; but this insurance policy is not WHY the PC is bought, it is unintended social benefit.

PCs that are shared are getting appliantised by various restrictions.

Regulation, not jsut technology, made it hard to make the telephone network useful for anything apart from telephoning.

ATT had control of network adn devices attached to it

Courts forced ATT to accept non-ATT devices on phone system

Court principle in 1940s: "as long as network is not harmed, 3rd parties can hook up to it".

The modem on the "open" ATT network was first generative opportunity for hobbyists and businesses

Firt online services were along IBM model but consumer-facing; one company supplied the lot.

In 10 years of business, between 1984 & 2004, Compuserve hardly changed its array of offerings.

Compuserve piggy-backed on ATT network, phy layer generativity, but did ot itself offer 3rd party acces for generativity.

Information service companies were charging by the minute - so why didn't they want to cultivate 3rd party apps?

Probably no one actually saw the commercial potential.

By early 90s, a small umber of networks offered slightly differentiated, closed, monolithic services.

The answer is probably deep in organisational behavior: mature markets encourage exploitation of niches, not radical exploration of new models.

Hobbyist BBS encouraged the opposite incentive of proprietary netowrk: bandwidth limitation meant that users had to be as quick as possible.

FIDOnet - the original p2p file sharing system - got over some of the bandwidth limitations.

FIDOnet was a great kludge, but had some real issues - like wrong numbers in the call list ... and irate humans at the end of them.

FIDOnet problems were exactly the problems that ATT worried about when it wanted to limit 3rd party appliances on the network - who was to blame for quality of service.

Internet offered flexibility of FIDOnet with robustness of proprietary network.

PC Internet analogy: Internet connects to proprietary networks and eventually replaces them

From the first, Internet is designed to pass information along diverse computer platforms by defining protocols.

Internet was low level: allow 2 nodes to exchange information ... and that is all; no prejudgement of what people want to do with that information.

Internet builders were academics who could devote time to the project and yet were doing it for the innate interest; $10m or so spent on Internet between 1969 and 1990 --- great return on investment, from a social point of view.

Environment of creation is critical in determining the nature of the system: Internet pioneers were trying to get the most out of existing infrastructure and diverse netwroks; contrast business logic of FedEx network.

The motto among Internet pioneers was, "We reject: kings, presidents, and voting. We believe in: rough consensus and running code;" they were building it for themselves, in a sort of Kantian environment of universalisability of the user.

1991, backbone privatisation context also allowed for decentralisation of network ownership and operation.

Internet -> Windows PC link built by a single academic hobbyist

1991, first opening of Internet to non-research uses; by mid 90s, proprietary networks were branding mainly as ISPs --- generative boom was gathering unbelievable speed.

Internet was built away from commercial logic, no trying to guess what the conusmer wanted; just like the hobbyist PC.

Basic trade-off: consistency, quality and reliability of service (closed systems) versus breadth of uses, trust of open systems.

Intenret design choices made sense given a sociology, not just an negineering environment.

Procrastination principle: 1984 Internet engineer academics formuilate argument that only universally useful features should be built into the network, with other specific work being done at the end-points.

Internet design assumed basic trustworthiness and competence.

Banks are built for robbers, airports for terrorists ... the Internet was built for colleagues.

AOL/CompServe had identity management built-in: wanted to know who you were for marketing purposes ... good for accountability, too.

On Internet, identity-management was left to specific service providers or apps to choose.

There's good and bad in internet's anonymity -- can't pursue bad behavers, but can't get the freedom fighters against oppressive governments either -- but an engineering decision has ended up havign huge social impacts.

Internet as "network of networks" loosely cobbled together entails the inability to develop quality of service, or bandwidth agreements, that apply between any 2 points.

"bit egalitarianism" is assumed in the network, and fixes have been found along the way (eg Akamai & network caching).

Crucible of development - collaborativ academis -- strangely made something that thrived beyond it and gave world this wildly generative environment.

Is the appliance and the proprietary network an evolutionary dead-end?

The generational pattern of generative technology: small-world success, big-world expansion through commercial forces, limitations of the model; limitations now may threaten the ethos that amde it so good, but limitations are real.

That is proven, with cell phones, iPods, calculators. More broadly, it can be said that nongenerative appliances coexist with generative devices. It's not so clear whether one subverts the other without a lot of force behind it.

I'm not convinced of the direction here -- the Brother or Smith-Corona appliance typewriter was supplanted not by a generative product, per se, but by Word Perfect, WordStar, Microsoft Word, and other competitors. No open-system revolution brought this about, only high-stakes competitive market cannibalism. It was not the open programmability of the PC and Mac that inaugurated the market, but the plug-and-play capability of the new product, coupled with the price point.

Better known as the "acoustic coupler," which gives the layman a better idea of its shape and function.

You mean mainframe, correct? Also worth a mention is CBBS and the other hobbyist networks that predate CompuServe, et al.

iPhone needs Apple permission before you can tinker; Apple can remote modify it; it is a controlled system.

ahah ! - sterile the iphone ?... not so... it has now many app. done free, as the next widgetbox web2 ways - and it is a very great inspiration too, an industrial asian has produced a similar product better...and of course as usual, protections attract hackers to modify or to use free: it's done...

The dark side of openness - viruses, spam, crashes - makes the closed philosophy of the iPhone attractive to consumers.

generative creativity = dangerous; closed consumption = "safe"

This is a very important tension to consider, and one that will take time to work through.

iPhone / Apple II story is representative of the Internet: it is moving form the open to the tethered, the generative to the controlled.

The original PC is a blank slate.

Apple II is a user object; iPhone is a consumer object (open / closed)

Boxed appliances are fine, but they endanger the NEXT ROUND of innovation --- innovation-types they themselves were dependent on in the past.

If the generative aspect is ursurped by closed appliances, the ecosystem dies as a productive source of further innovation

When there is separation of PC ownership and use, interests diverge: stability versus "wilderness"

As long as a PC can run code it is given, and especially with broadband connectivity, the dangers are there, and are serious.

like a natural ecosystem, open and generative platforms can provide habitat for dangerous pathogens (viruses, invaders, etc.)

These dangers add support for lockdown and interference by the surveiling centre --- with all the dangers this implies in authoritarian states.

lockdowns as monoculture... safe, but sterile

Lockdowns and box-appliances will destroy the creativity of IT, and to stop this happening we'll need new technologies, communities that sustain the right ethos and a sense of public purpose.

The everlasting battle will be between open, Web 2.0 systems and services and "bottled power" as exemplified by the iPhone.

- users (vs. consumers?) stay in the generative loop via Web 2.0 applications

free services and ...ware = battle for money may change to battle for communities influences and direct human exchanges