Some things ya wish you never said

Its big failure was that it wasn't compatible with anything. DEC vehemently fought the revolution, and paid the ultimate price.

You just haven't let the story go far enough. ;-)

I think it's been there for a while. Some just haven't noticed. ;-)

huh?

That would be amazing considering the year.

Myself, I would see no reason other than cost back in 1969 working at dec. I worked in displays, and troubleshot a 4 high rack, which was a asynchronous serial line multiplexer. They were making computer displays. I was in the mist. It was TTL, and my computer, a PDP 8/I.

Guy working on the display, kept going in the bathroom to wash boards. Only way to get them to work.

Greg

Yes, I think the Surface Pro is almost there (maybe one or two more iterations). The pen input is really nice.

Best regards, Spehro Pefhany

True. The same people want the answers, without being able to frame the questions. There are also those that consider computing a fashion statement.

I see it somewhat differently. The typical pad and smartphone are primarily display devices. They're great at watching videos, reading email, reading books, viewing some web pages. They're kinda mediocre at searching, voice command, and number entry. They're nearly unusable for any input more complicated than menu picking. If one tries to turn a pad computer into a full featured computing device that can function in place of a desktop, the peripherals required will rapidly turn it into a tangle of wires and boxes. Pads and smartphones are for viewing someone else's creation. However, you want to create content, it's best done on a more traditional desktop or laptop.

These same users are also not into maintenance. One that I know all too well has decided that she doesn't need to do anything to protect her work, files, email, etc because if anything goes wrong, I will somehow magically resurrect them from the dead. No amount of logic will convince her to backup anything, even to iCloud. I'm waiting for the inevitable disaster.

Spice and schematic capture on an iPad:

In the 70's an "internet" (lowercase i) was the term applied to an "internetwork" -- a group of (inter)connected computers. What we would now call "a network" (or even a private network or an intranet).

..on the "cloud".

No, the original development of TCP/IP was for "internetworking", i.e. interconnecting different types of network, what we now call a WAN.

I've used it to connect Arpanent, Bitnet, VNET, and IBM CM on a token ring.

Cheers

Phil Hobbs

I said nothing about TCP or IP: An internet is a group of interconnected ("INTERNETworked") computers. ISTM that we colloquially call these things "networks" (or LANs). A WAN exists on a larger geographic scale. It is also *an* internet -- and need not use IP (e.g., ATM).

All of the devices (computers) on *an* internet must share a common communications protocol -- else they can't talk to each other. A protocol cares not (small lie) what the underlying hardware mechanism might be but, rather, how the "content" is interpreted.

The Internet (uppercase I) is *an* internet that happens to use The Internet Protocol as this communications protocol. The TCP/IP model elides a few of the layers in the OSI model -- IP representing the "Internet" layer (corresponding approximately to the "Network" OSI layer). TCP and UDP ride, side-by-side, atop this in the "Transport" layer (corresponding to the OSI layer of the same name).

As with many protocols, you can choose to run IP over many different hardware media *and* atop other underlying protocols. This allows things like Ethernet-over-ATM (LANE). Gateways and bridges are your friends. :> Just like you can have a CAN-to-IP gateway! Or IP-over-"EIA232" (SLIP).

"It's all just software! (and a wee bit of hardware)"

If you don't personally own copies of *all* of Stevens' texts, it's a worthwhile investment. I bought them when I wrote my second TCP/IP stack ("Sheesh! There has got to be an easier way to do this...") -- they weren't in print when I wrote my first. A shame he won't be writing any more :-/

Stevens : networking :: Knuth : algorithms

Comer also has some texts that deal with these issues with a lighter touch if you aren't interested in all the dark corners and efficiency of implementation.

This is what we have been using the last 20 years (called the World Wide Web). The home "computers" are just visualization engines a bit more versatile than for example a VT100 or 3270 terminals. The web servers are just reborn centralized computing.

Except for hard core gaming, look what current home or mobile "computers" are used for, i.e. mostly running some www browsers. So these devices are just glorified IBM 3270 or Minitel terminals :-).

The real computers are the centralized servers feeding info into these private terminals over the internet.

IMHO, Olsen was not that far off.

I still think it is much easier to maintain a centralized system with

In a PC network, there is always the issues such as is the PC powered up, connected to the network, is it of correct OS version etc. etc.

That's certainly true. I like having control over my stuff, though, especially the backups of my life's work and my project plans. Plus when you write most of your own software, cloud computing is a crock--every time they change the APIs your code breaks.

Cheers

Phil Hobbs

That seems to be a bit US centric view.

At least in Europe, telecom companies tried to use the X.25 (HDLC) on transportation layer connections and X.400 for e-mail.

Typically accessing some TCP/IP e-mail site, a reference was made to RFC-822.

Agreed. Though how you define "dumb" is entirely subjective! :> E.g., I work with X terminals, here. Each of them would qualify as "the equivalent of a large computing center of the 60's" despite the fact that all they do is serve up a "user interface".

Yup. Has the "user" done something to alter the device's configuration? Disconnected it from the network (to work standalone)? etc.

Note that you (your IT dept) can offer as many or as few applications as it wants with the centralized model. You just have to decide *who* makes those decisions.

A big advantage of centralized model is that your "files/environment" come *with* you instead of being stranded *at* "your PC". From the administrator's perspective, it makes it a lot easier to ensure everyone is using the latest version of an application (unless they don't want to!), backups are easier (no need to pull all that data

The big downside of centralized is that it opens the door for the BofH syndrome. :<

Probably. :>

A network adopts one (or more) protocols to allow the devices on it to communicate with each other.

Connecting networks together requires a (possibly different) protocol (or more) to allow the *networks* and their constituent devices to communicate with each other.

E.g., all the devices on an XNS network "speak" XNS. All the devices on a Modbus network "speak" modbus.

There's nothing magical about the Internet Protocol (IP) that makes it special -- OTHER THAN ITS UBIQUITY! (owing in large part to free/open implementations early on that could easily be copied instead of requiring a reinvention of a slightly different shaped "wheel")

Remember "bang" addresses for mail??

What aspect(s) do you consider "really nice" and which still lacking? I've been evaluating different pen implementations and haven't found one that seemed to really "work" (in ALL the ways you expect a pointing device to "work").

E.g., picking/clicking icons and menu items vs. freehand "drawing" or selection vs. positioning a "cursor" (text or graphic) within an object vs. "character/gesture recognition", etc.

I haven't decided which usage factors enhance and hinder the experience. E.g., parallax plays a big role for fine positioning unless you hold the pen at an oblique angle to the writing surface (so it and your fingers are out of the way). How well "contact" is debounced (think hysteresis) also seems to contribute to usability.

I have a tablet PC that I am reasonably happy with for "drawing" use -- esp informal drawings. But, using the pen for fine positioning is tedious. And, *any* use in which the drawing surface (i.e., tablet) isn't rigidly supported is miserable!

What's US-centric about it? TCP/IP was developed in the US under a DARPA program, primarily by Cerf and Kahn, and later standardized (RFC-714).

We were talking about the origin of the term "internet", which comes from "internetworking", i.e. connecting together heterogeneous networks. See e.g.

Cheers

Phil Hobbs

I don't know about yours but mine is a bit better than 80x24 green. ;-)

Really? You think your thermostat or your microwave oven timer runs in the cloud? ;-)

The picture may have made a point a couple of decades ago but tThat's changing with bluetooth and WiFi. Cords are becoming a thing of the past (except, of course, the power strips full of wallwarts). The "Surface's" keyboard, while not a real keyboard, goes a long way to making a "tablet" into a PC.

I'm debating whether it's worth it to retrofit my house with Cat5. The house was built six years ago but they forgot to add Inet (or Fannie ripped it out when they "rebuilt" it). About now it's doubtful that it's useful.

Some learn from others, some don't even learn from their own mistakes.

Ick. Just the thought makes me ill.