Wikipedia:Reference desk/Archives/Computing/2016 April 19

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

April 19

How did the 1968 NLS system accomplish video display?

The NLS (computer system) that was used in the 1968 The Mother of All Demos connected via leased telephone lines, to ARC's SDS 940 computer 48 km away. It seems hard to find any documentation on the NLS device itself. Leased telephone lines are a real poor signal path for video signals or anything that has hard latency demands with a simultaneous need for high bitrates. So it must been generated locally. Anyone knows how this was accomplished? keep in mind that in 1968 these things were really hard to accomplish. Bytesock (talk) 02:39, 19 April 2016 (UTC)

The Wikipedia article cites this Wired article, which says "we beamed two channels of video along two microwave links up to San Francisco, bouncing them off dishes above the airport [...] we rigged up a homemade modem – 2,400 baud – to get signals from my console in San Francisco back to SRI over a leased line." So I think the phone line was only for unidirectional keyboard and mouse data. I edited the Wikipedia article to more accurately match what the Wired article says. -- BenRG (talk) 03:09, 19 April 2016 (UTC)

The article uses another link. But thanks anyway for solving the missing piece. This means the "terminal" was in essence like a keyboard-mouse-video console you connect nowadays to KVMs to setup servers etc. Essentially no computing resources at all unless the keyboard and mouse MCU counts (did they even have an MCU?). Computing were expensive then, but it still boggles my mind that it took so long before the commercial aspect got up to speed. The 1970s were about being able to drive a monitor at all. And interfacing 1 kB memories at kHz speeds with paper loading or magnetic tape storage. Btw, how the SDS 940 really accomplished the graphics output is also kind of "in the dark". It seems to resemble the Amiga copper chip somewhat from the description. And is the modem 2400 baud or bit/s? Bytesock (talk) 04:02, 19 April 2016 (UTC)

I think back then baud was synonymous with bit/s, see e.g. Baud#Relationship_to_gross_bit_rate. SemanticMantis (talk) 14:59, 19 April 2016 (UTC)

Yeah, it's almost impossible to deduce what the real speed would have been. But now I have also found one source that claim 1200 "baud" and another that claims 2400. Bytesock (talk) 15:49, 19 April 2016 (UTC)

this is also how modal editors like vi (as opposed to things like EDIT.COM and with a point and click interface) came to be. Drawing windows, menus etc, even in pseudo-graphics, much less pixel graphics, is only practical if you have direct access to the graphic card's frame buffer, but not if your display unit is connected via a slow serial line and you need elaborated control codes to do anything beside line breaks. BTW, this need not be limited to text and can be generalized to vector grpahics. This is the idea behind Unix'/Linux' X Window and Windows' GDI, which use their internal vector languages and furthermore NeWS, a windowing system which employed PostScript, today normally used in laser printers (!). And yes, terminals do have computing resources (the VT100 was built around the 8080, for example) but don't expose them, except accidentally (Datapoint 2200). The terminals thing is also why there is no distinction in Linux between graphical and "console" programs but there is one in Windows and this is decided at compile-time (a flag in the executable file tells Windows whether to open a console window for the program). It's also why Linux has a "terminal emulator" rather than a "console" to start with. What does it emulate? Why, the VT100 (or something very much like it.) Asmrulz (talk) 23:52, 19 April 2016 (UTC)

I don't think the demo had what we'd really call a bitmapped display. As Framebuffer#History notes, even research machines weren't really generating images from RAM until a year or two later. As I understand it, the demo is effectively text mode, with a mouse cursor. A character mode display of that era would be a mostly hardware based character generator system, which takes character codes (maybe ascii) from a RAM page, uses what's effectively a hardware lookup into a character ROM, and strobes that out to the analog raster. The mouse cursor is effectively a hardware sprite, where again the image is generated by hardware working off the raster sync signals, with the software just setting up the image and a few control registers - this achievable in affordable consumer grade technology by 1977's Television Interface Adaptor, so it's reasonable (but, don't get me wrong, still very impressive) for Englebart's team to a built a simple custom one in 68. -- Finlay McWalter··–·Talk 00:01, 20 April 2016 (UTC)

Point-to-point transmission of analog TV signals is effectively outside broadcasting, which the BBC was able to do over a radio connection by 1938. Their history of the efforts to live OB the 1953 coronation talks about them using a custom balanced and co-ax lines but also "ordinary telephone cables with video repeaters" - although obviously the BBC had lots of pull with the GPO (the telephone company), so surely those lines were manually patched at the exchange, and not run through a switchboard or electromechanical switch. It's easy to see how SRI, without comparable influence over Bell System, would opt for their RF solution instead. Their compositing of the camera video signal with their computer video signal would again be done electronically (i.e. in analog hardware). -- Finlay McWalter··–·Talk 00:19, 20 April 2016 (UTC)

Criminal modus operandi database ?

There are fingerprint and DNA databases, but how about one where you can search for a criminal with a similar MO, but in a different jurisdiction, perhaps even another nation. Does such a DB exist ?

I can see that there might be difficulty in categorizing and searching items, like if one profile said "victims were bound", another said "bindings are used", and another referred to "ligatures" (not to mention foreign language versions). So, it's not a trivial task. StuRat (talk) 18:57, 19 April 2016 (UTC)

"The database houses modus operandi, signature aspects, crime scene descriptions/photos, victim and suspect details, and other pertinent information." [1]. By "you" I assume you mean "a person" and not you, StuRat. You, StuRat, cannot search that FBI database unless you have access via a law enforcement agency. SemanticMantis (talk) 20:24, 19 April 2016 (UTC)

Sounds like a step in the right direction, but:

1) It makes no mention of being accessible for additions or searches in other nations.

2) It made no mention of it being accessible by attorneys. This would seem to be important if you are a defense attorney and can find a dozen other criminals in the area with the same MO as the one for which your client is charged.

3) It made no mention of being accessible by crime victims or the private investigators they hire to help to solve crimes the police can not.

4) It listed "150,000 open and closed violent crime investigations". That's only a tiny portion of the violent crimes in the US, isn't it ?

5) Nonviolent crimes might also be useful, say if an arsonist who started by setting dumpster fires has moved on to occupied structures.

6) Also, I'd like to know how the searches get past the problem of different terminology I mentioned in my first post. StuRat (talk) 20:36, 19 April 2016 (UTC)

In a UK context, I would assume that the Crimint system has some of this functionality, but I'm not about to draw attention to myself by making further enquiries! See also the UK-specific articles Police intelligence and Police National Computer. {The poster formerly known as 87.81.230.195} 185.74.232.130 (talk) 13:51, 20 April 2016 (UTC)

Virtual reality

I was watching BBC today, they were displaying something about Shakespeare play. The interesting part was it was getting done in 4D; a camera records the reality ‘’live’’ and then displays it in T.V. Of course, you have to detail the artwork of the virtual reality from beforehand, for example, create the virtual character fully in order to pupate the real person, and so on. Does anybody know what the software is called? -- Apostle (talk) 19:10, 19 April 2016 (UTC)

Is this [2] what you saw, about a game called Play the Knave? If so, it was developed at the UC Davis Modlab [3], and the official website for this VR Shakespeare is here [4]. It seems to me that much of the software would have been custom made, but also there is probably a lot of work done by open source libraries. Here [5] is a list of some of the research publications associated with the project, and that might tell you more. Finally, if you use the "contact us" button on the site, you might get a better description of the tools involved. SemanticMantis (talk) 20:23, 19 April 2016 (UTC)

Yes, its the right one. Thanks -- Apostle (talk) 18:27, 20 April 2016 (UTC)