Talk:SGI Dogfight

External Links - Broken link to JongWon Kim's lecture PDF

The link to JongWon Kim's DIC 1623 lecture is broken, and there is no archived version on either the Wayback Machine or archive.today. I can't find any more up-to-date content on the Gwangju Institute's website that might correspond to a more recent version of that lecture, but I don't really know what I'm looking for.

Does anyone know what was in those lecture notes and why it was relevant to the SGI flight simulator?

AstridRedfern (talk) 17:00, 7 May 2022 (UTC)

The original URL for this was: http://netmedia.gist.ac.kr/~dic1623/lectures/lec02-origin.pdf AstridRedfern (talk) 21:49, 17 August 2022 (UTC)

Since the link is dead, and since I couldn't find any archived versions, I've reluctantly decided that I'll have to delete it from External Links. AstridRedfern (talk) 21:49, 17 August 2022 (UTC)

I've emailed Professor Kim and asked him to send me a copy of these notes (assuming he still has them - 2005 was a long time ago!) AstridRedfern (talk) 14:22, 19 September 2022 (UTC)

The atc/Air Traffic Control program

I can't find much information on this.

First of all, it was supposedly "added" to flight, but I can't find any evidence that it was a later addition. For all I know it could have been part of the software from the very start!

The two mailing list posts below are the only information I have:

> Good morning to all you pilots out there. About a year and a half ago I received a TCP/IP version of dog that required a routine to be running on one of the network machines (called ATC) and each machine on the net had to reference that file in order to play dog. This version of dog tended to crap out at various times and could handle at best 4-5 players. SO, last Xmas, I loaded up the XNS kernel and XNS dog and had the pilots boot the machines in XNS to play during the 'holiday season'. Before I reload the XNS dog again for this holiday season, is there a better functioning TCP version than the one I have? I can't tell from the mailgoing around about dog if its the same one or not, but I have a feeling it is.

and

> We have a 3130 with GL2-W3.6; dog, shadow, etc. don't work. The manual says that in 3.6 they are TCP/IP, but I don't know if that is true. I made the above changes, but they didn't help. I have been told that a server program, atc (air taffic controler), must be run, but we don't have it. — Preceding unsigned comment added by AstridRedfern (talk • contribs) 19:29, 7 May 2022 (UTC)

Looks like I forgot to sign the above. Apologies. AstridRedfern (talk) 19:58, 7 May 2022 (UTC)

Those two mailing list posts came from https://web.archive.org/web/20070717025813/http://ftp.arl.army.mil/ftp/info-iris/info-iris.txt.05 - it's already in External Links, btw AstridRedfern (talk) 09:42, 23 July 2022 (UTC)

I've decided to cite the two posts anyway, since they are at least evidence that atc actually exists! AstridRedfern (talk) 15:21, 20 August 2022 (UTC)

The list of available planes

The list states that the set of available planes in the initial version was as follows:

Cessna 150

P-38 Lightning

Boeing 747

F-16

F-15 (F-14 model; wings spread)

F-14D (wings laid on)

https://github.com/lkesteloot/alice/blob/master/alice4/software/flight/main.c is a source file from a recent port of flight to some ARM-based custom hardware. In this file, all of these planes except the F-14 are available, and no others. This suggests that the port might be of a fairly old version, and that the F-14 was one of the planes added in later versions.

However, it could also be that the F-14 was deleted from the port for some reason, so I can't be sure of this.

AstridRedfern (talk) 19:57, 7 May 2022 (UTC)

I've just remembered that one of the sources I cited mentioned the F-14 as only being available in the GT/GTX/VGX version of the software. I don't know much about Silicon Graphics hardware, but if that version was not the original version, I'll definitely amend that list. AstridRedfern (talk) 20:05, 7 May 2022 (UTC)

The date when the first version was written

"Flight was written in the summer of 1983[citation needed] for the SGI IRIS series of computer workstations."

The book "Networked Graphics" (https://www.google.co.uk/books/edition/Networked_Graphics/76C_quJqVXcC?hl=en&gbpv=1&printsec=frontcover) (Anthony Steed, 2009, published by Morgan Kaufman) does state that it was "originally built in 1983." But it says nothing about the summer, and it's possible (going by the publication date) that it used this article as its source. It doesn't cite its sources inline, but it is also possible that the source was one of the publications cited at the end of Chapter 1.

AstridRedfern (talk) 13:46, 22 July 2022 (UTC)

The same is true of the following statement:

"In 1984, networking capabilities began to be added. Initially, two stations were connected by serial cables. This allowed about 7 frames per second. By SIGGRAPH 1984, XNS support was added, allowing play over an ethernet.[citation needed]"

Even if the book did not use the article as a source, though, it does differ slightly. It mentions XNS support, but does not mention it being "By SIGGRAPH 1984". Nor does it mention the frame rate. So it might (or might not) be a source for part of that statement. But not all.

I'm considering adding a Google Books link to the "External Links" section. AstridRedfern (talk) 13:51, 22 July 2022 (UTC)

Another:

"Probably in early 1985, dog was created" (article)

versus

"The version called dog appeared in 1985 and introduced combat." (book)

And:

"In 1986, UDP broadcast protocol capability was added (using port 5130). Information was transmitted via broadcast packets and at frame rate, meaning that the program made intensive use of network resources and even a small number of players was capable of saturating an Ethernet. So while it was probably the first game to use the Internet Protocol Suite, the game could not pass through a router, and thus could not be played across the Internet itself.[citation needed]" (article)

versus

"This version was an early demonstrator of the use of the User Datagram Protocol (UDP) (see Section 2.2), but because it sent information at the graphics frame rate, it caused a lot of network traffic. The initial version worked over the Ethernet using the TCP/IP protocol suite, but it didn’t actually work over a router and thus didn’t support internetworking." (book) AstridRedfern (talk) 13:56, 22 July 2022 (UTC)

Sorry, got the author information wrong - it was Anthony Steed and Manuel Oliveira. AstridRedfern (talk) 13:58, 22 July 2022 (UTC)

Manpages

I've found another IRIX manpage link at https://nixdoc.net/man-pages/IRIX/man6/flight.6.html - but I can't tell which version of IRIX it's the manpage for. AstridRedfern (talk) 21:46, 17 August 2022 (UTC)

Developers

The "Developers" section states that the manpage for the IRIX 6.5 version of the program lists the following developers and credits them for the following:

* Gary Tarolli (original version)

(The manpage does indeed say this.)

* Network communications David "Ciemo" Ciemiewicz (UDP Broadcast) and Andrew Cherenson (IP Multicast).

The manpage talks about UDP multicast. It mentions broadcast being present in older versions, but says nothing about IP multicast. It does list these two developers as responsible for network communications, but does not say anything about which protocol/etc. each individual dev contributed.

* Rob Mace (involved by 1988, listed as responsible for IRIX 6.5 version, as well as the geometry for F-14, F-18, and 727 aircraft, and the new instrument panel)

The article's phrasing should reflect the fact that other developers were also involved in the new instrument panel, and the F-18 and 727 geometry. Their names are given below, but the article looks as though it is contradicting itself and assigning sole credit to different people for the same thing in all three cases.

* Barry Brouillette ("contributor", geometry for F-16 and P-38)

* Marshall Levine ("contributor", new instrument panel)

(Again, wording needs to make it clearer that Rob Mace also worked on the instrument panel. Also, the manpage doesn't explain why these two are described as "contributors" when the others aren't.)

* Thad Beier (F-18 geometry)

* Marc Ondrechen (727 geometry)

(Again, we should improve the article's phrasing, to make it clearer that Rob Mace was also involved in both these things. Otherwise correct.)

* Sound effects by Chris Perry and Chris Schoeneman

(The manpage does indeed credit the two Chrises for sound effects)

AstridRedfern (talk) 15:58, 20 August 2022 (UTC)

There was also nothing in the manpage about Mace being "involved by 1988". I haven't found a source for that date, but I have found one that shows he was the primary coder on this by 1989. I will cite that. AstridRedfern (talk) 16:44, 20 August 2022 (UTC)