Date: Sun, 10 Jul 94 04:30:07 PDT From: Advanced Amateur Radio Networking Group Errors-To: TCP-Group-Errors@UCSD.Edu Reply-To: TCP-Group@UCSD.Edu Precedence: Bulk Subject: TCP-Group Digest V94 #144 To: tcp-group-digest TCP-Group Digest Sun, 10 Jul 94 Volume 94 : Issue 144 Today's Topics: DOS (5 msgs) DOS and Packet drivers Need NOS for Windows TCP/IP Question unsubscribe Send Replies or notes for publication to: . Subscription requests to . Problems you can't solve otherwise to brian@ucsd.edu. Archives of past issues of the TCP-Group Digest are available (by FTP only) from UCSD.Edu in directory "mailarchives". We trust that readers are intelligent enough to realize that all text herein consists of personal comments and does not represent the official policies or positions of any party. Your mileage may vary. So there. ---------------------------------------------------------------------- Date: Sat, 9 Jul 1994 07:28:52 -0600 (MDT) From: Klarsen Subject: DOS To: TCP digest I have been waiting almost 10 years for the general ham world to have a real DOS computer so I can get him to run nos. We are about there now. And what happens? All the serious thinkers are off with plans to use LINUX as the operating system of the future. I do admit the advantages of linux, it's free and it works. In defense of dos there are millions of homes and offices that have one or a lot of dos computers, many with the windows ver 3.1 running and several good programs for work related purposes and a few for fun. I have talked with a ham in San Diego who can now get used dos computers of the 386 class for under $100. It would then seem logical that hams will in the next few years be able to upgrade from their Commador to a 386 dos computer. From here it seems we need to talk about users running faster equipment. It is not unreasonable to suggest a CARD type modem that will give the average ham a 9600 baud switch that nos can attach to in a clean way. It should have some non-oscillscope way to adjust the audio levels and should come with serious documentation. The card will use the computers power supply and clocks and such so that the card can be produced cheap. It should be cheaper than a TNC-2 is today ($125.00). We could talk with G3RUH and see if he is interested and if the thing could be built for the above figure. I see the local network guys putting up x1j nodes on 2 meters and producing list's of radio's that will support 9600 baud (true FM transmitters). I have always wondered if the Alinco dj580 could be used for a 9600 baud system. It might make a good bit wise full duplex regenerator node.... Well the card is important because the new packet ham doesn't want to buy a tnc and then buy a modem for that. Better he buy 1 card and he is set to join in at 9600 baud. Wonder if DRSI would like to make this card? karl k5di ------------------------------ Date: Sat, 09 Jul 1994 11:21:53 -0400 From: "Brandon S. Allbery" Subject: DOS To: tcp-group@ucsd.edu In your message of Sat, 09 Jul 1994 07:28:52 MDT, you write: +--------------- | now. And what happens? All the serious thinkers are off with plans to use | LINUX as the operating system of the future. I do admit the advantages of +------------->8 Huh? Johan's targeting MS-Windows and NT; others are working on OS/2; and, last I heard, Phil was using 386BSD. Alan Cox and I are the primary Linux folks, and we're following different paths (Alan with kernel AX.25, my own being a JNOS variant which will allow use as a "traditional" NOS or as a simple AX.25 server to allow interfacing between AX.25 activity and Linux network clients and servers). None of *us* is arguing that we have the One True Way. [Many of the JNOS folks are looking at Linux, though, mainly because my own work makes a nice almost-drop-in replacement for their existing configurations.] It is, however, missing the point of a 32-bit multitasking environment to use it that way; I hope to be able to provide a migration path from a traditional NOS-like setup to something that actually *uses* the system. I believe Johan is working in the same direction with respect to NT; I think Phil cut the rope with his new stuff, though, and I know Alan Cox did. The point of these development paths is not that Linux, or any other particular operating system, is the way of the future. It is that DOS by itself has reached the point where it isn't viable; the 640K limit has become too severe a limitation. Witness Windows 3.x and --- more tellingly --- the upcoming 32-bit Windows 4. Heck, if I'd had any decent comm drivers for the DV/X djgpp distribution 3 years ago my own work would probably be in DESQview/X instead of Linux now. A DOS client setup is still doable, but any reasonable server configuration strains available memory and has done so for several years. It's only now that true 16/32-bit environments are becoming viable (read: inexpensive enough) for amateur radio networks. ++Brandon -- Brandon S. Allbery kf8nh@44.70.4.88 bsa@kf8nh.wariat.org Friends don't let friends load Windows NT (tnx Sun) A Linux iBCS2 developer ------------------------------ Date: Sat, 09 Jul 94 13:28:03 From: kz1f@RELAY.HDN.LEGENT.COM Subject: DOS To: tcp-group@ucsd.edu > | now. And what happens? All the serious thinkers are off with plans to > use LINUX as the operating system of the future. Brandon, I think his point was (similar to mine before the brush war) is that DOS users are pretty much left holding the bag. The Microsoft platform(s) solution doesnt exist yet and the fact they are still running DOS may mean they arent running enough to make it viable to upgrade to something else (OS/2 or NT or, yes, even Linux). No one capable of putting mosaic, for instance, into DOS NOS is going to, they are off on other platforms or they havent stepped up to the plate. I think the answer to his implied question is either: 1) thats life in the big city, or 2) find someone capable and willing to own NOS on a DOS platform. And, what will probably be the case is the former. Walt ------------------------------ Date: Sat, 09 Jul 94 12:25:00 -0000 From: mikebw@bilow.bilow.uu.ids.net (Mike Bilow) Subject: DOS To: tcp-group@ucsd.edu Cc: kz1f@RELAY.HDN.LEGENT.COM On Jul 07 09:46 94, kz1f wrote: k> Although I havent written Unix k> drivers, I have written OS/2 device drivers. My comment then k> still stands, one really has to want to do something k> sufficiently badly to start writting device drivers to k> accomplish it. There's an important difference. In Unix or OS/2, you have to write a device driver if you want to do certain things. For example, there is no proper way in either operating system to service a hardware interrupt other than by writing a device driver. In DOS, you can get away with writing these kinds of basic facilities directly into applications, and it became the standard way to do things for performance and other reasons. For example, how many DOS programs try to use the BIOS serial port services, even to do things that are not time critical, such as settings the baud rate? -- Mike ------------------------------ Date: Sat, 9 Jul 94 13:20 PDT From: bruce@pixar.com (Bruce Perens) Subject: DOS To: TCP digest , Klarsen DRSI already makes a 9600/1200 card. I think they list it for $300. There's no radio included, so add $100 for a TEKK (cheaper than an HT) and something for cables, antenna, feedline, power supply, etc. There's also the Gracilis card/radio/software package, which is about the same as the above but works out of the box and costs $100 more. What you get is something that's OK for connecting to a BBS, and doesn't have much of a user interface. The user interface is going to be a _BIG_ problem for the average ham. Maybe Windows would be better for this reason. Bruce Perens AB6YM ------------------------------ Date: Sun, 10 Jul 94 01:17 EDT From: glg@balrog.k8lt.ampr.org (Gary L. Grebus) Subject: DOS and Packet drivers To: A.Cox@swansea.ac.uk >The big problems with the Linux TCP/IP stack and AX.25 I had >have been > >o Use of large windows - DOS tcp stacks let you choose small windows >o Large MTU (ethernet is 1500 right....) thus large MSS >If the other stuff is well implemented then the initial RTT is way too fast >so you tend to get a couple of SYN frames out to fast but once thats happened >it settles down quite respectably. That's one thing that works better in *BSD networking code. You can associate an irtt and mtu with each route, so both your Ethernet and radio links have good defaults. Unfortunately, there are also application timeouts which may be too fast. Name resolution seems to be one of them, both on the my 386BSD system, and on the Windows system with Trumpet Winsock. /gary K8LT ------------------------------ Date: Sat, 9 Jul 1994 16:27:34 -0400 (EDT) From: Bill Horne Subject: Need NOS for Windows To: tcp-group@ucsd.edu Jack, N1REU, has just gotten started on TCP/IP, and needs a nos which will run under Windows. Please tell me what's available. Thanks. Bill aa1es -- E. William Horne bhorne@lynx.neu.edu (Internet) aa1es@switch.foxboro.ampr.org (TCP/IP) aa1es%switch.foxboro@WA1PHY.#EMA.MA (AX.25) (617) 784-7287 (home phone, voice only) ------------------------------ Date: Sat, 09 Jul 94 00:22:55 CDT From: route66@ddl.chi.il.us (System Administrator) Subject: TCP/IP Question To: tcp-group@ucsd.edu ok, TCP/IP Question for you all... I know you can do Telnet/FTP over TCP/IP. But, can I do other stuff over TCP/IP such as Mosaic/Gopher/Archie/IRC? If so, how? -Bye ------------------------------ Date: Sat, 09 Jul 94 11:14:27 EDT From: JonesA427@aol.com Subject: unsubscribe To: tcp-group@ucsd.edu unsub jonesa427@aol.com tcp-group@ucsd.edu ------------------------------ End of TCP-Group Digest V94 #144 ******************************