NetBSD runs on VAX, and has support for running a.out binaries from 4.3BSD. NetBSD also includes a IPv6 stack (KAME), and they have the best cross-platform source-based package management system, pkgsrc. Aside from how you're going to install it (and you're going to need, from the VAX's perspective, a giant hard drive - 200-400MB is what the base system alone will need), you'll also need a *lot* of patience while Apache and friends compile. Good news, though, your stuff won't compile to mind-boggling (even for modern machines) large binaries because NetBSD/vax has VAX support in it's ELF ld.elf_so. Don't know what the virtual memory subsystem is like for NetBSD/vax, but if it's anywhere as good as Linux or any *BSD on modern machines, you'll have copy-on-write fork()s and libraries that share read-only segments across unrelated main executables, which will be extremely useful in an environment strapped for free RAM. On Mon, Jun 4, 2012 at 7:12 PM, Jonathan Katz <jon at jonworld.com> wrote:

^^^^ Wow! Does it play chess? Tic-tac-toe? Global Thermonuclear War? :oD

I was really hoping the V in the acronym was going to be VMS. It's just another *nix box unless you're running VMS. -- "The Direct3D Graphics Pipeline" free book <http://tinyurl.com/d3d-pipeline> The Computer Graphics Museum <http://computergraphicsmuseum.org> The Terminals Wiki <http://terminals.classiccmp.org> Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com>

Personally, I tend to think that if you have a VAX, it would be more fun to use VMS, WASD (a native web server for VMS with lots of cool features including support for all the usual web languages plus DCL) and the database of your choice (either native RMS tables for simple stuff, Rdb if you can get it or MySQL if you can't) than to try and get some BSD-based system set up. Mike

On 06/05/2012 12:38 PM, David Brownlee wrote: I think this is a wonderful piece of work. Congrats on getting it running! I'm sure that was no easy feat, especially with the VAX's lack of IEEE floating point support. (which everything seems to depend on nowadays, whether it does one whit of math or not!) -Dave -- Dave McGuire, AK4HZ New Kensington, PA

In article <CAGN_6pbt1bssrnPPdmmcxm-menfsHtYyG7Sea6UAUM-MgEimUA at mail.gmail.com>, David Brownlee <abs at absd.org> writes: Yes, I don't mean to belittle the accomplishment; VAX is different enough from x86/ARM/MIPS that having *nix running on it with a modern software stack is interesting, I was just hoping it was going to be VMS. -- "The Direct3D Graphics Pipeline" free book <http://tinyurl.com/d3d-pipeline> The Computer Graphics Museum <http://computergraphicsmuseum.org> The Terminals Wiki <http://terminals.classiccmp.org> Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com>

On 5 Jun 2012, at 21:32, Dave McGuire wrote: As Phil Schiller would have it (well known for delivering the MOST boring Apple keynote segments ever) "It's all about the Hardware AND the Software". ;) An up-to-date AMP stack is not 'just a webserver' though, it includes modern versions of widespread languages like PHP 5.x and MySQL 5.0 or 5.5. Other scripting languages and databases are available (and I encourage people to look at them at least in passing). I appreciate web servers on VAX are not new, hell a lot of the Internet's popular early academic and institutional web content was served out off VAXen (probably running UNIX or VMS) back in the late 80s/early 90s. -- Mark Benson http://DECtec.info Twitter: @DECtecInfo HECnet: STAR69::MARK Online Resource & Mailing List for DEC Enthusiasts.

On Fri, Jun 8, 2012 at 5:16 PM, Jochen Kunz <jkunz at unixag-kl.fh-kl.de>wrote: Apache and PHP for a couple of years. It worked well. At some point I upgraded the hardware to a VS4000/60. I had trouble with NetBSD 1.6 and later and finally stopped using the vax for my web server in about 2004. -chuck

Richard wrote: I can't imagine why not. TECO is Turing-complete. I wouldn't want the job of writing it, though. I've got less productive things to do with my time.

On Sun, Jun 10, 2012 at 12:34 PM, David Brownlee <abs at absd.org> wrote: I couldn't get the packages to build and it was usually something floating point related. I think Perl was one that I looked into the most.

I'm trying to figure out which adapter I should get to attach my wyse 60 to the console of a microvax III (3900). I know I need a MMJ jack to DB25 male cable, but there are a few flavors. At first glance it appears that I either need H8571-D or H8575-D, however its possible I ned H8575-E. Someone know which one to purchase? I have the BC16E cable required... Kevin

No, I won't, and you won't blow a (real) fuse, or an IC, or anythjing if you try it. One good thign abotu RS232 is that it is designed to withstand having signals shorted or mis-connected. I wouldn't quite do that. I'd grab one of thos little DB25-DB25 adapters with LEDs to monitor the most important signals. Plug that into the dapater first (with the adaptor conencted to the (powered-up) Microvax) and see which line (2 or 3) it's transmitting on. One will be driven, the otehr won't be. Then if it's driving pin 2, you need a null-modem. I have solved many RS232 problems with little more than that. The worst thing I had to handle was an HP RS232 interface that correctly (as per the standard) did the half-duplex turnaround with the control lines. Problem is, almost nothing else does that... Most of the time I look of odiities in the interface (many printers put a busy/ready line on pin 11 for soem unknown reason), then use the LED monitor thing to see if it's a DTE or DCE. 99%+ of the time the cable I then wire works first time. The things I keep around are : PC/AT (DE9) to DB25 adapter DB25 socket to DB25 socket null modem with the flow control lines looped back -- that is 4-5 and 6-8-20 on each conenctor (not connected between the ends), 2/3 crossed over between the connectors and 7 wired straight through DB25 socket to DB25 socket null modem with the flow control lines crossed over between the ends. That is 2 crossed with 3, 4 corssed with 5, 6 strapepd to 8 on each conenctor and crossed with 20 (from the other conenctor) and 7 wired straight through A universal gender cable. I've not seen this commerically, it's easy to make. It's a couple of feet of 25 way IDC cable with a DB25 plugn and a DB25 socket crimped on near each end (4 conenctors total). Does away with the need for gender changers The LED monitoring adapter I mentioned. An assortemt of conenctors, multicore cable and a soldering iron for the permanent job. I do have a breakout box, adn even a datacomms tester. They are very useful -- when I need them -- but most of the time the above bits get the job done and are quicker. -tony

Err, right... While I can understnad not wantign to spend hundreds or thousands of dolalrs on tools and test gear just to get one machine up and running, I do think that spending $10 ro so to get the machine working is well worth it (hint : If you had to pay me for my time in sorting it out, it would be a lot more expensive than that :-)), and anyway, an RS232 tester is something you'll ened aagian and again if you work on classic computers. Can you stil lget tyhose LED adapters? At one time just about every PC shop sold them, but now, since RS232 is out of fashion, I;'ve nto seen them on sale for several years. It's not hard to make one, but... The other trikc is to see what flow control lines are being driven, and it noting works, to try loppign them back to what they normally pair with (e.g. if pin 4 -- RTS -- is being driven by a device and you can't get it to sand anything, try connecting RTS to CTS (4 to 5)). You migth end up with a thing that drops data due to buffer overrun from time to time, but you sort that out once you've got it sending something. -tony

I'ev never seen RS232 tersters or breakout boxes with anything other than DB25 conenctors. Were there ever ones with DE9 connectors, for example? Sure. The point of my (and others) post was that if you get _any_ adapter you can then use the RS232 tester to figure otu which is TxD and which is RxD, and if necessary make up an adapter using easy-to-get DB25 conenctors to lin kit to your terminal. That might not be as neat as having the right adapter, but it will get the thing working. -tony

Hence my comment (in anotehr messagee) to use the LED adapter to see what flow control lines the device drives. If it drives something, it's possible it requires the counterpart to be asserted before it will do anything. Another trick I've used is to use the diode test feature of a DMM (or an analogue multimeter) to check every pin o nthe DB25 to ground both ways round. You will pick up the input protection diodes of the receivr chips and the output stage of the drivers. Thus tyou can tell which pins go somewhere, and which are not connected at all I assuem it also powers down the drivers for any flow control outputs. That is plain evil. What device does this? A third thing you cna't test with the light adapter on its own is an RS232 drvice that's powered from the RS232 signals, like some (most?, certainly not all) serial mice. I think it's fair to say that no solution works in _all_ cases. That doesn't mean the LED adapter is not handy to have around. It works in enough cases that it's a useful thing to have in the toolkit. -tony

On 14 Jun 2012 at 17:30, Dave McGuire wrote: Linear Technology EIA transceivers ; e.g. LT1080--pretty much have always had that feature--with an ON/OFF pin. --Chuck

I've seen severail RS232 driver chips which have a logic-level input to turn off the SC0DC ocnverter and power them down, but Iv'e nto seen one that is controleld by a signal on one of the RS232 input lines. That seems like a misfeature to me. -tony

On Wed, Jun 13, 2012 at 1:59 PM, Tony Duell <ard at p850ug1.demon.co.uk> wrote: I have one, don't know the original vendor, but it's badged by Black Box. DE9s on both ends. I carry it around with DE9-DB25 adapters. Very compact. For those that attended VCF East 8.0, I used it to figure out what adapters would be required to hook up a VT420 to a USB serial dongle on my laptop so I could give my demonstration of Scott Adams adventures and my own ZDungeon more of a vintage feel than you'd get on a 1-year-old Dell. -ethan

Fry's still carries them. Not super cheap, but I still see a few in stock in the two Fry's shops I frequent. -- ------------------------------------ personal: http://www.cameronkaiser.com/ -- Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com -- A different taste in jokes is a great strain on the affections. -- G. Eliot

Hi Is the //c connector any different than a IIe? Dwight

dwight elvey wrote: Yes. And IIRC, the //c+ connector is different than the //c connector.

If you go this route, I'd love to have one of your main boards...Mine is fried and looks like one of the PALs. I already xchanged the dedicated ICs with a //e and found no TTLs fried :'(

Danke, Dave! But I have a PAL burner (Beeprog, from Elnec). The problem is exactely the equations :o) They are not known :( And I'd love to get a newer //c board, with the memory expansion pinout. I'm desiging a SRAM card for the //E and would love to adapt it for the //c Thanks a lot Alexandre :o)

'Twas always the way. I very rarely go to such places since they're unlikely to have anything I want. Even when my PC wasn't _quite_ os old-fashioned, they enver had the cable or wahtever that I wanted... I can rememebr one time I visitied such a palce, they had a few shelves for 'operating systems'. All that was there was the installation and upgrade kits for Windows. They sold Linux, but classed it as a 'book' (!). Hmmm... OK... Well ,that depends... It's not a waste of time if they;re no logner avaialble. And the one I got was unrleiable from new due to an excessive numebr of dry joints inside. I resoldered the lot and it's been fine for getting on for 20 years. Resodlering one probably takes less time than making one, but... I can. It was a couple of months back. I'd bought and then repaired a piece of test gear [1] which had a helpfully-labelled 'RS232 (DTE)' connector on the back. And it was clear from the setup options on the display how to make it use the RS232 port rather than the GPIB prot, and how to set the baud rate. So I did, and found it would send nothing. Strapping the handshake lines back got it to output a prompt to the terminal. Alas I've not got any further, I can't figure out what commands to send. [1] A rather nice telephone line simulator. We talked about those here soem time back as a device for woriking with and testing old modems. I decided I wanted somthing I could use for this and came up wit ha sort-of design. I then foudn there were 3 options, all of which would cost aobut the same : Make the design I'd come up with and debug it Buy a cheap, new device for the purpose, which wouldn't do anything more and would probably be hard to repair if it failed Buy a non-working professional line simulator and repair it. On the grounds the last would provide me with the best device in the end, I went that route. If I'd not been able to repair it, I'd guessed that many of the bits in it would be useful for my homebrew idea. As it happened, it took me quite some time to find the fault and then a few pence worth of parts (a couple of LM311s and an LM337) and a few minutes to fix it. Well, there are 6 microprocessors and a total of over 200 ICs spred over 10 PCBs in the darn thing ;-) -tony

On Jun 10, 2012, at 8:41 PM, Dave McGuire wrote: Well, of course, if you want to be a purist about it you can always replace the unwieldy mess with the correct cable later, when people are actually going to *see* it. Getting it running is hardly the time to be a stickler unless it's actually going to damage something. - Dave

In interest of having the quickest and most reliable solution, I took a look into the dec vt1300 and VR320 x-terminal I have, and found out that not only can it do x-windows, but if you ensure the rear s-3 switch is in the correct position (down), you can connect the console from the vax 3900 directly to the printer port of the vt1300, and use the "DTE" command from the vt1300 boot logic to get a dumb terminal. Works wonders. This eliminates the need for me to connect the vax systems to my wyse 60 enitrely, especially since the dec keyboard has additional gold keys that make things much easier in the long run. Nothing like having to figure out that the macintosh thin keyboard keypad "-" character is the "do" character to get into EDT/TPU or EVE command mode. Kevin

On Sat, Jun 9, 2012 at 7:23 AM, Kevin Reynolds <tpresence at hotmail.com> wrote: Here are some helpful references: http://h71000.www7.hp.com/wizard/padapters.html http://www.lammertbies.nl/comm/cable/dec-mmj.html

On Jun 11, 2012 3:30 PM, "TeoZ" <teoz at neo.rr.com> wrote: about the RJ45 ethernet to MJJ adapters (I have one Amphenol 482560001)? Is that a coupler that you plug an MMJ cable plug in one end an an RJ45 cable plug in the other?

Are you sure this has anythign to do with ethernet? My guess is that it's sued to linke DEC serial ports (MMJs) to the modern 'structured wiring' found in soem buildings. You'd use them in pairs, one ant each end of a building cable so as to connect the 2 serial devices. And, please, it's not RJ45 unless you are taling about a particular US telephone wiring scheme. 10baseT ethernet, etc does not use an RJ45 conenctor. It's an 8p8c. -tony

--- On Wed, 6/13/12, TeoZ <teoz at neo.rr.com> wrote: Adapters to go from DB25 to "RJ45" [1] connectors are very, very common - however there is no standard pinout for RS232 on RJ45. When you get such adapters new, they are sold as kits - the RJ45 jack has wires coming from it, crimped onto pins that are left loose. You are required to plug the pins into the DB25 in the required configuration. As such, when you find used adapters, they're likely to be in any possible configuration - you'll have to pop them apart and see how they're wired, and rewire where necessary. On the subject of MMJ - they are physically the same size as a 6p6c modular plug, so you can easily file the clip off such a plug and use it in an MMJ jack - it won't secure in, of course, but it'll work. Not that MMJ's stay in particularly well as it is - that stupid offset clip means that the connector sits in at an angle sometimes, and then occasionally makes poor contact. The only real MMJ cable I have has issues plugging into my MicroVax 3900's console port - which is slightly damaged from corrosion from the battery. It'll randomly stop working, and you have to remove and reinsert it. The filed-down 6p6c plug doesn't have this problem. DEC had lots of great ideas. MMJ was most definitely not one of them. Heh. -Ian [1] Yes, I know that "RJ45" refers to a specific wiring of a particular Registered Jack, not the 8p8c connector itself. But the tech industry has standardized on referring to 8p8c modular plugs and jacks as RJ45 - regardless of the term's initial meaning. This is similar to referring to 9 pin DSUB connectors as DB9 - despite the shell size not being B, it's E - the correct name is DE9. But the majority of the industry calls that connector a DB9. "Common usage" and "exact, correct, technical term" aren't always the same thing.