• Please review our updated Terms and Rules here

PDP-8 Ramblings

vrs42

Veteran Member
Joined
Nov 24, 2009
Messages
999
Location
Beaverton, Oregon
I've been recently working on tooling to extract files from DECmate WPS floppies. I've also been looking at compatibility issues with Edge and my website, so-much-stuff.com. And thinking about building WPS8, looking at OS/8 build issues, not to mention PQS8, COS, and other neglected PDP-8 operating systems. Thinking about how there needs to be a cross-assembler for MACREL, etc. etc. How my website uses HTML 1.0, SVN, and Eagle CAD, none of which are considered cool anymore. How I'm neglecting hardware projects like the Omnibus memory. And basically, on and on like that.

I'm at an age now, where this summer the government informs me that I'm old enough that I should be required to subscribe to socialized medicine. I can understand that. I feel that if I'm lucky, I have perhaps a couple dozen "year long" projects left.

So, I'm starting a thread to ramble about what still needs doing, and how to prioritize what I work on, and maybe even how to keep things going after I'm gone.

Vince
 
Vince.

I suggest you make a list of your projects and then put them in order of what you want to do most to what you want to do least. Take into account the fun factor.

Once you have your list you are all set.

I'm not far behind, I just turned 62.

We are all here to support which ever you choose.

Mike
 
I suggest you make a list of your projects and then put them in order of what you want to do most to what you want to do least. Take into account the fun factor.

Once you have your list you are all set.
Perhaps you are reminding me that I'm under no obligation to do stuff that isn't fun :).

I'm aware of that, but I also have these crazy notions about what I'd like to see happen, and so I worry.

I'm thinking I might use this thead to describe various things I'm either working on or thinking about working on, and maybe someone will get inspired to take some of them on.

One thing that's certainly fairly low on my list is the website. The last time I upgraded the engineering drawings from Eagle 4 to Eagle 6 (I'm licenced for 7), it took me months. I think there are serveral issues with the website currently:

The current implementation is a little bit of PHP (to list directories and fetch modify times) and plain vanilla HTML. I made a conscious decision early on to avoid cookies, client side scripting, and other potentially evil stuff. There's also no reason to consider stuff sensitive or secret, so it was all http:. Recently the HTTPS everywhere movement led to me needing to allow HTTPS fetches of the pages because otherwise the browsers won't render them. But now apparently I need to walk all the pages and fix all the URLs to be HTTPS also, or the pictures won't render, etc. I also see that right-clicking download doesn't work either, unless you specifically override the security warning. All fairly easy to fix, but basically annoying make-work.

The use of Eagle for the engineering drawings is an albatross. The drawings should ideally be converted to KiCad, I think. That would allow more folks to enjoy them without feeling pressured to buy stuff. As I mentioned above, there are so many drawings that the relatively easy transistion from one version of Eagle to another took a lot of work. Converting to KiCad is an effort I'd class as a "year long project".

The use of SVN is also potentially an issue -- folks tend to look to github for stuff like that, and are not likely to have subversion installed. This makes it difficult for them to mirror the site or even to download a whole directory. On the other hand, the backing store for the site is *huge*. With my mediocre internet connectivity, I typically have to re-issue the SVN command a half dozen times to finish an entirely fresh check out the entire 78G before timing out. This is also a few times over what sites like github are willing to provide for free.

The hosting isn't much of a problem, though the fellow who helps me with that is older than me, so it's unclear which will survive the other, etc. Mirroring would be great, but I never seem to get an answer when I ping folks about it. My hope is that a few folks have managed to download the whole mess with SVN, and can a way to rehost the content whenever so-much-stuff finally bites it.

Vince
 
How long is your "short list" and what is on it?

Before Covid, I went to several HAM flee markets each year, and every time I see stuff and go thinking "I am going to build something neat with that". I am a sucker for things that have a CRT in it. My attic is filled with boxes of "going to build something neat with it". I started to get rid of a few of those projects, because they probably will never materialize ...
Henk
 
How long is your "short list" and what is on it?
Well, it's longer than I can remember (which is perhaps saying less than it used to).

What I tend to do is look at what I was working on most recently and whatever someone asked about recently. That usually sinks more time than I've actually got, so stuff gets pushed down (and possibly lost) on the stack.

So far this month, I reached a stopping point on the PQS8 build, helped Kyle locate some paper tape images for the PDP-12, started looking at what it would take to export the PDP-8/L drawings to an FPGA, did source reconstructions for the known versions of the random ISZ diagnostic (except the LINC portion for the PDP-12, which remains unfinished), wrote forensic tools to extract files from DIAL LINCtape images (thanks Dave!), and tried to understand WPS floppy formats. I eventually ended up porting a version of Kevin Handy's DMPROCES tool which I got from Dave's FTP site. That gets the text files off and converted, but doesn't shed any light on WPS itself, so that got me looking at trying to build the WPS code on ibiblio, which reminded me of all the other source code that Charles has archived there that no-one is looking at, the need for a MACREL cross-assembler, wondering if I was ever going to get back to debugging the boot/memory board on the 8/A, and generally leading me up to starting this thread.

Vince
 
Hmm I recognise the problem... I also like S100 stuff, other old computers and vintage electronics. What I usually do is take a step back. Just determine how much time and money some projects will cost. And I try to weight in how happy I will be when a project is finished. I can assure you that a lot of stuff has left the building without even touching the electronics of it. Especially when things were also waiting for a little while.

If I want to do all my projects and ideas I need to become 300 years old or so. That is not realistic, so I'm happy when projects go to people who also have fun of working with that old stuff. Don't forget, others like to fix stuff and write software too. That is part of the hobby.

Stuff like converting schematics to another program version? Meh, in my opinion that is useless. The documentation is there, and there are tools which people can use them self to convert your files. DEC stuff is well documented so someone can even draw it again. And within 10 years kicad is obsolete and there is something else. Don't go there, converting files is a waste of time.

I think the best what we can do is preserve and provide as much info and tools as possible to next generation collectors. So when you have a skill to complete some historic work because you know about unique stuff, then that is your goal...

So what I want to say is: " give a man a fish and you feed him for a day; teach a man to fish and you feed him for a lifetime "

Regards, Roland
 
I started to make a list...not sure if all the items will mean anything to others:
----
The website apparently needs to do more work related to "https everywhere".

The PQS8 builds for versions previous to 8z remain undone.

The PQS8 cross-build should be finished, so that file dates and dependencies
can be accurately maintained.

Build trees similar to the PQS8 ones should exist for OS/8 and the other
systems (COS, MULTOS, TSS, RTS, etc.), too.

The DEC "big list" is still running a little below 50%. Need to locate the
missing files.

For the DEC "big list" files that do exist, source needs to get found or be
reconstructed, as well as build technology.

The DECUS "big list" is still running a little below 50%. Need to locate the
missing files.

For the DECUS "big list" files that do exist, source needs to get found or be
reconstructed, as well as build technology.

Related to the "big list", a lot of documentation needs to be refiled.

The "big list" tends to obscure where things are, by presenting them by
part number. Some way to find them more easily is needed -- possibly by
using the "qfNNNN" mechanism to group them?

Schematics should get drawn or redrawn for missing DEC boards, as well as
complete peripherals and CPUs which haven't been done yet.

Various boards (M220 M8340 M8341 M837, etc.) need to get to where they can
be manufactured and distributed.

The netlist-to-verilog path needs more work and more excercise. Now that
the 8/i is done, an 8/L should be relatively easy. Then more peripherals
and the PDP-12 CPU (with FPP support) would be awesome.

8/A, DECmate, and gear of a similar vintage is full of programmable logic
devices, which need to be dumped and archived.

The boot/memory board needs a better understanding of the 8/A bootloader
issue, so that a single bootloader board will work in both machines.

Related to that, the 8/A needs to get reassembled and debugged.

The M837/M8650 CPLD enhancement for the boot/memory board needs to be
prototyped and debugged.

The Flip Chip Tester has become long neglected. Finish that work on the
negative logic tester, or abandon it.

The Flip Chip Tester needs more module descriptions/test cases.

The 8tools disassembler really needs to be enhanced to disassemble LINC
code.

8tools for the LINCtape need more work.

8tools for the DECmate need lots more work.

8tools needs a MACREL compatible cross assembler. Most of the post-OS/8 OS
work, and even the OS/8 CCL program, would benefit from this.

The 8tools paths for converting vintage media and extracting files are
outstripping the neglected transfer in the other direction (back to
vintage media -- rx01, rx01, etc.).

Possibly related to that is an old idea about using handlers in dumprest.
How best to do that with respect to non-standard media, like reading
LINCtape on TD8E?

Explore that idea of using the LINC processor to implement a C compiler.

Explore that idea of emulating the LINC on the PDP-8, possibly as a runtime
for that C compiler.

There's also that idea about a C compiler that shares the mk40 Algol runtime.

Finish the assimilation of the old directory trees from Windows 7, ideally
*before* the migration to Windows 11 or Linux or whatever is next.

Related to old directory trees, the "dirdups" tool to find duplicate
directories need more work/debug.

There are a pile of website archive folders, which need to be assimilated.

The pile of archive media images needs to keep growing.

The DS310 floppies need to get archived.

SIMH could use model specific settings for PDP-8 CPUs.

SIMH could use LINC/PDP-12 emulation.

SIMH could use an integration of the VC8E (and other) graphics options.

There was that crazy idea for a template matching disassembler...

The IOB6120 project box is still laying there...

The W076x box is also still sitting there...

The PDP-12 needs attention.

The PDP-8/i needs attention.

The PDP-8/L needs attention.
----
Vince
 
I don't have a PDP 8.

But I do have my first computer (an H89), a couple of pre and post IEEE 696 S100 bus machines, 3 PDP 11's, a medium sized safe securing my firearms (we haven't been to the range since COVID), a pretty substantial stack of spares for my old race car, and most of the driveline for a street rod project, all stashed in a garage that needs serious thinning in preparation for a move to another city.

At some point, many realize that they have more yesterdays than they have tomorrows.

I will be 73 in June

You are not alone.
 
At some point, many realize that they have more yesterdays than they have tomorrows.
I will be 73 in June
You are not alone.
Thanks!

Hopefully we've got a few years left yet :). I hope that folks will see something they'd like to work on (either on their own or with me), to improve the chances that thing gets done.

For example, I don't know anything about SIMH internals, so I haven't done anything about those projects.

When I saw the LINC instruction set, I thought "I wish I'd had these when I did my C compiler". Likewise, when I saw the Algol runtime and it's amazing code density.

I've got crazy ideas about how to do curation, but it's a bigger job (even restrained to just the PDP-8 and PDP-12) than I can do alone, etc.

Vince
 
Vince, I'm probably a little older than you, I approach each day thinking I have at least a few days left. I realized some time back that I'll never do everything I wanted to. So, I do what I have to (family stuff) and what I want to, when I can. I've went through all the software changes OrCad to KiCad, (I'm also into cars, wood working, astronomy,everything changes). I remember stuff that worked 40 or even more, years ago, but isn't a part of anything now. So just do the best you can and enjoy it. I certainly appreciate the help you have given me, Mike
 
It is interesting how old age, Covid-19, the up-coming World War III and the resulting world-wide famine makes us rethink and review our perspective on life and what is important and what is not.
 
So just do the best you can and enjoy it. I certainly appreciate the help you have given me, Mike
Thanks! I do enjoy it, or I wouldn't be able to do so much of it :).

I'm interested in folk's ideas about "what's needed" or "what's next" in PDP-8 preservation.

Vince
 
Well, I toke the PDP8E and the ASR33 to a church fair. We were looking for help in our volunteer IT group. The PDP8E generated some interest with the kids. You know blinking lights and clattering. I was thinking about talking to the sixth grade teacher about having an hour or two regarding basic computers and using the PDP8E as the example. Computers today are just a small all enclosed pack, Arduino's and PI's. They generally work the same but the PDP8E is all pulled apart so to speak. Sort of gives a more detailed picture. To preserve something you need someone to preserve it. Mike
 
Thanks! I do enjoy it, or I wouldn't be able to do so much of it :).

I'm interested in folk's ideas about "what's needed" or "what's next" in PDP-8 preservation.

Vince
I've been hoping the CJL P?S/8 work would be completed and released to the community. As matters currently stand it's likely, IMO, to head into final oblivion due to its unreleased status. While it's not as exciting as hands-on hardware (for some) it seems to be like a good bridge "back" from key-less teeny-tiny hip-pocket devices to the types of devices and their work-environments that got us to where we are today. We're fortunate that there is still an active community of HW maintainers, but I can quite easily see the unreleased P?S/8 becoming an apocryphal tale (it's already in that status, IMO), rather than something "alive". Such a cryin' shame. (YMMV)
 
Vince, don't underestimate the work you have done already... You have a tremendous amount of information on your website. I think you have to determine what you want to accomplish for your self first. Maybe you are seeing possibilities which others do not see. For example Doug's project with combining the terminal and diskserver on one serial port, I think that is brilliant. But you need someone who thinks about it and has also the knowledge to build it.

So maybe you have idea's to improve the possibilities with the old machines. Or like projects like P?8. I have no knowlede about that, but it would be great if that will be available with some documentation and bootable disk images...
 
I've been hoping the CJL P?S/8 work would be completed and released to the community. As matters currently stand it's likely, IMO, to head into final oblivion due to its unreleased status. While it's not as exciting as hands-on hardware (for some) it seems to be like a good bridge "back" from key-less teeny-tiny hip-pocket devices to the types of devices and their work-environments that got us to where we are today. We're fortunate that there is still an active community of HW maintainers, but I can quite easily see the unreleased P?S/8 becoming an apocryphal tale (it's already in that status, IMO), rather than something "alive". Such a cryin' shame. (YMMV)
I've been doing considerable work with P?S/8, collecting the source code and reconstructing the build for the most recent version. (That's got a thread here on VCForum.) I have it online in my svn repository, though I haven't yet constructed an HTML page about it for the website. That together with the PEPS binary release, I think establishes it as a "real thing" and no longer the vaporware that it was for so many years. That said, there is potentially another year's work to reconstruct the older versions, create installation media for versions other than DECtape or RK05, etc.

P?S/8 has been interesting to me, as it dovetails with my recent acquisition of a PDP-12. There's a lot of PDP-12 influence, I think, in P?S/8. (Or maybe it's LINC/8 influence,) There's something in that lineage too, about the higher capability of the basic machine and the insular behavior of the folks that use(d) them. (Basically, they don't seem to need each other as much, and so they don't seem to form the same collegial hobbyist groups. Or maybe they do, but I am just an outsider?)

Whatever the reasons, there are traditions there, and they can be seen influencing things like COS and ultimately the DECmates (which run COS, and WPS as well). But a lot of what went on in those labs and offices seems largely lost now.

Vince
 
Vince, don't underestimate the work you have done already... You have a tremendous amount of information on your website. I think you have to determine what you want to accomplish for your self first. Maybe you are seeing possibilities which others do not see. For example Doug's project with combining the terminal and diskserver on one serial port, I think that is brilliant. But you need someone who thinks about it and has also the knowledge to build it.

So maybe you have idea's to improve the possibilities with the old machines. Or like projects like P?8. I have no knowlede about that, but it would be great if that will be available with some documentation and bootable disk images...
I agree that Doug's project has great potential for renewing the 8/i, 8/L, and straight-8 communities (and their old machines). Those machines seem to be largely dust collecting showpieces at the moment, at least compared to the community of Omnibus machine users.

There are probably a pile of projects that I see, that few if any others would be prepared to tackle. (Perhaps some appear in that list I posted, which didn't generate much "Ooh, I want to do that!".) My collection of hardware and software does great opportunities. For instance, I think I've only seen Bill Cattey express similar ideas about the software curation and preservation stuff.

Maybe Bill can be pursuaded to do a P?S/8 release for the PiDP? I think that would really amount to doing whatever he'd like about organizing and preserving the source in his tree, then creating (or including) a bootable RK05 image for the PiDP release. That would get P?S/8 into more hands!
Hmm. That would, of course, aggravate the documentation issue. There's plenty of source material in Charles' notes though.

Vince
 
(That's got a thread here on VCForum.)
Thank you for mentioning that. The usual search attempts failed me, but (et voila!) "CJL" dug it out. I'm very glad to read of the ongoing work! Will review, looking for that binary release of PEPS :->. A release for the PiDP-8 would, IMO, be a great way to engage a wide audience of aficionados. Calling Bill Cattey ...
 
Another brainstorm...

Take Kyle's TU56 emulation, port it to Arduino or whatever, interface it to the vintage gear. This could add extra drives, substitute for broken ones, etc. while running the proper vintage software.

Optionally slave the motors on a real drive (Michael's broken head drives, say), to get realistic visuals.

Possibly add TU55 and earlier drive compatibility, configurability about g888 location, etc.

Vince
 
Back
Top