Hacker News Reader

Comment by throwaway5959 | Hacker News Reader

I attended the SoCal Vintage Computer festival, went to a talk and when they asked for questions, seeing no one else asking questions, I asked if anyone wrote malware for these systems now that they’re internet connected. The inventor of the device (Thomas Cherryhomes) immediately belittled the question and asked why anyone would want to do that (the irony seemed to escape him, given that he built a network interface card for a device obsolete around the time I was born, but I digress). Fortunately Jon Decuir, a former Atari engineer also on stage giving the talk gave a decent thoughtful answer quickly thereafter, but that was enough.

VCF SoCal was my first and last vintage computer festival.

View on HN

Technically a valid question, but imho a problem if you take the network oart seriously.

Most of these festivals are more about shared nostalgia than objective appreciation of technology from a specific time.

Like watching one of those videos from LGR. Watching him power up an old computer and hearing that "thunk" sound is beautiful.

And slighty offtopic, but I think nostalgia is fun when shared with people. Being nostalgic alone is a mix of warm and sad.

Yeah and being nostalgic around people who don't share it is difficult, since if you talk about it then they will find you insufferable.

Unrelated but I can remember when a friend of mine came back from a long, glamorous trip and suddenly stopped sharing in our usual moments of nostalgia. They acted like they were embarrassed. It sucked because it made me believe that they were ashamed of who they used to be, and I felt sorry for them. This person has since made great efforts to adopt a newer, more glamorous persona. I miss when they were genuine. I miss my friend.

I met with some old college friends and we were discussing some of our antics at the time. I sort of cringed when being reminded of the lame stuff I had been doing at that time. That was me then, and now I'm a different person. I don't think I'm not being genuine though.

Seems a bit harsh to judge an entire event by the conduct of a single speaker, but to each their own, and all that.

To answer your question, yes, people are still writing malware for old systems. They don't even have to be network connected: computer viruses can spread perfectly fine from other media as well. People still send each other floppies in the mail.

It's not as prevalent as on modern mainstream systems of course, but for example, a new Amiga virus started spreading a couple of years ago. The Jackal bootblock virus overwrites the bootblock on floppy disks, possibly rendering them unusable.

https://www.vht-dk.dk/amiga/desc/txt/jackal.htm

Thanks for answering the question.

With regard to the show, I didn’t want to pile on, but here we go:

There weren’t that many talks (the primary thing I was interested in), the trade show section was pretty small, and there wasn’t really much for newcomers to explain why this old hardware deserves preservation.

Vintage computer meetups, like vintage car meetups, are mostly chummy in-group gatherings for the already initiated. It's where geeks of a particular brand can talk shop. (I usually don't attend them and instead go to demo parties[1], which are in no way any less esoteric.)

> There wasn’t really much for newcomers to explain why this old hardware deserves preservation.

It does and doesn't - much like old cars, record players, fighter jets, trains, radios and phones. A handful can be kept around for posterity, as museum pieces. There's a point in not forgetting how they work, since most modern stuff is to a large extent refinements and complexity layered on top of previous knowledge.

But mostly, it's just a fringe hobby. Old computers deserve preservation because someone cares deeply about them and loves tinkering with them. Just like with old cars, record players, trains and radios.

[1] https://en.wikipedia.org/wiki/Demoscene

"Seems a bit harsh to judge an entire event by the conduct of a single speaker [..]"

The answer was extremely rude. Had they said, they got rid of the documents, because they'd valued them differently from Jason, that'd have been rude and disappointing, but maybe understandable to some degree.

At least these things happen. A former landlady of mine had donated a rare and precious motorcycle she inherited from her father to a local museum. When she coincidentally learned about 10 years later that the museum was about to sell it, she bought it back for a lot of money. Donations can sometimes turn out not the way we want.

In my opinion, saying they just kept the boxes was appalling and beyond disrespectful and I think Jason's reaction is understandable. Of course we've only heard one side...

I was replying to the parent comment.

Sorry and thanks for pointing it out. I should have read more carefully.

As the speaker in question here, I wish to reply.

I am formally apologizing for minimizing the question. It was not my intention to be derogatory.

But to formally address your question:

The same issues that affect any FujiNet device, are the exact same which affect any IoT device. The problem sets are one and the same, and are addressed by a combination of disciplined test driven development (which we are now doing), and auditing (which we need people to help with.), as well as leveraging fixes from the upstream vendor framework (ESP-IDF).

Since this issue is very close to your heart, would you like to help address this issue directly? All of the issues that the FujiNet team addresses are a direct result of champions who drive them forward.

Thank you for your time, Thomas Cherryhomes, Firmware Engineer, The FujiNet Team.

The 8-bit era was an interesting time because DRAM memory had finally become cheap and there was no clear cut market leader that had a definite average over the others. This lead to a lot of interesting innovation in the market, some truly unique designs, companies who bet big, some won and some bet it all and lost. The software industry for micros was also just getting started.

There was this whole period where machines were getting small enough that an individual person might be able to afford a computer, and away from unit record systems that were basically glorified accounting machines that worked with punched cards and paper tape into things such as refreshable text and graphics displays, sound chips, etc.

Why preserve it? Well if you want to know where you are going it helps to know where you've been; and yes, there's a heavy dose of nostalgia there for those of us who grew up with these machines.

Have you never wondered what the GUI and mouse was like BEFORE Windows and Mac? You could actually try it on a Lisa or Xerox Star, an Amiga or Atari ST, or using GEOS on the C64..

Joe Decuir is the man - there are a number of really good presentations of his on youtube. And the parent is talking about Fujinet, which is all the rage:

https://fujinet.online/

It's too bad Cherryhomes shot down an interesting question.

I just watched the video from that presentation and it didn't sound like anything so egregious that it would make you want to never attend a similar event was said (we couldn't hear your actual question very well). I get that there's a little more to your decision than that, but still.

It is not possible to infect 8-bit computers from the 1970s and early 1980s with malware. The operating system is on a dedicated chip that would have to be physically replaced with a new chip that contains the crippled OS. This is in stark contrast to more modern computers where the OS can be modified by software (e.g. malware).

Is the question about infecting the 8-bit computers connected via the ESP32-based FujiNet device, or about infecting the FujiNet device itself? As Cherryhomes himself points out, that device is VASTLY more capable than the vintage hardware/software to which it's connecting. A compromised FujiNet running in a trusted network would be a significant security issue.

He's intensely invested in it, and sometimes comes off frustrated at the lack of interest (based on what I see on mastodon, happen to share instances). The suggestion of malware was probably somewhat offensive as it is antithical to bringing disparate retro system users together.

Sorry that happened, but I think he is an unusual case. Hoping to provide some understanding for why it played out that way even though your question was reasonable.

Security through being offended at the idea of needing security is rarely a winning plan.

They seem to be ESP32 based wifi modems talking to 8-bit machines over a variety of interfaces. I don't think exploits compromising the 8-bit machines are a serious concern or even likely. Presumably the question was just being playful curious.

The IoT-style modems, maybe, but that wasn't what was asked. I'm not sure what the architecture is like for the network so I can't really say if there are security concerns there.

I have various electronic/computer things that various people would silently scream if they knew I was tossing them. I'll probably make a nominal effort to find a home but just a nominal effort and I bet they'll end up in someone else's attic after they've played around for a day or so.

I'm the 'inventor' (seems a bit of a weird term) of the Spectranet, a similar device for the Sinclair ZX Spectrum, although it's wired ethernet rather than wireless (when I designed the device, there weren't simple-to-interface devices like the ESP32 readily available, however, the Wiznet W5100 chip for wired ethernet was, and is a perfect match for an 8-bit system).

It would be possible to write malware for a Spectrum using this device but I don't think anyone ever has - needless to say, a Sinclair Spectrum isn't a secure computing platform. It would certainly be possible to write a worm.

Back when I was a teenager, we had an econet network of BBC Micros/BBC Masters and Acorn Archimedes at our school, and I did indeed try to write a worm for that platform (essentially, attaching itself to a user's !Boot script - essentially adding some 6502 code to the start of it to allow the program to spread itself to other !Boot scripts - the BBC micro allowed you to attach 'hooks' to its system calls, so as long as the user didn't overwrite the memory where your program was or do FX 200,2 followed by Ctrl-Break (a precaution I always took before logging on :-)) you could keep a small program memory resident. Especially in a Master 128.

I gave up because I realised the teacher who ran the computer lab had started taking quite a close interest in the code I was writing, and I wasn't entirely confident that he didn't know 6502 asm (one day I forgot to PROT my system, and I noticed a slow down while running the assembler, the slow down a sure sign someone was using VIEW to view my screen. So I ran PROT, which disabled that, and within 30 seconds the guy was in the room and looking over my shoulder! He knew that I knew, and he knew that I knew that he knew what was going on) Afraid of being caught and banned, I abandoned the project. (For the avoidance of doubt, the 'malware' in question was a bit like the idea of the Morris worm, simply to spread itself but with no payload).

However something similar could be done on other 8 bit systems. The "mal" part of the malware would be extremely limited - with such a small amount of RAM, no multitasking and no memory protection, anything you load into memory isn't going to last long once a user loads a real program.

The real target of any malware would be things like the TNFS (network filesystem) daemon, as to be able to run on vintage hardware it's got to be written in plain C and there could be bugs, and the filesystem protocol being very simple and unencrypted is not secure (nor is it designed to be, the preface on the README for tnfs does tell users they are mad if they try to store any data they want to keep private). The other target I guess in the case of ESP32 based devices is the ESP32 itself, but that's not really malware targeting the vintage system.

Reply on news.ycombinator.com