Hacker News new | past | comments | ask | show | jobs | submit login
Trying to make sense of CompuServe server hard disk images (medium.com/mpnet)
133 points by harel on April 2, 2019 | hide | past | favorite | 59 comments



Out of core context:

the article mentions "The Digital Antiquarian" ( https://www.filfre.net ) which in turn has an article about the game "Wing Commander" ( https://www.filfre.net/2017/04/from-wingleader-to-wing-comma... ) => this is funny:

> At the beginning of August, Snell unceremoniously booted Chris Roberts, the project’s founder, from his role as co-producer, leaving him with only the title of director. Manifesting a tendency anyone familiar with his more recent projects will immediately recognize, Roberts had been causing chaos on the team by approving seemingly every suggested addition or enhancement that crossed his desk. Snell, the brutal pragmatist in this company full of dreamers, appointed himself as Warren Spector’s new co-producer. His first action was to place a freeze on new features in favor of getting the game that currently existed finished and out the door.

(Chris Roberts is the one creating the game "Star Citizen" https://en.wikipedia.org/wiki/Star_Citizen - I was one of the first buying/financing the game in 2012, but I admit of having lost interest while waiting for the "final" version to be released...)


I too am a founding backer, also totally lost interest.


I wonder if twenty years from now, someone will be archiving one of the big email providers.

"In the early 21st century, there existed large cloud 'email' organizations. Here I found a trove of 'hard disks' images from one of the companies, HotMail. I've been digging around and look at these email addresses, it appears this image contains all the 'email' for around 10k users. Here is one where a couple users with the usernames 'joesmith' and 'andreasanchez', engaged in playful romantic flirting. Although by digging further in the emails it appears one of them, 'joesmith' is newly married to 'andreasmith'. Here is thier wedding invitation. Fascinating digital archeology"


This is what is happening with old letters as well. If you look carefully through antics markets you can often find old letters.

Random example: https://www.ebay.com/itm/1683-Old-Historical-Handwritten-Let...


Of course that's exactly what will happen. If you put your data in the cloud, it's one day going to be publicly available. If you don't like that, you should consider running private servers for your data.


That's assuming they do not properly destroy their data, which they are legally obligated to do as far as I understand. You could also just encrypt your data before uploading.


> which they are legally obligated to do as far as I understand

Says who? Do their ToS even address what happens if they're bankrupt?

It sounds like Apple can legally do whatever they want if they go bust: https://law.stackexchange.com/questions/30810/what-happens-i...

I expect the same is true of all the other major cloud providers.


It would be nice if we ever got any good privacy legislation for it to include a requirement that a company keep money in a trust to retain services to erase servers with private information of public citizens. That, along with some specific rules about what private information can be collected or retained, might greatly change the economics of the "online free service supported by massive data mining of customers" space, which IMO would be a very good thing.


My understanding is not all cloud providers go through the same great lengths to ensure all data is correctly destroyed, instead of just "Dark" but I sadly don't have any references I can site, just antidotes from people who work in these data centers.


You meant anecdotes, right?


And cite, right?


The question is if they still exist


Or encrypt your data before uploading it to "the cloud" (aka, somebody else's computer).


Encrypting something today is no guarantee that it’ll still be encrypted in 20 years.

If you encrypted something in 1999 with 56-bit key DES, your data is no longer private.


Reminds me of the bit from Cryptonomicon, which as it happens was published in 1999:

> [Randy] has pointed out to Avi, in an encrypted e-mail message, that if every particle of matter in the universe could be used to construct one single cosmic supercomputer, and this computer was put to work trying to break a 4096-bit encryption key, it would take longer than the lifespan of the universe.

> "Using today's technology," Avi shot back, "that is true. But what about quantum computers? And what if new mathematical techniques are developed that can simplify the factoring of large prime numbers?"

> "How long do you want these messages to remain secret?" Randy asked, in his last message before leaving San Francisco. "Five years? Ten years? Twenty-five years?"

> After he got to the hotel this afternoon, Randy decrypted and read Avi's answer. It is still hanging in front of his eyes, like the after image of a strobe:

> I want them to remain secret for as long as men are capable of evil.


If you encrypted something in 1999 with 56-bit key DES, you were using a cipher that at that time was not just controversial or considered weak by leading cryptographers (that was already the case in 1977), but also practically demonstrated to be broken.

If anything, the fact that brute-force still seems to be the most practical attack against DES shows that (symmetric) encryption lasts surprisingly long.


Something like a recent episode of The Orville perhaps? Where they find someones phone from the 21st, download all the messages, calls, etc and uses the holodeck to recreate what it might be like.


Wait a minute there's a public archive of Compuserve data noone has any clue what it is? What if it's PII? It wasn't that long ago to make this a non-issue.


As much as I'm fascinated by this work and the author's tenacity, as a past user of CompuServe I'm quite alarmed at the prospect of my data falling into the hands of a random tech junkie blogging on the internet.

Verizon, if you're out there - please get a handle on this breach before my lawyers call you.


Yes, I was a CompuServe user back in 1988. I'm very concerned about my teenage "CB" chat logs falling into the wrong hands.


Past Senate and current Presidential candidate Robert Francis O'Rourke felt a need to apologise for his adolescent fiction, which he posted online. You might be surprised what you posted back then. I know that when I've gone back and reviewed my own history I've been surprised what I posted for posterity as a kid. I didn't understand that it'd be out there for the rest of my life.


I didn't understand that it'd be out there for the rest of my life.

Neither did anyone else, which means it isn't going to matter in the long run for anyone who is just starting out today. If everybody has skeletons in their closet, no one does. And everybody pretty much does.

Remember that one of the two major parties in the US has willfully abandoned any pretense at moral authority by supporting Trump. The only threat Beto faces in that regard comes from his own party. The correct reaction would be for him to own his juvenile fiction, not apologize for it.


:-) Although I was no longer a teenager and don't think I ever used chat, it's hard for me to imagine just about anything in my early online presence that would cause me more than very mild embarrassment today.

Yeah, someone should have scrubbed these. But it's hard for me to get too worked up.


Eh, buy the ticket, take the ride.

Don't type anything on a keyboard today that you don't want to see on the front page of the newspaper tomorrow. I learned that lesson at 300 baud, long before I had ever heard of CompuServe, but that didn't stop me from forgetting it a time or two when I got my CS account. So if there's anything in this archive that would embarrass you, well, at least you're in good company.


Completely scary, what a massive security breach just casually thought of as a digital challenge.


What if it's PII?

If anyone finds data for 72167,7570, send a copy back to me.


...but seriously... I think I still remember my "license plate" number for prodigy.net, which I think was 7 digits alpahnumeric. I've very occasionally searched for it thinking there was maybe an archive or footprints from "those days" but nothing has ever come up.

Quite literally the transition of physical paper => digital forums => internet (pre-crawlers/pre-archive) => where the data could have been lost before it had a chance to be "modernized".


Perhaps the CompuServe PII wasn't covered by modern 'privacy' policies that have users 'consent' to the company doing whatever it wants.

A big company might be vulnerable to class action from a bunch of people who are tech-savvy and affluent enough to have been using Compu$erve.


I wonder, who is responsible in a case like this? Is it the current owners of CompuServe for not clearing the PII before trashing the machines?


This is very common. I've bought machines at auction with databases on them.


Yes, though not as bad as it used to be, IME. It seems a lot of ordinary people now know to pull or wipe drives before disposing of a computer.

I've acquired over 100 used computers, and, once I saw how common leaving data on them was, I began always very promptly DBAN-ing any disks, and wiping SSDs. If it has a non-SATA interface, I destroy the medium. Similar with some kind of wipe of smartphones.

Besides wiping the disk/device seeming a responsible and considerate thing to do, it avoids remote yet severe liabilities.


You know that you can't truly "wipe" SSDs, right?


It's not that simple.

A secure erase should wipe all blocks (and since it is done by the firmware, physical blocks, not logical ones), but there have been instances of buggy drives that don't actually do it properly.

Some SSDs also transparently encrypt all data written, and the key is replaced on a wipe command. Even if they don't wipe the data fully, it's now unusable, if the encryption works correctly.

Quite old SSDs might not support any of this.


> Even if they don't wipe the data fully, it's now unusable, if the encryption works correctly.

Unfortunately, you can't trust the SSD firmware developers to have their shit together [1].

[1] https://techcrunch.com/2018/11/05/crucial-samsung-solid-stat...


Which is why the if is in there :(


I thought all SSDs support ATA password.


I don't think ATA password necessarily means the drive content is encrypted, or helps you fully wipe a drive you obtained from somewhere else.


The no longer existent corporate person is responsible. Same as if a real person copied all your info, published it, and died.


They do still exist though, and they are a subsidiary of Verizon. So is Verizon responsible even though nobody at Verizon had anything to do with it? Does the person who redistributed the data have any responsibility?


> So is Verizon responsible even though nobody at Verizon had anything to do with it?

Imagine if CompuServe still existed, but they had undergone 100% employee turnover in recent years. They'd still be liable, right?

It seems to me that when Verizon bought CompuServe, they calculated that its assets minus its liabilities totaled less than the purchase price and hence it was a good deal. If we treat leakable data as a legal liability, it should send the right financial signal - eg, maybe as a company you don't want to collect much.


Imagine trying to find or even implement exploits against such a system:

> The CompuServe system used non-standard PDP-10-based hardware along with a custom version of TOPS-10. As a result, booting it in one of the available emulators is next to impossible. But I hoped the disk images could be attached to a standard TOPS-10 system as secondary disks somehow.

Especially when all those manuals weren't freely available...


I dropped so much cash on compuserve as a kid... ugh the pain and realization when my folks figured out what it was..


For people who don't know what the costs were like, here's this page from a 1988 Whole Earth catalogue giving some pricing information.

http://tinypic.com/view.php?pic=2janfrd&s=7

Compuserve $11 per hour, had "more than 250,000 subscribers".

The Source $8 per hour.

Delphi $6 per hour had a loyal but small (less than 10,000 users) following.

BIX $9 per hour.


CompuServe charged more if you connected at a higher baud rate, like $12.80 per hour for 14.4k... I think - I haven't worked there for 22 years, so my memory is hazy.

There was also an ecosystem of apps to script connections to minimize time you spent online. You'd pick the forums you wanted to scrape new messages from, fetch them in one session, compose replies offline, and post the replies in a second session.


It was the same thing on private BBS systems. There were a bunch of applications and scripts that basically let you go online, do your business as quickly as possible, and hang up. I remember I primarily used a system running PCBoard and Qmail was the software that let you do most things offline.

ADDED:Private systems didn't (ever?) have a time charge. The bigger ones usually still offered some lines for free users but offered preferred access for subscribers.


I remember Delphi had one called DIJ/Delphi Internet Jet, I want to say it actually did everything in QWK format but it's been so long I don't remember. I do remember using QWK to get a lot of stuff off of WWIV<->UUCP gateways and FidoNet.

One of the things that kept me at the time on Delphi was 1) X-Files had a huge fanbase there and Chris Carter would regularly join the forums and 2) You could break out of the menus and get down to the VMS shell which is how I first learned something other than DOS/Windows.

I still racked up massive phone bills even with the offline readers.


That's cool. Was there any data caps?


On a 14.4kbps connection it took more than 10 minutes to transfer a megabyte, so the hourly billing rate was enough of a cap.


Long distance charges may apply.

I grew up in Ashtabula, Ohio, and the closest access numbers made the toll charges the greatest cost.


One of the peculiarities of telephone rates in those days was that it was very common for intra-state calls outside your local calling area (depending on your phone plan there was often a "free" local calling area and an extended region of some sort as I recall) to be more expensive than interstate calls.

ADDED: BTW, in the early eighties, long distance rates were starting to come down but they were still in the tens of cents per minute so something over a dollar/minute in today's currency.


Ugh. Yeah I had a few $400 bills in the late 90's. CS charges plus local long distance to the POP. Being addicted to Worlds Away didn't help either.


I spent nearly $400 on CompuServe playing "Island of Kesmai" as a kid.

It was the first and only time I ever saw my dad visibly tremble with rage and turn red. ($400 80's bucks is a lot of 2019 bucks to spend on a video game.)


I saw one of the CompuServe PDP-10 clones running at VCF West last year in Mountain View! It's this one on Youtube: https://www.youtube.com/watch?v=N1CvBwt5nfg


GEnie seems to be another online service lost to time: https://en.wikipedia.org/wiki/GEnie


There were a lot of them although Compuserve was at least one of the biggest. (It's the one I used.) There were also BIX, AOL, DELPHI, and Prodigy, and others that aren't popping to mind.

There were also a vast number of small BBSs ranging from a system in someone's bedroom to commercial--though still relatively small--operations, some of which became early local ISPs. At the time, one mostly preferred to make local calls because of the expense of long distance.


"Digital archeology" isn't the right term, "digital preservation" falls under archival science, whether formally or informally.


Except preservation on its own is not very useful. You need some sort of understanding of the data on many levels hence "archeology" (that may not be the best term but I think it gets the point across).


They find some accounting details, rather than something more interesting.


I don't understand the downvotes, it's an accurate summation of the [edit result of his work]. The journey [edit: up to and including that result] was an interesting read.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: