Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

DOS is easy to emulate - and dosbox does a great job of it, even in a web browser.

Windows 3.1, 95, 98, Me are less easy to emulate.

Note that that seems to have impacted the preservation of old games and programs. Plenty of dos games are all over the web and still quite popular, yet most stuff from the Win 9x era has almost entirely vanished due to the difficulty of running it on modern hardware.

Archivists take note - if you want something to live for a long time, it needs to be easy to emulate. And in turn, that means it needs to be both very common, and have simple API's so someone in the future can be bothered to make and maintain an emulator.



> most stuff from the Win 9x era has almost entirely vanished due to the difficulty of running it on modern hardware.

The tricky part is that this applies even if you're using a VM. I learned the hard way that Windows 98 isn't compatible with Ryzen CPUs, even through VirtualBox. I had to try again on another PC with an older Intel CPU.


A patch is available [0] to allow Windows 98 to be virtualized on more modern CPUs including Ryzen CPUs. It patches the "TLB Invalidation Bug". [1]

[0] https://github.com/JHRobotics/patcher9x

[1] https://blog.stuffedcow.net/2015/08/win9x-tlb-invalidation-b...


For early-ish windows 98 era machines, 86Box is a very good option.


DOS may be easy to emulate and re-implement because it's a single task operating system that does not do much. Most of hardware is accessed directly, and needs to be emulated instead. We enjoy great compatibility because of the enormous leap in performance since then (the slower the system the easier it is to simulate correctly on a modern one), and the combined knowledge of all the ins and outs collected during the PC boom by software authors and hardware makers implementing and re-implementing compatible devices.


I've had great success running Win 95 games on modern hardware. I just had to do it in Wine, amusingly enough.


Frustratingly, wine for windows isn't a thing...


Except it is (at least for the use-case of 16-bit apps that are unsupported by a 64-bit Windows) https://github.com/otya128/winevdm although no updates since 2021, but maybe it was "good enough" for whatever they were targeting.


I wanted to play certain games from that era (Spiderweb's Exile series), and the best solution I found was to just play the MacOS versions with SheepShaver.

You can technically get Windows 9x software running in a VM, but not without laggy video/audio in my experience.


> Archivists take note - if you want something to live for a long time, it needs to be easy to emulate

how do archivists have a say in this?


Some archivists make decisions about what to archive. Something that isn't going to be runnable in the future would be a poor choice if you only have limited resources.

Also, some archivists have the choice to convert media. For example, rather than storing a Wordperfect document, perhaps it is best to convert to PDF. Rather than storing the ROM of an 80's arcade machine, or the whole machine, perhaps it is best to store an MPEG video of a playthrough. Rather than storing the data on a floppy disk in a filing cabinet, perhaps it is best to store the data on a server which will be kept up to date? Well resourced archives might be able to implement emulators - but then the question remains how should that be done - Is it okay to have a PDP11 emulator that runs on dos, emulated by dosbox in windows XP, emulated again by virtualbox on Windows 11?

A big part of being an archivist is making decisions of what to keep, what not to keep, what form to keep it in, and when to convert it.

There is no consensus - some archives knowingly keep data and software that they have no way to open/run, in the hope someone might bother in the future. Others keep dependency tables to ensure that they always have some combination of hardware and software to run/open any stored material.


Personally I'm of the opinion that we should focus on storing as many bytes of data of human endeavors as possible, and not worry about emulation/search/cataloging.

Future people will have better solutions to all these problems, and every bit of effort we put into organising our archives today is effort taken away from collecting more bytes.


This means that you care about byte counter instead of actual content.

For some hardware, the number of people who can make it work has already diminished a lot. You can gather some of the knowledge today, “future people” won't be able to. What's the use of collections of data that can't be used?


True - and those are some of the reasons that my opinion is not common in the archiving world.




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

Search: