Hacker News new | past | comments | ask | show | jobs | submit login
MenuetOS: An entire operating system in x86_64 assembly (menuetos.net)
127 points by gorenb on Oct 29, 2023 | hide | past | favorite | 45 comments



2009 might be earliest HN commentary, but MenuetOS dates back to at least 2005 according to the Internet Archive.

https://web.archive.org/web/20051215234548if_/http://www.co....

https://web.archive.org/web/20051030022655if_/http://www.men...


Kinda weird to use IA for this instead of just looking it up on Wikipedia

    Initial release: May 16, 2000; 23 years ago (32-bit)


I don't see anything "kinda weird" about that.


It is weird to anyone who didn't grow up being told to not cite Wikipedia as a source of information in any authoritative capacity, being told to instead dig for and cite the original source instead which a tool like Internet Archive is useful for.

Culture shock stemming from a generation gap, I guess.

(Obviously this distinction is not important for most comments here on Hacker News which are just written in passing, but if this was an academical setting you would, presumably and hopefully, get laughed out of the room.)


I disagree with the sentiment, the IA is a very useful source, but, as shown here, it provided the wrong answer. In general, you can't trust it to tell "how old is something", only to show that it is "older than".


"Since at least 2005" is not a wrong answer. It's imprecise, but perfectly accurate. And since there is no oracle of absolute truth to consult, it's not like it's wrong to seek out empirical answers like this, it's just one fairly general approach to find a new upper bound. Of course, in this case there is a better answer, but still, there's nothing weird or wrong about the approach.


I mean.. it took more work and produced a sub-optimal answer?


More effort has been made to argue that it's a bad approach than the actual effort it takes to go to archive.org, enter a URL, and see the first capture date. That all to say, it's not any appreciable amount of effort. It's really at worst a similar amount of effort to searching Wikipedia for an answer.


I think the "5 * 9 is at least 40" from Rick and Morty is relevant here : https://m.youtube.com/watch?v=IxqcYJEXr6I


I am not commenting on the accuracy or value of the information provided by Internet Archive, I am speaking concerning the phenomenon that not citing Wikipedia is "weird".


Wikipedia may not be a citable source itself, but it can still be a good place to find links to other sources more reputable.


Major life hack. You may not be allowed to cite Wikipedia, but there's nothing stopping you from using wikipedias citations.

Also the whole "don't use Wikipedia" thing is idiotic. There's a bunch of tech illiterate teachers who think Wikipedia is full of misinformation, and impress that view on students. So lots of people think Wikipedia is useless, while in reality it's a fantastic source of information and you can easily find the sources of that information.



Earlier than that: I remember booting it on my Pentium 90 before 2003


Mandatory mention to KolibriOS[0], an open-source fork of MenuetOS.

0. http://kolibrios.org/en/



mom said its my turn to repost it!


Why did they go closed-source for the 64-bit version?


> Menuet64 is closed source because M32 was forked and new copyrights just slapped at the beginning of practically all 32bit source files (multitasking, drivers, GUI, networking..) without any permission. You can take a guess about the country in question.

https://board.flatassembler.net/topic.php?t=22194


How does that make sense? That has always been and will always be possible with any open source project. Should every other project go closed source too just because Russia can fork it and say "I made this"?


Yeah, that excuse makes perfect sense if you don't think about it.


Closing down the source was likely an emotional, rather than a rational reaction. The author simply got pissed and didn't want this to happen again.


It makes sense that the fact you are making a big deal of it while you didn’t get to say anything about apple macos since it was never open source in the first place.


MacOS and MenuetOS aren't equivalent, not even up to isomorphism.

Menuet is far from being a practical OS. It's a very interesting proof of concept/research project, but its usefulness evaporates almost completely without the source.


That seems like an awfully petty reason. (and does the forum censor the word "Russia" or something?)


> You can take a guess about the country in question.

For the curious, it appears to have been done by people in Russia (state-sponsored or not is unclear)


It would be ridiculous for a state to sponsor a toy OS. I think it's absurd.


No proof whether it's state-sponsored or not, and this is called out in the thread. I thought I had carefully chosen my word when I said "people in Russia", where not every people in that country is working for the state ;)


ReactOS?


I wouldn't classify ReactOS as "toy", and I believe Russia stopped sponsoring it many years ago.


They must have at least one customer where this makes sense.


Is the IntelHDA audio driver only available for 64-bit MenuetOS? According to the release notes, it was first included in version 0.96X, which seems to be 64-bit.

The developer has only made one 32-bit version (0.86) availabe, though. Would be great to have past releases available as well.

EDIT: Apparently the 32-bit versions have not come too far in the last 12+ years -- in 2011 (when the Intel HDA driver was released), version 0.85 of M32 was downloadable: https://web.archive.org/web/20110615210928/http://www.menuet...

I'm learning some assembly (on FreeDOS), so MenuetOS is interesting. Was hoping to test it (+ the HDA audio, for 24-bit playback -- and maybe try to make sense of the audio driver source, for a hobby project) on a Dell Mini 9, which is 32 bit. It's time to move on then, I guess.


Intel HDA audio driver is available for Menuet64. M64 also has under 1 ms audio latency, or even under 0.1 ms audio latency. Part of the latency setup is adjusting the scheduler interval up to 0.1 Mhz.


Thanks! ("Time to move on" = mmmaybe put those 32-bit machines aside, ohwell; I'm long time aware and still curious about MenuetOS, also as a huge fan of minimalist systems. The Dell Mini 9 is a really cool fanless netbook, though, esp for the ~5€ I paid for it, ha. Would be great to test MenuetOS with audio on it as well. That M64 latency is surely impressive.)


Menuet (and Kolibri) are most likely the only OSes written in asm which are still under active development until today...



Ah ARM. Nice.


Is there any examples of practical uses of this OS or is this more of a passion project?


Check under "screenshots".

Could be just the vehicle someone needed to get into x86 coding on bare metal.

And serves as a nice counterpoint for "optimizing compilers outperform skilled assembly coders". Regardless of whether that's true or not.


I wonder if using this makes the most sense from a "security by obscurity" standpoint. It doesn't seem likely that blackhats would spend a lot of time hacking into such a marginally used OS.


It only takes one.


there are better fringe OS:es if you care about security.

QubesOS for example

https://www.qubes-os.org/


This is at least 24 years old.


The news page indicates a new release on 2023-10-26, which is impressive diligence.


Menuet & Kolibri are still actively maintained though.




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: