An interview with two of the developers:

http://www.goodgearguide.com.au/article/315421/we_talk_devel...

http://en.wikipedia.org/wiki/KolibriOS deserves a mention too, GPL fork of MenuetOS.

That's not just a GUI -- it looks like it supports alpha transparency. That's certainly more than I expected out of a floppy disk.

You can do a heck of a lot in 1.44MB. The original Mac had a complete GUI plus some applications running from a 400K floppy, plus a 64K ROM.

People forget.

Well, I assume they're usiing straight-forward, direct-to-memory framebuffer manipulation. No X protocol to get in the way, no unavailable GPU drivers. Simple alpha transparency shouldn't be hard to do. When did Porter and Duff write their paper, '83?

Adding alpha transparency for a framebuffer based GUI need not take more than a few hundred bytes.

A few hundred bytes is precious space on a floppy.

Interesting and impressive.

The 64 bit floppy image runs fine under VirtualBox if you select a 64 bit OS.

Virtual Box? It's great that people can try it out and see what it's capable of, but doesn't that sort of defeat the point? I mean, if its real claim to fame is that it's written in assembler than it seems that running it inside a C (is VB C?) application on top of a C operating system would seem to miss its best qualities.

Now if I could only find a floppy around here...

Well if you want to try it out for a few minutes, as I guess most people on here will, it beats rebooting your PC (or figuring out how to even make it boot on a MacBook). It's not like it's full featured enough for daily use anyway.

I can't really see your point to be honest.

What's a floppy?

Something that's not to be copied, IIRC.

D-D-Do-Do-Don't...Don't copy.. Don't copy that floppy!

Linkage: http://www.youtube.com/watch?v=XWf_jbrpn4o

Not Ready Reading Unit A: (A)bort? (R)etry? (F)ail?

I have to admit, that's some of the friendliest looking assembly source code I've seen. http://www.menuetos.net/e64.asm Granted, I haven't looked at assembly in over 15 years...

Decent -- nothing special (if you consider decent as being good code written by smart people). FASM was an interesting choice and I'd like to see their justification for that.

I think this kind of projects are good reminders that modern computers have incredible amounts of processing power and storage capacity. If only all that power could be harnessed efficiently....

I also think that legacy cruft is hindering development in computers. One reason why (GNU/)Linux was able to advance so quickly was that it had much less legacy than other systems, allowing more freedom for developers to do stuff better way.

Reading the changelog belies the complexity of that project... I would dearly love to have the time to take a look at the code for it.

Well done!

Is this belie definition 2, "To tell lies about, esp. to calumniate by false statements", definition 4, "To give a false representation or account of, to misrepresent; to present in a false character", definition 6, "To call (a thing) false practically, to treat it as false by speaking or acting at variance with it; to be false or faithless to", or definition 7, "To show to be false, prove false or mistaken; to falsify (expectations, etc.)"? I don't see how any of those makes sense, given that it's an action by some person being said to belie an attribute of something. (Maybe that would make sense if you were one of the developers and reading the changelog demonstrated that the project's complexity were somehow "false"?)

(OED 1ed. vol. 1 p.782, 806 of 1270 in the PDF)

"belie" can also just mean "to be in contradiction with".

Which I assume was the definition being used. As in, "the simplicity of the changelog is in contradiction with the actual complexity of the project".

In that case, definition 4 should fit the bill. Reading the changelog gives a "false representation" on the complexity of the project.

All of these senses have the meaning, loosely, "to be in contradiction with"; my confusion was how any particular one of them applied, and which one. In a way, senses 4 and 7 are opposites: in one case it's the belier that's false, and in the other sense it's the thing belied.

Thanks for that, yes that is indeed the way I am using the word. A bit old-school perhaps...

I would rather see a hand-optimized version of L4, packaged with a kernel for Debian.

Or an OS written in Forth.

"ForthOS is a complete, standalone operating system for the PC. It includes command line, compiler, debugger, editor, and filesystem."

http://www.forthos.org/

Implement a Forth on the OS. If the rest of the OS is in ASM anyway, you can do high-level work in Forth.

Eugen, I have a copy of F-83 if you want it.

Thanks, I used to hack Forth a bit back in my Amiga days.

I still have my Novix 4000 board I bought used somewhere.

I remember similar projects many years ago: - QNX - SqueakOS - SqueakNOS

Linux, ColorForth, GEM, Kontiki, Oberon, the Macintosh System, Microsoft Standalone BASIC, Microsoft Windows, etcetcetc.

Amazing. The hardest part probably was to find a floppy drive to read / write the floppy. I don't think I have installed one on my last 2 rigs.

So does having all the labels in UPPERCASE save that much space?

Wow it's improved since I tried it about a year ago

Which one is more active, KolibriOS or MenuetOS?

Cool, but can I run Erlang on it?

Seemed like a good idea when they first got it out. It seems now like it was soo long ago.

Maybe it is interesting for embedded devices, but seems they've lost their purpose. It was never meant for business, but again, what's the point?

You're reading this on a site called "Hacker News" and you have to ask that? How about: the challenge.

There is a difference between a challenge that is of use for this world, and a challenge just because.

I don't see technological, logical, or knowledge progress in the attempt to write complete OS in assembly (anymore). I've done it myself, in high school, just for fun, a 16bit. That is also when MenuetOS came out. Now I spend my time more wisely. I've actually went on doing things that benefit the society the technology and the users.

Not that menuetOS is bad, but I suggest they reorient them selves so that there is that usable part. And I can see space for it, in embedding computing f.i. Far from having no respect for their work, which I have my hands in from day one.

So, essentially your point is that because you have already done something others should be spending their time more productive ?

To some the challenge really is all there is. I'm very happy to see some counterculture to the bloatware that is the norm.

This is an amazingly compact and neat hack, it incorporates a very fast and lightweight GUI that can be used for general purpose computers just as easily as for embedded systems.

Hmm. I don't think I am making my self very clear. I think it is a great challenge, and I know what it takes to do so.

However, as Novash says - a reason: "that was faster", "couldn't find a good one in this language", "friendly enough", that seems to be missing. It was there long time ago when they started (for those of you who remember), but today everything shifted. I don't go building theories anymore on the statement that the world is flat?

So stop attacking me because I gave an opposing opinion and suggestion for improvement. It is great what they do, but I am sure like any other programmer, they too would like their code and idea used, not forgotten.

This was not about productivity or stupidity of projects, so reread my comments before you burst. I also do stuff all the time just to see if it can be done. Don't all of us? But again, "if it can be done" solves a problem. Redoing same things just because, now that's something worth discussing.

ok I get it now. Sorry for the misunderstanding!

Fair enough, and good point.

I do stuff 'just because' all the time.

I coded a program to parse a text file into a OpenCalc Spreadsheet just because someone asked me if it could be done.

I coded a program to solve Sudoku because a friend made one and challenged me to make one that was faster (his was).

I am still coding a whole MUD Engine in C# simply because I couldn't find a good one in this language and nor a friendly enough one in any other language I know. It is moving slowly since I want it to be generic enough so others can use it.

I still, every now and then, solve problems on Valladolid.

I can see your point that they could spend their time more wisely, but I beg to differ. A lot of people love to code and coding becomes the purpose on itself. The knowledge they are gaining WILL prove to be invaluable to them and to others and the fact that they are doing something because they want, not because they must will drive them much farther than if it was a commercial project. It will give them both the reasons and the will to hone it to perfection. I suggest you to read "Jonathan Livingston Seagull", if you haven't.

It does not mean that every challenge that you take on must be of some use to other people. You're forgetting the fun aspect of doing such things. The same would be said to those who create their own graphics engines, compilers, and heck, even Lolcode. Sure, they don't become the next "big" thing in our industry, but the sense of accomplishment of pulling off a project of that magnitude is valuable enough in of itself.