OSDev.org

Kevin wrote:Hm, then the versions I downloaded must be seriously stripped down...
DexOS: No network; well, let's call it a basic GUI; basic CLI; seems to actually have an assembler if you use the CLI - but where's the editor?

I told you you need the old ver
http://www.dex-os.com/old/index.htm

http://www.dex-os.com/old/editor.htm
http://www.dex-os.com/old/DexServ.htm

And basic as in not a desktop GUI

This works in two ways: the graphical front-end is more like a video game launcher than a traditional OS, and programmers can easily access the bare metal for maximum performance. It's an intriguing concept, and the take up of Android and iOS in the last couple of years has shown that traditional desktop computing metaphors are starting to look long in the tooth.

guyfawkes wrote:Example see here what BareMetalOS is useful for:
http://www.youtube.com/watch?v=uEMwSsra1EU

With respect, ca. 10000 primes in 10 secs on a 1.5GHz dual core? Awesome how far BareMetalOS got, but you shouldn't link this video as a pro, it's more like a contra. You are doing it wrong.

guyfawkes wrote:And basic as in not a desktop GUI

Two buttons, one of which launches the CLI, and the other one a menu with five entries that start small demos. How would that be more than basic?

I downloaded the old version and at the first sight it looks a bit more interesting indeed. Still can't find an editor on the image I downloaded. And I don't see much more of the network functionality than the displayed IP address in the GUI and a "Networking" menu entry that doesn't seem to do anything when I select it.

Insert something about ASM developers and their genetic need to brag about their work.

My kernel is 100% assembly too.

Kevin wrote:

guyfawkes wrote:And basic as in not a desktop GUI

Two buttons, one of which launches the CLI, and the other one a menu with five entries that start small demos. How would that be more than basic?

I downloaded the old version and at the first sight it looks a bit more interesting indeed. Still can't find an editor on the image I downloaded. And I don't see much more of the network functionality than the displayed IP address in the GUI and a "Networking" menu entry that doesn't seem to do anything when I select it.

Oh look, the wii menu is just a couple of buttons, oh no the i-pads menu is the same.
Whats this the new win8 is the same
http://i.i.com.com/cnwk.1d/i/tim/2011/1 ... t-menu.png

And in dexos, just type ide <enter>

There are plenty of successful operating systems (hobby and non hobby) written in ASM and HLL languages. I just dont see where the OP is getting at... sounds awfully biased.

neon wrote:There are plenty of successful operating systems (hobby and non hobby) written in ASM and HLL languages. I just dont see where the OP is getting at... sounds awfully biased.

My point is yes there is well written C and ASM OS's, but the way rdos seem to be picked on in this post
http://forum.osdev.org/viewtopic.php?f= ... 91&start=0
Which was then locked, just because he dared to say you could make a 100% ASM OS

One of the main point that was put up as to why you should not code a 100% ASM was long term maintainability.
Well i have post example of long term ASM OS's that are still being maintained after many years.

At the same time most hobby C OS are abandoned or rewritten for the 100th time.

Its not that ASM coders brag about their work, it more it takes years and years to get noticed.

I would argue that any project without commerical sponsorship is a hobby OS project. Therefore, by that margin, we can include in the list of HL OSes

• FreeDOS (Most of the kernel and command.com are C; other portions various, often C or ASM)
• ReactOS
• Syllable (which is a fork of the dead AtheOS RDOS mentioned above)
• Haiku

Which amount to:

• An MS-DOS clone (a non-trivial project being as nigh on every DOS application uses some "undocumented" functionality)
• A Windows clone (which has similar issues to the above w.r.t.. compatibility with buggy apps)
• A somewhat POSIXy but otherwise rather different OS. (e.g. no VTYs for a start)
• A clone of BeOS, itself a somewhat POSIXy but otherwise different OS (in a similar way to Haiku)

I'm sure you can come up with other projects which are similarly advanced...

I think its possible to say that once a C hobby OS gets sufficiently advanced, it has a higher probability of attracting developers than an ASM OS (being as there are far more developers who know ASM than C) and it finds it easier to gather momentum. At that point, people seem to stop thinking of them as a "hobby OS" in the same way they do smaller projects.

One of the main point that was put up as to why you should not code a 100% ASM was long term maintainability

All assembly language code is doomed to unusability in the long run by nature due to its tie to a particular architecture. The only types of software that can survive this transition is those written in HLLs. My argument against 100% OS's isnt maintainability, its compatibility.

Also...many hobby OSs written in C rarely need rewrites.

The reason UNIX came to power is that it was the first (IIRC, although there's probably some research system I'm missing) OS written entirely in a portable HLL, so that it could be ported easily to the wide variety of architectures present at the time. The value of portability is huge.

Wider use and commercialization doesn't have much to do with the quality or language used. We all know that many people used Windows in it's earlier versions when it was no more than a beta, and that MSDOS was a hack that should never have been released. The acceptance of Linux was just luck, good timing, and the fact it claimed Unix compability. Linux, or Windows, are not creative OSes, but old architecture (Linux) and commersial junk (Windows).

And to be able to port toolchains has no priority to me. I could provide the complete OpenWatcom toolchain for RDOS, but I don't have a valid reason for why to do it.

guyfawkes wrote:One of the main point that was put up as to why you should not code a 100% ASM was long term maintainability.
Well i have post example of long term ASM OS's that are still being maintained after many years.

And I guess my main point was that ASM OSes are kept small because otherwise they would lose the rest of their maintainability.

I'm not necessarily saying that this is bad, they just have smaller goals with respect to functionality and try to optimise other things, like performance. Which of course results in what I saw yesterday when I tried some of the OSes you listed: You get more small (albeit nice) demos than useful applications.

What I don't understand is what is "more successful" about these OSes. I knew Kolibri before and of course it's a great achievement. But most of the rest of your list didn't really impress me much. Not that they are bad, but I would call them "nice" at best. Maybe it's in a different way, but it's about the same level of unimpressiveness as if you take the average C OS.

Kevin wrote: And I guess my main point was that ASM OSes are kept small because otherwise they would lose the rest of their maintainability.

RDOS currently has 195,000 source-lines (290,000 including comments / headers / empty lines) ASM code in the kernel, and 20,000 lines of C++ code in the user-level classlibrary which was written specifically for RDOS, and a set of ported toolsets (JPEG, PNG, LIBMAD, ZLIB, XML). That is not exactly a small project, which proves your statement about maintainability is wrong.

Owen wrote:I would argue that any project without commerical sponsorship is a hobby OS project. Therefore, by that margin, we can include in the list of HL OSes

• FreeDOS (Most of the kernel and command.com are C; other portions various, often C or ASM)
• ReactOS
• Syllable (which is a fork of the dead AtheOS RDOS mentioned above)
• Haiku

These OS's have had commercial sponsorship, eg: dell (freedos), wine (ReactOS) etc.

Also OSDev is the main forum/wiki for hobby OS development, all the above ASM OS's i named there main developers are members of this forum, How many of the developers of the OS's you posted are members of this forum ?.

As for small demos, yes they are small that's the main reason we use ASM.
We code our apps from scratch, not badly port them.
All the ASM OS's are all different from each other, other that the supper cool Freedos, the rest of the C examples are just linux or BeOS clones
Example:
Haiku code base is a mix BeOS, homebuilt code or derivatives of existing open source software .

Syllable Desktop is an original, modern operating system design, in the tradition of the Amiga and BeOS, but built using many parts from the GNU project and Linux.

Syllable Server is a small and efficient Linux operating system.

guyfawkes wrote:These OS's have had commercial sponsorship, eg: dell (freedos), wine (ReactOS) etc.

Wine is an open source project itself and certainly not a sponsor. I also doubt that Dell sponsors FreeDOS.

But even if they did: Does the success of a hobby OS vanish when it's become interesting enough that a company invests in it?

Also OSDev is the main forum/wiki for hobby OS development, all the above ASM OS's i named there main developers are members of this forum, How many of the developers of the OS's you posted are members of this forum ?.

Believe it or not: There is OS development outside this forum. Really.

As for small demos, yes they are small that's the main reason we use ASM.
We code our apps from scratch, not badly port them.

And that's fine. I just disagree that it automaticaly makes these OSes more impressive or at least interesting and "successful".