I actualy saw the source code for MS-DOS (IBM in the early days shipped PC-DOS branded version and non IBM kit could buy MS-DOS). I can't recall if it was version 4.x or 5.x (think it was 4.x) and the only thing that struck me was the use of some Sun boot loading code without any visable (c) signs in comments apart from soemthing along the lines "This is from Sun for boot loader to work...".
But history does have its ability to make things unclear over time so can see why people argue about 2000 year old books, let alone software in our own lifetimes.
Only thing I feel sad about is that OS/2 was not open sourced 10 years ago.
Still in another 10 years or so people will be writting stories about how Linux was a rip of from Minix and other historical things without any hope of changing the past.
You may very well be right, though I would of expected some (c) signs in the source (assembler). Sadly I don't have access to those sources any more, so I won't be able to verify it. Also more I think about it the more I believe it was 4.x I saw and that with what you pointed out about Sun may very well be the case, especialy given the additions to volume ID'd (more than one disc) being added for that version.
One day the source may officialy be publicly available and that will be a interesting day. That said it does appear as if the 6.x source code was leaked and illegaly available from a quick google, but that's another story.
DOS was (is) a loose collection of small (optional) utilities, a tiny bootloader (IO.SYS) and a limited command interpreter -- command.com and the bootloader was all that was strictly necessary: hardly a "heart". Memory management was lousy (and bizarre) and usually since I was using PharLap memory management it was largely redundant anyhow. You could use any number of thrid party memory managers or even roll your own. Disk caching/buffering was just awful and it wasn't until version 5 that it started shaping up. V5 was actually a very good version, after years of rewrites and improvement.
Microsoft bought QDOS in good faith and are therefore not in the frame for anything. For once in my life I'd defend them here. I don't even know why Gates's name is even being connected to this story.
And yes, Gary Kidall was a genius. But in this one instance, MS are free of any wrongdoing in my view.
This is just an advertorial for CodeSuite. Srsly, there was "no way" to compare them until now, we read, then further down the page we read his methodology: download the easily available source code to both from the web, and search them for common strings.
My guess: “Comerr” was a routine that handles either communication errors or command errors
When this guy encounters "foo" and "bar" his head will explode.
I expect better from IEEE, this is not even at the level of TechCrunch!
My thoughts were "if you have to look up what 'jnz' means, you probably aren't qualified to do this analysis, much less come to the definite conclusion that no copying was done"
It appears that he looked for string similarity in the source and binaries.
Another, and imho better, way to find code plagiarism is to build the call graph of the programs and perform graph isomoprhism to detect similarities. It is much harder to conceal plagiarism from that sort of analysis.
I was never under the impression that MS-DOS (QDOS) was literally stolen from DRI, but that QDOS was a clone of CP/M back in an era when software copyright wasn't clear. Microsoft licensed QDOS but didn't buy it outright, and later made a deal with DRI to cover its ass. Both the original developer of QDOS and DRI ended up with the rights to sell DOS products. (As did IBM.)
Years later, DRI was selling its own version of DOS (DR-DOS) which was widely considered technically superior to MS-DOS or PC-DOS (e.g. At one point it supported preemptive multitasking and could also run Windows 3.x iirc.)
The real "theft" was that when the original developer of QDOS went bankrupt, its mos valuable asset was its DOS license. But when it tried to sell it Microsoft sued and forced the owner to sell the license to Microsoft for a very low figure. Meanwhile, DR-DOS was killed using a combination of deliberate incompatibilities and eventually Windows 95.
I never used DR-DOS, but I remember greatly preferring PC-DOS to MS-DOS because the default text editor was so much nicer.
The fact the article makes zero mention of DR-DOS is a bit puzzling.
13 comments
[ 4.7 ms ] story [ 46.3 ms ] threadBut history does have its ability to make things unclear over time so can see why people argue about 2000 year old books, let alone software in our own lifetimes.
Only thing I feel sad about is that OS/2 was not open sourced 10 years ago.
Still in another 10 years or so people will be writting stories about how Linux was a rip of from Minix and other historical things without any hope of changing the past.
RIP Gary Kildall.
http://en.wikipedia.org/wiki/Sun386i
One day the source may officialy be publicly available and that will be a interesting day. That said it does appear as if the 6.x source code was leaked and illegaly available from a quick google, but that's another story.
EDIT: keep forgetting HN comments aren't Markdown.
Microsoft bought QDOS in good faith and are therefore not in the frame for anything. For once in my life I'd defend them here. I don't even know why Gates's name is even being connected to this story.
And yes, Gary Kidall was a genius. But in this one instance, MS are free of any wrongdoing in my view.
My guess: “Comerr” was a routine that handles either communication errors or command errors
When this guy encounters "foo" and "bar" his head will explode.
I expect better from IEEE, this is not even at the level of TechCrunch!
man strings # strings, man!
Years later, DRI was selling its own version of DOS (DR-DOS) which was widely considered technically superior to MS-DOS or PC-DOS (e.g. At one point it supported preemptive multitasking and could also run Windows 3.x iirc.)
The real "theft" was that when the original developer of QDOS went bankrupt, its mos valuable asset was its DOS license. But when it tried to sell it Microsoft sued and forced the owner to sell the license to Microsoft for a very low figure. Meanwhile, DR-DOS was killed using a combination of deliberate incompatibilities and eventually Windows 95.
I never used DR-DOS, but I remember greatly preferring PC-DOS to MS-DOS because the default text editor was so much nicer.
The fact the article makes zero mention of DR-DOS is a bit puzzling.
http://en.wikipedia.org/wiki/DR-DOS