Hacker News new | past | comments | ask | show | jobs | submit login

> Under what license? Can we view, modify, redistribute?

Just speaking for myself, I come down pretty firmly in the "Don't know, don't care" camp for both questions.

In a world where FreeDOS exists, there just isn't any practical value to having the source code to MS-DOS 2.0 available under a truly open license. That code's only interesting as a historical exhibit. If all we can do is look at it and go, "Huh, so that's how that worked", that's fine with me.




The important question, I think, is: can the FreeDOS folks read the MS-DOS source, and then continue to work on FreeDOS?


Is this a real judicial concern or just a philosophical question? I read code on a daily basis with all kinds of different licenses and I also contribute to open and closed source on a daily. Should I go make an appointment with my lawyer tomorrow?


Some companies have license agreements that forbid reverse engineering their software if you have used it / worked on it. It may be even worse if you have actually looked at the source code of the closed source alternatives.

INAL, But if you are contributing to projects on tangential areas you should be OK.


It's safer if they don't.

Unless Microsoft licenses it under a license with the most explicit and extensive protections against patent lawsuits.


Utility patents only last 20 years. MS-DOS 2.0 was published in 1983, which is 31 years ago. So, while it's copyrighted and you can't redistribute the code or the compiled binary, you can go to town with the algorithms, should you want to.

As a bonus, the availability of this code might make prior art challenges to extant patents much easier.


Is MS-DOS even patented ? As far as I remember, software patents were not granted until fairly recently in History (usually from the 90s).


Oldest patent I see that looks like it could be for MS-DOS is #4,779,187, "Method and operating system for executing programs in a multi-mode microprocessor", which was issued in October 1988. That wouldn't apply to DOS 2.0, of course.

The bigger concern would be copyright. That might be a fairly plausible argument for why FreeDOS developers shouldn't be reading this source. Though realistically the odds that Microsoft would actually pursue such a case are remote at best.


I would call the odds "beyond remote."

The chance that you are going to get MSDOS expressive structures in FreeDOS in a way that is even arguably beyond deminimis is nil. You don't usually just move code between implementations easily, and chances are pretty good that the FreeDOS folks use fairly different approaches than the MSDOS folks. The argument is that if you read the code you might be tainted, but I don't see this as a significant argument here for software that's 30 years old.


Issue date is not the only relevant date - patents get extensions and expand in scope. I would imagine the FAT patent that Microsoft uses to get some revenue from every SD card sold comes from the dawn of personal computing.


> patents get extensions and expand in scope

Yeah, but usually you cannot get extensions THAT easily, and extensions are limited in time. I don't think you can expect extensions to last too long either (at most you can add a couple of years of exclusivity).


Most of the FAT patents of the '90's finally expired in November.

Now exFAT is much more recently encumbered.


I couldn't find the answer but did come across this interesting historical take on software patents.

http://www.nytimes.com/1994/02/24/business/microsoft-loses-c...

> The Patents and Trademark Office, which began issuing software patents only a decade ago, has been accused of having inadequate expertise in software to judge what is truly novel.

Nice to know nothing has changed in a decade.


Two decades, even.


I think they could solve all their problems by just hiring a handful of geeks and OSS enthusiast's like myself to lay down some ban hammers.

While thinking about it a little more deeply let's give them the benefit of doubt for five minutes and assume they did hire a bunch of geeks to do this. Is there any possibility the problem simply can't be solved easily? What seems complex and patentable today may in fact seem completely obvious and novel in ten years when it actually shows up in court or even just 5 years from now once it finally gets reviewed and accepted.

In searching for prior art, is it actually possible to go back in time on the wayback machine and check for these types of obvious issues with the claim?

Something to think about anyway. Disclosure: kiwi here. We did some form of reform already, havtn t noticed anything adverse thus far.

Would be interested to hear if anyone has come across problems yet.


It doesn't really matter. MS has plenty of new patents they could use to tie you up in the court system for years. Fighting that would probably bankrupt you.




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

Search: