Amiga OS 3.1 Source code leaked yesterday

[Knoeki]

it is even more sad to see quite some of the remaining active users happily participate in distributing this unauthorised copy, compromising years of development work and all possible future versions of AmigaOS and therefore biting the hand that feeds them against all common sense and in blatant violation of copyright.

How is this biting the hand that feeds? I think the people interested in the 3.1 sources couldn't give a flying fuck about OS4.

If anything, these sources might help improve OS3.x and make it something better, and I think that's what's Hyperion is afraid of, that OS4 becomes even more irrelevant than it already is. (:

[absence]

Originally Posted by Olaf Barthel

Both Cloanto and Hyperion only have licenses to use certain material, but the licenser alone can contribute historic code to the Computer History Museum.

Hyperion is only a licensee, but some copyrights were apparently transferred to Cloanto ("As the holders of the Commodore/Amiga copyrights [...]"). If Cloanto owns the copyright to the source code for Amiga OS 3.1 and earlier, they're the ones who can open source it, assuming it doesn't violate existing contracts, e.g. with Hyperion. That's where lawyers come in.

[Thalion]

Originally Posted by Knoeki

How is this biting the hand that feeds? I think the people interested in the 3.1 sources couldn't give a flying fuck about OS4.

If anything, these sources might help improve OS3.x and make it something better, and I think that's what's Hyperion is afraid of, that OS4 becomes even more irrelevant than it already is. (:

As i can understand that they are not happy with the release of the source code, i think it is even more sad that Hyperion wants to give the community a guilty conscience now. I think Hyperion could be very happy that there is such a good and active community which still supports its old Amiga Hobby.

[absence]

As was mentioned earlier in this thread, Hyperion is possibly obligated by their Amiga OS licensing contract to react.

[copse]

The date on the files in web file directory listings for this leak, are 2013. That's the archive itself as posted to the web. I have no idea what the dates are inside the archive, as I've not obtained it to look.

Let's also be clear about how easy it was to find. It was named 'amiga os source code 3.1.tar.bz2'.

Simple logic dictates that if they really cared they could have googled and found it, as many of us did. Much much earlier than this conference tweet "leak" announcement. But no, the it must be protected argument is weak because it provably has not been. That the web file directory in question has only now been taken offline in reaction to this tweet, is irrelevant to the long term previous easily discovered availability.

[Olaf Barthel]

Originally Posted by copse

The date on the files in web file directory listings for this leak, are 2013. That's the archive itself as posted to the web. I have no idea what the dates are inside the archive, as I've not obtained it to look.

Let's also be clear about how easy it was to find. It was named 'amiga os source code 3.1.tar.bz2'.

Simple logic dictates that if they really cared they could have googled and found it, as many of us did. Much much earlier than this conference tweet "leak" announcement. But no, the it must be protected argument is weak because it provably has not been. That the web file directory in question has only now been taken offline in reaction to this tweet, is irrelevant to the long term previous easily discovered availability.

Ahem. Please do not assume that the people involved were negligent or otherwise unfit to attend to the task at hand.

When the leak occured, the contents of the archive had to be assessed: what had leaked, had the AmigaOS4 source code repository been compromised?.

Figuring out if the archive contents had proliferated was not straightforward, e.g. the data was repackaged and published under different names on other servers (and that did not include the various bittorrent seeders).

Then the archive contents wound up on GitHub at possible the worst time, on New Year's eve when it turned out nobody was "minding the store" and would take down the material. It took a week for that to happen.

A press release had to be prepared, too. Nobody likes to read them, but somebody has to make the best effort to explain the situation and the consequences. A thankless task, but still necessary.

The thing is, it is complicated, challenging and on more than one level unfair what fate throws at you at times. Best of luck to you that you may never have to face a challenge such as this.

[Knoeki]

Considering I'm not in the business of pushing a niche OS for niche hardware, I value my chances of this never happening quite high.

[copse]

Originally Posted by Olaf Barthel

Ahem. Please do not assume that the people involved were negligent or otherwise unfit to attend to the task at hand.

When the leak occured, the contents of the archive had to be assessed: what had leaked, had the AmigaOS4 source code repository been compromised?.

You misunderstand me. I am not talking about the reactive defense of the Amiga property, as you are. I am talking solely about a simple demonstrated lack of interest in proactively defending the property.

The archive I found on the day of the twitter "leak" was in an open web directory, with a modification date of 2013. It was trivially locatible via google, given it's filename. I'm not the only one who noted this.

It should be very achievable to set up a google alert, or some similar thing to notify of the appearance of archives named similarly to 'amiga os source code 3.1.tar.bz2'. Or even to google it once in 3 years. It certainly got taken down fast after the twitter announcement.

[UberFreak]

All this noise about ancient source code, when an actual HDF image with OS4.1FE classic pre-installed is easily found.

The reality is, almost everything is out there, can't really fight it.

[gulliver]

I went digging thru the archive and found the following BINARY/EXECUTABLE unreleased versions of AmigaOS disk files that are NEWER than what went out with AmigaOS 3.1:

Filename, Size, Version, and Date

utilities/more 12752 40.3 (2.2.94)
utilities/installer 106644 1.27 (24.10.93)
utilities/clock 13164 42.1 (4.2.94)
utilities/cdaudio 15528 42.4 (8.3.94)
prefs/cdprefs 10624 42.3 (21.3.94)
libs/player.library 11128 40.14 (22.2.94)
libs/locale.library 17272 42.1 (23.8.93)
libs/iffparse.library 5852 42.2 (16.8.93)
libs/commodities.library 8236 42.1 (12.7.93)
libs/asl.library 38416 42.2 (15.7.93)
libs/68040.library 43808 42.1 (26.7.93)
l/CrossDOSFileSystem 27640 40.24 (21.10.93)
devs/printer.device 27940 41.10 (8.2.94)
classes/images/led.image 2924 42.1 (24.2.94)
classes/images/label.image 4484 42.1 (12.4.94)
classes/gadgets/tabs.gadget 6008 42.3 (23.2.94)
classes/gadgets/button.gadget 4868 42.2 (7.3.94)
classes/datatypes/pcx.datatype 4160 42.1 (31.7.93)
classes/datatypes/macpaint.datatype 2532 42.1 (2.8.93)
classes/datatypes/ico.datatype 2024 42.1 (30.7.93)
c/wait 976 42.1 (12.7.93)
c/sort 2280 42.1 (9.8.93)
c/protect 1504 42.2 (14.9.93)
c/list 6484 42.7 (14.9.93)
c/iprefs 11876 42.1 (11.8.93)
c/binddrivers 1248 42.1 (27.8.93)

I did a quick test, and they all seem to work quite nicely

[xArtx]

Originally Posted by Knoeki

How is this biting the hand that feeds?

A company is never the hand that feeds you,
it’s the other way round, unless the company is actually donating food!

I wouldn’t have expected Cloanto to be concerned about the release of source code
that can’t be built again with modern day computers.
I’d only have expected them to be concerned about compiled ROMs.

That’s enough for me to remove myself from the discussion.
Mike has done me a favour in the past without anything to gain (or lose)
for himself, and I really appreciated it.

It’s difficult to maintain the moral/ethical high ground when you are on the wrong side of the law,
as unfortunate as the situation is.

My point of view about the Amiga might be different because I think the Amiga ended with the CD32,
and in order to properly Amiga, a large group of users must all have the same, or very similar hardware.

[absence]

Originally Posted by xArtx

It’s difficult to maintain the moral/ethical high ground when you are on the wrong side of the law,
as unfortunate as the situation is.

Please don't equate law with moral, ethics, or even justice! There are plenty of laws that are immoral, unethical, or unjust, and there are many ways to be immoral, unethical, or unjust without breaking the law.

[xArtx]

You’re right there, but still in this case, it will only really work out if something is given away rather than taken,
and it looks like that won’t happen. At the start of all this, I was kinda hoping none would care :

[Olaf Barthel]

Originally Posted by absence

Please don't equate law with moral, ethics, or even justice! There are plenty of laws that are immoral, unethical, or unjust, and there are many ways to be immoral, unethical, or unjust without breaking the law.

I always wanted to use this quote, now seems to be the right time: "To live outside the law, you must be honest."

Originally Posted by copse

You misunderstand me. I am not talking about the reactive defense of the Amiga property, as you are. I am talking solely about a simple demonstrated lack of interest in proactively defending the property.

The archive I found on the day of the twitter "leak" was in an open web directory, with a modification date of 2013. It was trivially locatible via google, given it's filename. I'm not the only one who noted this.

It should be very achievable to set up a google alert, or some similar thing to notify of the appearance of archives named similarly to 'amiga os source code 3.1.tar.bz2'. Or even to google it once in 3 years. It certainly got taken down fast after the twitter announcement.

Good point, I'll give it a try.

However, post-hoc it seems to be clear that the pattern to look for would have been "amiga os source code", but after the leak that pattern may no longer yield useful information.

[funK]

Hey, what's going on in here?
IIRC OS3.1 sources were already leaked back in late 1990s/early 2000s: I seem to remember that even laire supposedly used those leaked sources to write parts of his infant MorphOS, back then.

[copse]

Originally Posted by funK

Hey, what's going on in here?
IIRC OS3.1 sources were already leaked back in late 1990s/early 2000s: I seem to remember that even laire supposedly used those leaked sources to write parts of his infant MorphOS, back then.

Perhaps the difference was in how public the leak was, and how widely available was the source code made. In this case, worldwide on the most common media (twitter and github). A world of difference, one might say, without even knowing much about the first leak.

Personally, I think the word leak is a common bad choice. The source code was leaked decades ago, and this was just more recent publicity. A bit like scientists seeing canals on Mars in the 90's and newspapers on a slow news day in 2015 going "THERE ARE CANALS ON MARS" because they never came across the information before.

[Gzegzolka]

I have one question. We all thinks how those code could be used to improve wb 3.1, but what about wb 3.5 and wb 3.9 ? How much they are updated and compatible?

Wrong, the day that the Archive went up I contacted one of the Admins of Github who is a close personal friend. I told him about it, informed him that it may be an issue. I was told they had to wait for a takedown order. The day it was taken down was the day that they were informed formally.

Originally Posted by Olaf Barthel

Ahem. Please do not assume that the people involved were negligent or otherwise unfit to attend to the task at hand.

When the leak occured, the contents of the archive had to be assessed: what had leaked, had the AmigaOS4 source code repository been compromised?.

Figuring out if the archive contents had proliferated was not straightforward, e.g. the data was repackaged and published under different names on other servers (and that did not include the various bittorrent seeders).

Then the archive contents wound up on GitHub at possible the worst time, on New Year's eve when it turned out nobody was "minding the store" and would take down the material. It took a week for that to happen.

A press release had to be prepared, too. Nobody likes to read them, but somebody has to make the best effort to explain the situation and the consequences. A thankless task, but still necessary.

The thing is, it is complicated, challenging and on more than one level unfair what fate throws at you at times. Best of luck to you that you may never have to face a challenge such as this.

[Olaf Barthel]

Originally Posted by RobertJDohnert

Wrong, the day that the Archive went up I contacted one of the Admins of Github who is a close personal friend. I told him about it, informed him that it may be an issue. I was told they had to wait for a takedown order. The day it was taken down was the day that they were informed formally.

I am sorry, there is more to it, and I stand by my statement.

The day the repository went up (Thursday, December 31st 2015) a DMCA takedown notice was submitted to GitHub. That takedown notice was not processed until Monday, January 4th 2015, but not executed. The notice came back with a request to resubmit it.

This was frustrating, as you can imagine.

[Olaf Barthel]

Originally Posted by Gzegzolka

I have one question. We all thinks how those code could be used to improve wb 3.1, but what about wb 3.5 and wb 3.9 ? How much they are updated and compatible?

The 3.5 and 3.9 updates incorporate enhancements and additional bug fixes suggested by some of the material then available from the Commodore archive.