DOS ain't dead

I could use advice from the collective wisdom...

Is it admissible to publish modifications, fixes and the like for widely accessible utilities such as, e.g. the old PC Magazine collection which is available with sources often from a load of sites. The question comes from the programs being copyrighted by the publisher (not even the actual authors) despite being open source.

In the event I wished to make privately made fixes available to the community at large, what of the following ways could be (im)proper in your opinions :

- publish fixed binaries along with modified sources (including the original) ?

- publish modified source only (in diff form) ?

- other ?

- are answers different in case of small bug fixes, versus large enhancements, modifications, or derived work ?

I'm assuming the publisher (Ziff Communication Co., if even it still exists) can't practically be contacted over such queries.

---
Ninho

Publish modified binaries + modified sources.

---
DOS-u-akbar!

Hi Laaca.
> Publish modified binaries + modified sources.

Is it OK with respect to program sources carrying, and executables printing copyright statements such as (example) :

copyright db "(c) 1990, 1991 Ziff Communications Co.",10,13
author db "PC Magazine ",254," Douglas Boling"

---
Ninho

> Hi Laaca.
> > Publish modified binaries + modified sources.
>
> Is it OK with respect to program sources carrying, and executables printing
> copyright statements such as (example) :
>
> copyright db "(c) 1990, 1991 Ziff Communications Co.",10,13
> author db "PC Magazine ",254," Douglas Boling"

I don't know the status of Ziff Davis, and I'm no lawyer. I know they forbid redistribution since a long time (re: Garbo). I'm sure Arjay knows more about this. Anyways, I think they actually used to sell (a year or two ago) it online still as a subscription or bundle. Yeah, maybe I'm remembering wrong, can't remember the URL, who knows.

In any case, you're probably only safe NOT publishing binaries but instead binary patches (debug scripts) and/or non-context diffs, IMHO.

http://en.wikipedia.org/wiki/Ziff_Davis
http://en.wikipedia.org/wiki/PC_Magazine

>>
>>
>> copyright db "(c) 1990, 1991 Ziff Communications Co.",10,13
>> author db "PC Magazine ",254," Douglas Boling"

> I don't know the status of Ziff Davis, and I'm no lawyer. I know they
> forbid redistribution since a long time (re:

Yeah, I heard that too. OTOH many many repositories carry their utilities,
apparently without fear of reprisals, and I mean serious looking places,
not the typical "war3z" house :=)


> Garbo). I'm sure
> Arjay knows more about this. Anyways, I think they actually used to sell (a
> year or two ago) it online still as a subscription or bundle. Yeah, maybe
> I'm remembering wrong, can't remember the URL, who knows.

It used to be by subscription I think. I don' know, either, if Ziff or some
successor entity still actively tries to sell those old proggies.
Considering the state of DOS in 2011, it could be a costly operation for
little benefit ;=)


>
> In any case, you're probably only safe NOT publishing binaries but instead
> binary patches (debug scripts) and/or non-context diffs, IMHO.

I'm grateful for your advice. To be sure I don't misunderstand, what do
you call "non-context" diffs, please ?

---
Ninho

> >> copyright db "(c) 1990, 1991 Ziff Communications Co.",10,13
> >> author db "PC Magazine ",254," Douglas Boling"
>
> > I don't know the status of Ziff Davis, and I'm no lawyer. I know they
> > forbid redistribution since a long time (re:
>
> Yeah, I heard that too. OTOH many many repositories carry their utilities,
> apparently without fear of reprisals, and I mean serious looking places,
> not the typical "war3z" house :=)

Well, most people aren't legally minded (nor me), which is why a lot of corners get cut. We shouldn't have to be so strict with each other, IMHO.

> > Garbo).

BTW, make sure to read this link, though I'm not sure it's very definitive anymore. But it'll give you some background history, at least.

> It used to be by subscription I think. I don' know, either, if Ziff or
> some successor entity still actively tries to sell those old proggies.
> Considering the state of DOS in 2011, it could be a costly operation for
> little benefit ;=)

Yet people do still sell DOS software in plenty of places (e.g. Gog.com), just not as a primary source of income, usually.

> > In any case, you're probably only safe NOT publishing binaries but
> instead
> > binary patches (debug scripts) and/or non-context diffs, IMHO.
>
> I'm grateful for your advice.

It's just random advice, not necessarily the best.

> To be sure I don't misunderstand, what do
> you call "non-context" diffs, please ?

"diff -c" is context (preferred by *BSD ??), "diff -u" is unified (preferred by GNU). Traditionally you'd use "diff" without either option for "normal" output, but it makes it harder if you have slight changes in the original or try merging several changes at once. (Actually, diff predates Larry Wall's patch by quite a few years. Originally I think it outputted a script for ed.) I mean, honestly, I don't personally think context diffs should be copyright infringement, but to be ultra conservative and cautious (since some people are extremely litigious), I would avoid it anyways.

I don't think that other diffs are not copyright infringement if context diffs are. Doesn't every diff format contain some of the original information? (Haven't looked at them and compared them very well now though.) And even if there was one that didn't, wouldn't merely new (added) source code contain enough implied information to be an infringement? Depending on the usage of the program's internal interfaces and such.

---
l

> Is it admissible to publish modifications, fixes and the like for widely
> accessible utilities such as, e.g. the old PC Magazine collection which is
> available with sources often from a load of sites. The question comes from
> the programs being copyrighted by the publisher (not even the actual
> authors) despite being open source.

Copyright issues are for lawyers, programming is for programmers. Put somewhere the original source code, your patches and the binaries you created, and if someone authorized will ask you to remove them, you simply remove them.

---
Glory to God for all things

> Copyright issues are for lawyers, programming is for programmers. Put
> somewhere the original source code, your patches and the binaries you
> created, and if someone authorized will ask you to remove them, you simply
> remove them.

100% agree

---
DOS gives me freedom to unlimited HW access.

>>>> "(c) 1990, 1991 Ziff Communications Co."

>>> they forbid redistribution since a long time

> Well, most people aren't legally minded (nor me), which is why a lot of
> corners get cut. We shouldn't have to be so strict with each other, IMHO.

Totally agreed...but as you said, I nor you aren't, God forbid! /Lawyers/

>>> Garbo).

> BTW, make sure to read this link, though I'm not sure it's very definitive
> anymore. But it'll give you some background history, at least.
>

Very instructive, thanks. So Wasa Univ had got explicit permission, which was
later removed unilaterally (and rather rudely). Hmmm! Unless that prior
permission had explicitly written words that it could be retired at the
permitter's sole initiative, ISTM & IANAL ;=) that a gift is a gift, full
stop; or, after a French legal dictum, "donner et retenir ne vaut", you
can't give and then take back. Whatever, Garbo's and UWasa's prudence are
understandable...


> Yet people do still sell DOS software in plenty of places (e.g. Gog.com),
> just not as a primary source of income, usually.

Yes, but let's face it, utilities from the old collections - while many of
them may still be usefull, are often obsolete and want to be revised,
which Ziff or whoever won't care to do. They /can't/ just cash on the lot
without doing anything else, at least morally. There should be a
peremption date like on a bottle of milk :=)

.....

> "diff -c" is context (preferred by *BSD ??), "diff -u" is unified

Oh, right, O.K...

Supposing smallish diffs, I would try to argue they constitute NO infringement,
along these lines : (M.Justice, Your Honor...) in order to make
effective use of the provided "differences", a user would have to get a
(hopefully) legal copy of the original program, which I do not provide.
For tehnical reasons, the "differences" do contain limited exctracts or
quotationsfrom the original work, that do not exceed fair use etc. etc.

What do you (including CM) think of the reasoning ? ISTM much preferable
to give hints in source form, when we have the original source, than
binary patches (that evoke the devil of "reverse engineering" + could be
suspected of malevolence)...



> (preferred by GNU). Traditionally you'd use "diff" without either option
> for "normal" output, but it makes it harder if you have slight changes in
> the original or try merging several changes at once. (Actually, diff
> predates Larry Wall's patch by quite a few years. Originally I think it
> outputted a script for ed.) I mean, honestly, I don't personally think
> context diffs should be copyright infringement, but to be ultra
> conservative and cautious (since some people are extremely litigious), I
> would avoid it anyways.

Thanks again, Rugxulo, plenty for thought

---
Ninho

> I don't think that other diffs are not copyright infringement if context
> diffs are. Doesn't every diff format contain some of the original
> information? (Haven't looked at them and compared them very well now
> though.) And even if there was one that didn't, wouldn't merely new (added)
> source code contain enough implied information to be an infringement?
> Depending on the usage of the program's internal interfaces and such.

Might it not depend upon how much has to be quoted from the original work?
Doesn't "fair use" authorize limited extracts for illustration purpose ?

Practical example off my head : to allow programs compiled by one of those "batch
compilers" to return the last errorlevel, as it should, we have to
minimally change the compiler from : mov ax, 4C00h  int 21h ,
to smthg like: mov ah,4Ch  mov al,err_rc  int 21h ,
plus hopefully modify a greeting message to indicate version changed.

Diffs such as this example can hardly be said to include enough original
content to infringe on copyright. I would be careful to mention that a
user's right to use the diffs is conditioned on his right to use the
original Pc magazine utility.

It's a real pain in the you-know-where that a full binary can't be
provided. I hate the idea of losing my time in making robust patchers
and/or debug scripts & the like. Pfff!

---
Ninho

If you are still reticent, IMO the best idea would be to ask the copyright holder if they consider your changes to be under "fair use" or they consider them copyright infringements and consequently will bring you into justice in case you publish them. Finally their view should be the most important for you.

---
Glory to God for all things

> If you are still reticent, IMO the best idea would be to ask the copyright
> holder if they consider your changes to be under "fair use" or they
> consider them copyright infringements and consequently will bring you into
> justice in case you publish them. Finally their view should be the most
> important for you.

You believe I can determine without any doubt who the real copyright holders are as of April,
2011? Track them, find somebody with authority to make a legally binding
statement on their behalf? Successfully get in contact with the responsible
person, possibly have to argue, negotiate,whatever... with any chance of getting
a binding statement of consent ?

I'm naive indeed, but there are limits even to my credulity...

What I will do is try to track original programmers and get their
opinions, but it's not a given I will find them or get answers, and in
except in cases where the stated program author actually holds the rights,
what I expect to get from them are just that, valued opinions. Like yours
here...

> > Copyright issues are for lawyers, programming is for programmers. Put
> > somewhere the original source code, your patches and the binaries you
> > created, and if someone authorized will ask you to remove them, you
> simply
> > remove them.
>
> 100% agree

Agree in principle with you and Khrushaw, but I'm collecting opinions and
trying to determine the best format - less contentious while still usable by non
experts. And also who will host the contents (does bttr-software have an
interest in such, for instance?)

> You believe I can determine without any doubt who the real copyright
> holders are as of April,
> 2011? Track them, find somebody with authority to make a legally binding
> statement on their behalf? Successfully get in contact with the
> responsible
> person, possibly have to argue, negotiate,whatever... with any chance of
> getting
> a binding statement of consent ?

If you really care about copyright issues and want to be legally safe in an absolute way, you have to determine and ask.

> What I will do is try to track original programmers and get their
> opinions, but it's not a given I will find them or get answers, and in
> except in cases where the stated program author actually holds the rights,

As far as I understood the copyright is not hold by the programmers, but by a certain company. Or am I wrong?

---
Glory to God for all things

> If you really care about copyright issues and want to be legally safe in an
> absolute way, you have to determine and ask.

Absolute ? I have no hopes for any absolutes in this sublunar world.

> As far as I understood the copyright is not hold by the programmers, but by
> a certain company. Or am I wrong?

I'm sure you're right. But I don't know what that company is (the Ziff group was dismantled, its assets shared), nor willing to go to hassles to get in touch with it (or its lawyers...). I don't intend to make money or get any kind of gratification out of my whatever work, correspondingly I'm not going to deploy considerable efforts, even less real money just to try and get a formal agreement (probably hopeless anyway).

---
Ninho

> Absolute ? I have no hopes for any absolutes in this sublunar world.

For absolute absolutes perhaps there are no hopes, but for relative absolutes there are .

> I'm sure you're right. But I don't know what that company is (the Ziff
> group was dismantled, its assets shared), nor willing to go to hassles to
> get in touch with it (or its lawyers...). I don't intend to make money or
> get any kind of gratification out of my whatever work, correspondingly I'm
> not going to deploy considerable efforts, even less real money just
> to try and get a formal agreement (probably hopeless anyway).

I understand, but in case you will change you mind, I think that Ziff Davis, Inc are those, Vivek Shah being now the owner.

---
Glory to God for all things

> I don't know the status of Ziff Davis, and I'm no lawyer. I know they
> forbid redistribution since a long time (re:
> Garbo).
Correct.

> I'm sure Arjay knows more about this.
Not a great deal more beyond what I vaguely remember from Ziff Davis's legal follow ups as at time I was running a legal BBS but had copies of the simtel archive locally (Walnut creek CD-ROM's). I was never approached by them but there was a lot of chatter on Simtel/over the BBS's at the time about it.

> Anyways, I think they actually used to sell (a
> year or two ago) it online still as a subscription or bundle. Yeah, maybe
> I'm remembering wrong, can't remember the URL, who knows.
Unfortunately you are remembering this all correctly, e.g: please see the over zealous boiler plate on PC Mag's subscription page for the 22 year old ANSI.COM.

Some further background here and here based on my own understanding of the removal requests from my BBS days. See also
Curtiss Priest's excellent Buried software utilities in Simtel and 'pcmag' archives the top of which touches on their leaking onto the "net" but importantly is also an informed well written/structured document on the utils.

Fortunately there are some sensible authors/writers who were connected with PC Magazine, e.g. Ethan Winer's Ethan Winer's page will be of interest for basic/asm programmers.

> In any case, you're probably only safe NOT publishing binaries but instead
> binary patches (debug scripts) and/or non-context diffs, IMHO.
IANAL, however I would recommend seeing boiler plate on download link above.