On Sun, May 15, 2016 at 6:52 PM, Rugxulo <rugx...@gmail.com> wrote:

> Dennis, I almost hate to bring this type of stuff up. It's almost
> flamebait because nobody can agree. So it's a waste of time.
> Nevertheless ....

I think we are talking past each other, but...

> On Sun, May 15, 2016 at 3:58 PM, dmccunney <dennis.mccun...@gmail.com> wrote:
>> On Sun, May 15, 2016 at 2:58 PM, Jerome E. Shidel Jr. <jer...@shidel.net> 
>> wrote:
>>> I can't imagine anyone taking stuff from a FreeDOS 1.2 release and
>>> *wanting* to issue it as a commercial product.  Rex released 4DOS as
>>> open source because it was no longer selling.  The world had moved on
>>> from MSDOS and 16 bit, and so had he.
>>> It is not an impossibility. For example , the current version of the
>>> commercial product SpinRite runs on a FreeDOS boot CD.
>> What has that to do with anything?
>> Spinrite is and always has been a commercial product.  The vast
>> majority of what ran under DOS back when was commercial.  The fact
>> that it *runs* under FreeDOS is irrelevant.
> It's pretty relevant. Without a "free" DOS, he couldn't (re)distribute
> a bootable CD at all. He'd have to make all his users find a
> compatible DOS elsewhere, which is not as easy as it sounds (anymore).

The key for Gibson was that he could pick up and use the FreeDOS
*binaries* to create a bootable floppy from which Spinrite could be
run without special license or cost.. The fact that he could get
*source* was likely irrelevant.  If a version of MS/PC/DR DOS was
freely available for reuse in binary format without licensing
requirements or cost, it likely would have gotten the nod.

>> It just means FreeDOS is compatible enough with MS/PC/DR DOS that
>> Spinrite *will* work under it.  That level of compatibility was a FreeDOS
>> design goal from the beginning.
> Yes, but compatibility means little if you can't redistribute (or
> easily acquire) the OS. There are many commercial, proprietary DOSes,
> but almost all of them have died (and can't be easily found legally).
> I'm not trying to overhype FreeDOS, but it's literally the only one
> who cares about that. Any one of them could've done it, but they
> didn't.

See above about reasons for using FreeDOS.  I don't argue that.  What
we are specifically discussing is the licenses that will let various
open source programs be distributed as part of an actual FreeDOS

>> And as I recall, Spinrite only uses DOS to load it.  It does not
>> actually use DOS once up and running, and has its own low level code
>> for disk access and testing.
> Great, but "barely uses" still means you have to have a compatible DOS
> ... unless he makes it like old PC booter games (no OS or only uses
> BIOS).

Which it appears Gibson may be doing going forward.

Part of the issue is that the floppy is an endangered species.  My
current desktop doesn't have one.  I have a USB floppy drive I can
plug in, and it's recognized as A: and will work, but I've never had
reason to use it.

Most software these days simply won't fit on a floppy, and gets
distributed on media as a CD or DVD, or as an ISO file that can be
burned to one ot to a USB thumb drive. Machines can be told to boot
from them.  For that matter, when I installed Ubuntu on my desktop in
dual-boot mode with Windows, I migrated the original Win7 Pro
distribution from hard drive to SSD.  Once it was up and running from
SSD, I re-partitioned the SSD drive from Windows to clear a raw slice,
burned the Ubuntu ISO to a bootable USB thumb drive, and rebooted and
ran Ubuntu from the thumb drive. It saw and installed to the raw slice
on the SSD, creating the desired ext4 file system for Ubuntu to live
on and run from.  The end result was a multi-boot configuration using
grub2 giving me a choice of Ubuntu, Windows on the SSD, or Windows on
the HD to boot from.  I've subsequently upgraded from Win7 Pro to
Win10 Pro on the SSD, and my multi-boot menu is Ubuntu, Win10 on SSD,
and Win7 on HD.

Floppies are unlikely to be part of systems Spinrite will be run on
these days, so a bootable floppy containing it will be irrelevant, and
FreeDOS unnecessary.

>> The issue is open source code in a FreeDOS distro being used in a
>> commercial product.
> I hate to nitpick, but please stop using "open source" to mean
> something other than OSI. Yes, it can be misused, and no, they
> probably can't stop you (trademark claims), but it's not beneficial at
> all to pretend that "open source" means just "sources available". Most
> people only refer to "open source" as OSI (or similar free software).

Unfortunately, nitpicking gets required.  There are a variety of
licenses under which source code is offered along with programs built
from the source.  Some of the licenses are incompatible with each
other, which means code you see in one open source project may be
something you can't *use* in yours, because the license under which it
is issued won't permit it.

This sort of thing makes my irony meter peg offscale, since the whole
point to open source in the beginning was to be *able* to re-use code
from other sources instead of rolling your own.

(As a Horrible Example, Gnu Public License v2 and Gnu Public License
v3 are incompatible with each other.  You can't use v3 code in a v2
project without re-licensing the project as v3.)

Another project I follow is Toybox.  This is an attempt by former
Busybox maintainer Rob Landley to create a replacement for Busybox.
The Google Android developers are enthusiastic supporters.  Android
uses an inferior product called Toolbox, but as Toybox commands reach
a stable point, the Android development infrastructure maintainer is
replacing the Toolbox versions with the Toybox builds.

Rob used to be a GPL supporter and recanted.  These days, he refers to
Gnu efforts as "Gnu/damnit", and doesn't want the GPL anywhere near
his code. (Toybox uses a BSD license.)  When Toybox reaches a 1.0
release, it should be possible to create a working Linux CLI system
where only the Linux kernel is GPLed.  I'm all in favor.

>> That may not be impossible, but it's so unlikely
>> that whether the particular open source license freely allows such
>> usage is something I wouldn't waste a moment worrying about.
> It's not unlikely or they wouldn't have bothered making such restrictions.

Highly unlikely, in the stated case.

>> As a rule, if you wish to incorporate open source code into a commercial
>> product, you are expected to get clearance from the author (and likely
>> pay a fee for the right to do so.)
> Not at all. Who told you that? You're pretty uninformed here. "Open
> source" always means able to use without charge. The term was designed
> to be business friendly so that they could hire developers (if needed)
> to improve existing code bases, similar to (but broader than) GPL.
> Even GPL was designed more to sell future development as a service
> instead of perpetual royalties just to use a single-user license of
> proprietary crud that can't be changed.

We *are* talking past each other.

I believe I mentioned earlier that there is no restriction in open
source licenses about picking up and using the code, and you can do so
without charge *as long as you, too, will provide source* to what you

There is also no restriction I have seen in open source licenses about
whether you can *sell* your product for money as a commercial product.

But if you *do* plan to sell your resulting product for money, you
almost certainly intend to make your product *closed source*, to be
*able* to charge for it.  People simply don't *buy* code freely
available in source form.

If you want to do *that*, you are expected (and likely legally
required) to contact the original author and negotiate a closed source
license for the code.  You *cannot* pick up open source code, make a
commercial product from it, and sell it *without* also making your
code open source under the same license, unless you get the permission
of the original author to do so (and getting such permission usually
involves paying a fee.)

There are exceptions.  One open source product I follow and use is
SQLite, an SQL compliant database implemented as a single library
file.  It's *widely* used (like, the Firefox browser keeps
bookmarks/history and other things in SQLite DBs), but the developers
don't know *how* widely, because it is explicitly *public domain*,
with no requirement you even mention you use it.

I have seen *no* open source projects where the *open source* version
is offered for sale as a commercial product.  I have seen some mixed
cases where an *older* version is available as open source, but the
current version is closed source commercial you must pay for.  An
assortment of open source projects provide a way to make a donation if
you use and like the code.  I am aware of *one* open source project
(Kovid Goyal's Calibre ebook database tool) that makes enough in
donations that the author makes his living doing it.  (The fact that
Kovid lives in India, with far lower living costs, is what likely
makes that possible.)

The fundamental problem facing open source code is "How do you get
*paid* for writing it?"  The general answer is "You get a job with
someone like Google or Facebook that uses a lot of open source code,
and pays engineers to hack on what they use."  Most folks contributing
to open source efforts are doing so as an unpaid sideline, and if they
write code for a living it's for closed source proprietary efforts.

>> If the idea is that only code
>> issued under an open source license that *doesn't* require you to
>> contact the author about commercial usage should be included in the
>> FreeDOS 1.2 distro, that's a profoundly silly notion.
> Silly? Aren't you friends with Eric Raymond? He's a very big "open
> source" (OSI) proponent. Heck, he co-founded OSI!

I've known Eric since before he became famous.  We met something like
30 years ago.  I strongly suspect I know more about his views on the
topic than you do.

> OSI was meant to 'promote open source ideas on "pragmatic,
> business-case grounds." '. And business obviously means money, but
> that doesn't mean paying (over and over again) for frozen software.
> I realize that there's still lots of proprietary software, and not
> everyone agrees with OSI or FSF. But there is a heavy push towards
> business-friendly "open source" / "free software". It's just easier
> for developers (and those who are willing to pay people to improve
> public software).

Going back to cases, what prompted this discussion was Rex Conn's open
source license for 4DOS, which indicated his source code couldn't be
used in a *commercial* product without contacting him.  There was a
question about including it with that restriction. I don't see that as
unreasonable, and it's actually standard practice in most cases.  The
implicit assumption is that a commercial offering will be closed
source, and you must contact the author for permission to use it that
way.  And I would be flatly astonished if anyone ever *did* contact
Rex about using the 16 bit code he released as open source in a
commercial product.

For that matter, I strongly suspect there are license
incompatibilities between stuff currently offered with FreeDOS,  in
the sense that you may not be able to lift source from one project and
use it an another with a different license.

Everything issued as part of a FreeDOS distro should be open source,
offered under licenses that permit providing the source along with the
binaries.  Whether any of the sources may be incorporated in a
commercial product offered for sale will be governed by the specific
license under which the source is offered.  The same will be true for
whether any of the sources can be used in other projects offered under
a different license.  It should not be a factor in whether its offered
in a FreeDOS distribution.

Mobile security can be enabling, not merely restricting. Employees who
bring their own devices (BYOD) to work are irked by the imposition of MDM
restrictions. Mobile Device Manager Plus allows you to control only the
apps on BYO-devices by containerizing them, leaving personal data untouched!
Freedos-user mailing list

Reply via email to