ANN: Microsoft goes Open Source

D

David Brown

When I wrote it, I thought there was a distinction. BTHOOM what
I was thinking about.


So the attributes are not changed. That is what I was ranting
about. It's a sore point in my little neck of the woods.

I gathered as much from your other posts later on.
Isn't branding equivalent to attributes, as in the copyright?
In our biz, the copyright didn't have the author of the code
because DEC owned it. Is it the custom now to place the
author's attributes in the copyright statement?

Normally code written while employed is "owned" by the company, so the
company name is in the copyrights. The actual author's name may or may
not be included.

Branding is a very separate concept to copyright attribution. Branding
is about what the user sees - on a Red Hat system, the user is left in
no doubt that they are using Red Hat Linux, because the name, logo,
colour scheme, etc., are all over the desktop and documentation. From
the user's viewpoint, they bought the distribution from Red Hat, because
they trust Red Hat to have put together a working package. But
underneath, the copyrights say who owns the code (and normally who wrote
it, modified it, and/or is responsible for maintaining it). Red Hat can
take code written by others (under open source licenses) and put their
own branding on it, but they can't change the copyrights. Similarly,
Cent OS can take Red Hat authored open source code and put their own
branding on it. It would be illegal for them to misuse Red Hat's names
or logos, and risky to come close (they could be in trouble if they
called their distro "Red Cap"), but as long as users are left in no
doubt about where the distro comes from, there's no problem. As another
example, when Microsoft incorporated the BSD TCP/IP stack in early
versions of NT, they had to keep the copyright notices in place (the BSD
license basically says you can do anything you want with the code except
claim you wrote it).
 
W

William Pechter

When I wrote it, I thought there was a distinction. BTHOOM what
I was thinking about.


So the attributes are not changed. That is what I was ranting
about. It's a sore point in my little neck of the woods.



Isn't branding equivalent to attributes, as in the copyright?
In our biz, the copyright didn't have the author of the code
because DEC owned it. Is it the custom now to place the
author's attributes in the copyright statement?

Nope... branding, as far as RH is concerned is trademark use and logo.
They don't even own the copyright on a lot of what they're shipping in
RHEL... Mozilla/Firefox isn't copyright RedHat.

An equivalent to RedHat is that they are basically the editor of a
book where the individual authors hold the copyright for each short
story but they have publication rights and have the copyright on the
compilation as a book. 'Course the GPL means the authors of the
individual works are allowing them free distribution withn the GPL
license language.
Thar be dragons in that process. I sure hope it works out well.

Works out real well. RedHat can sell their nice expensive support
contracts. Those organizations that feel the need for outside support
can purchase them. Those who want to do it alone and take the risk of
not having someone to call can do it that way.

I had RedHat 7.x with their maintenance before the corporate enterprise
changes.
Their NFS pefromance sucked and would cause corruption... I applied all
the updates -- no luck. Took a contemporary Mandrake kernel and the nfs
problem went away. The other boxes involved in testing was a NetApp
filer, a Sun Solaris Sparc, and RH6.2.

I then took the latest standard Linus 2.4.x kernel and built it. No
problems. Seems like RH had a problem specific to their stuff.
Took an hour to rebuild the kernel and a long night of testing.
Problem was resolved without any need for them to patch something for
me.

Not sure what I'd get as far as response on the old retail version...
I also wasn't too sure what their test facility was.

I know that DEC and IBM had the tools for reproducing this type of stuf
in the lab for development. But you had to get a long way up the
support chain to get stuff like this resolved.

Having open source means you can do your own stuff quickly -- if you
have your own Systems Programmer you can roll your own local fix. Much
more like the early days... before closed object only code.
<snip>

/BAH

Bill
 
C

Charlie Gibbs

It is certainly time for Apple to bring QT to Linux. Who cares about
WMV (like everything from that source), it's merely a lock-in play.

It's a non-issue. mplayer can already play them all. And it's
a lot less obnoxious than any other video/audio player I've seen.
 
J

jmfbahciv

Works out real well. RedHat can sell their nice expensive support
contracts. Those organizations that feel the need for outside support
can purchase them. Those who want to do it alone and take the risk of
not having someone to call can do it that way.

Has this been tested in court yet? Has it survived rapacious
destruction attempts? That is going to be the litmus test
of whether it works well or not. Oh, and longevity. None
of these have had that experience yet.

I had RedHat 7.x with their maintenance before the corporate enterprise
changes.
Their NFS pefromance sucked and would cause corruption... I applied all
the updates -- no luck. Took a contemporary Mandrake kernel and the nfs
problem went away. The other boxes involved in testing was a NetApp
filer, a Sun Solaris Sparc, and RH6.2.

I then took the latest standard Linus 2.4.x kernel and built it. No
problems. Seems like RH had a problem specific to their stuff.
Took an hour to rebuild the kernel and a long night of testing.
Problem was resolved without any need for them to patch something for
me.

Not sure what I'd get as far as response on the old retail version...
I also wasn't too sure what their test facility was.

I know that DEC and IBM had the tools for reproducing this type of stuf
in the lab for development. But you had to get a long way up the
support chain to get stuff like this resolved.

Having open source means you can do your own stuff quickly -- if you
have your own Systems Programmer you can roll your own local fix. Much
more like the early days... before closed object only code.

We were always against object-only distributions. It was PITA
to make the distribution tapes and it was stupid to not
allow customers to learn. Our customers were the most important
part of the cycle that made a product as perfect as possible.
Allowing them to have TECO access to the sources also began
that undefinable thingie called customer loyalty. If somebody
saw their code in our sources, we had a customer for life.

I understood why DEC wanted to have strict source control, but
I still think (this is my opinion) we lost more than we
protected by not sending out sources. It was no coincidence
that the time DEC waned coincided with the generation of
sourceless kiddies who had grown up to sign purchase orders.

/BAH
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

No members online now.

Forum statistics

Threads
473,744
Messages
2,569,484
Members
44,903
Latest member
orderPeak8CBDGummies

Latest Threads

Top