[5th Draft] Open Letter to CNRI: Request for clarification

[Tim Peters]

[Stuart Ford]
> Any change in the license that changes our ability to release
> products using Python will set all my hard work back to stage
> one, plus make me look the fool for recomending the language
> in the first place.

[phil <dawks at tesco.net]
> Exactly.  I even asked this question in this group and was
> assured that this was not a problem.

[Tim]
> Well, "this group" hasn't even seen the new license yet, and even
> if they had they didn't write it. ...

[phil]
> I asked this question over a year ago.

Ah, that would explain why I couldn't find your msg.  There's still nothing
to stop you from using the 1.5.2 release (which is what you were probably
asking about, yes?) in any way you please.

> I don't really understand why you're so agressive about it,

Don't know what you mean here.  I assumed that, like Stuart, you had real
current concerns about the ability to continue shipping products under the
new CNRI license.  I hope my reply was useful to Stuart, then.

> unless you know something i don't?  (which, having read quite a
> number of your posts i'm positive you do).

A lot about the history and the personalities, sure, but *everything* I know
about the current state of the license negotiations I've posted right here.

> There's a website and a newsgroup; i don't share a watercooler.

Actually, neither does PythonLabs <wink>.

> I added Python as a cool feature for my customers (i thought it was
> rather a nice capability and spent a _lot_ of my spare time
> implementing it) . It's a total irrelevance as far as marketing is
> concerned, and to be honest, i think i now agree with them (for
> different reasons). The sad thing is that i have to explain to my
> boss why i was such a wanker for recommending it to our customers,
> but i can manage.

Sorry, that lost me.  Why do you suddenly feel you're "a wanker"?  We don't
yet know whether the new CNRI license will make *any* difference to how
Python can be used.  BeOpen PythonLabs is pushing to make "no difference"
the reality.  If we thought that was a lost cause, we would have stopped
negotiating by now.

But while the OSI can credibly pronounce on the license's Open Source
status, and the FSF (via RMS) can do so for its GPL-compatibility status,
people using Python in proprietary ways are going to have to judge the
license for themselves.  This was also true of the CWI licence, the major
difference being that Guido could speak for the *intent* of that license but
cannot speak for CNRI's intents.  That doesn't mean their intents are evil!
It means you have to ask CNRI.  People on c.l.py are used to having Guido be
the final word on everything, and I repeat the bit about asking CNRI because
the fact that Guido is not the last word on this particular issue appears to
be hard to get across.

> Now that Stallman is involved and the idiocy has begun,

I think you misunderstand RMS's goal here badly:  part of being compatible
with the GPL is that the CNRI license cannot contain clauses that restrict
you *more* than the GPL does.  The effect of RMS's suggested changes is to
*reduce* the number of requirements imposed on you by the license.  If this
is idiocy, the world needs a whole lot more idiots <0.5 wink>.  This is
again no different than what was done in the past for the CWI license,
except that RMS had no problems with that license the way it was.

> i'm out of here and will solve the problem with VBScript (and it can
> be solved that way, it was my other less interesting alternative).
>
> I don't run my business on Python and it seems clear that anybody
> would be a fool to do so.

Your choice, but I urge everyone who isn't a self-confessed wanker <wink> to
hang on to see how the license issue turns out.  I've got no reason to
believe it will turn out badly for any Python user -- and I don't believe
you do either.

if-you-think-ms-licenses-don't-change-across-releases-you-
    haven't-been-reading-them-ly y'rs  - tim