From: John Thingstad on
On Tue, 13 Dec 2005 22:38:14 +0100, <alex.gman(a)gmail.com> wrote:

> I wrote this as an email reply to R. Fateman, but I now feel like
> posting it. (BTW, Richard, you probably meant to post your message in
> the first place):
>
>
> All I can say is that "we will ask you for a %-age of your future
> profit once you have your product ready for shipping" is a total deal
> breaker.
>
> I don't want to hire lawyers NOW to negotate with Franz when I'm just
> developing the product, and I don't want to get f*** (pardon my french)
> by Franz if I delay the negotiations until later. Score 1 for other
> alternatives.
>
> If I'm being unclear, image this situation: I'm Franz. You have naively
> developed a product using ACL, Allegro Cache, and other things that
> only I (Franz) can provide.
> You have two choices: pay whatever I ask you, or abandon the product
> (major rewrite included). As a business, I (Franz) will be inclined to
> ask you for just a little less than the cost of a major rewrite.
>
> Now if you are smart, you'll never get into this situation by either
> (1) Never using Franz (LW, CMUCL) or (2) negotating the distribution
> terms in advance (laywers, etc.). For most people (1) seems like a
> better alternative.
>

This is the response I got from Franz when asking about prising..

In addition, Franz Inc. offers two different types of Commercial licenses,
based on how an application is used and
deployed. Stand-Alone Application Systems or Client/Server Systems only
used internally, require a Corporate End User
License. And all applications distributed to customers require a Value
Added Reseller License. How do you plan on
distributing your application? How much do you charge for your
application? Does your application contain compilation
capability? Approximately how many applications will you distribute
annually?
To give you an idea of the royalties, these ones vary between $150 and
$800 for the Corporate End-User license, and
between 4% and 10% for the Value Adder Reseller license.

Hardly a question of being clueless about what to expect.

--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
From: John Thingstad on
On Wed, 14 Dec 2005 05:14:46 +0100, Pisin Bootvong <joesb.coe9(a)gmail.com>
wrote:

> I thought that if I bought MSVC++ and sell a product call
> "start-up-vc.bat" which does nothing but start MSVC. And if I sells it
> for 1$ and distributed MSVC++ with them, I can still get sued from MS.
> And yet MS didn't charge MSVC per distributed app.
>

This is a bit off topic but MS-VC++ compiler tools are avaliable for
free from microsoft and have been for a time now.
It is the IDE they charge for.

--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
From: Richard Fateman on
Pisin Bootvong wrote:

>
> And I can be sure Franz did not violate any of them because I never see
> their source code, right?

And if you did see the source code, you would know? Of course you would
not know then either! But presumably Franz stands behind their product and
claims (and will defend the claim that) it owns Allegro Common Lisp.

> And nobody that uses Franz in commercial also use CL-PPCRE, CL-PDF,
> SLIME?

If you are selling a product that relies on BSD code, you are in fact
assuming a certain risk. If you are selling a product that relies on
GPL code, you are assuming a certain risk, as well as an obligation,
in many circumstances, of making your own code free and open source.
As a business model this has some deficiencies.

And if you use Franz with those BSD/GPL licensed library then
> you don't have to be worried?

>
>
>>the business says, "Oh, about that Tahiti trip, who's your maintenance
>>person on 24-hour call." you say "Hey man, does a blackberry work in
>>Tahiti?"
>>
>
>
> This is quite nonsense. Your client never know, when a problem comes,
> whether it is your application's problem or the underlying Lisp's
> problem. So if a customer have a problem with a product (Your product
> X, not ACL or any Lisp) and cannot contact the you because you went to
> Tahiti, that is not a problem of using open-source Lisp, it is a
> problem of choosing the wrong shop.
Well, when I ask for help from a vendor of Windows-based software, I
am often told that I should mess around with the operating system, get
a download of something, etc. The vendors rely on Microsoft maintaining
a certain presence. We may argue that MS doesn't do everything we want
it to do, but it usually doesn't require downloading fresh source code
for the whole operating system, editing a makefile, and recompiling everything.
That's the kind of advice I get if cygwin is broken.


> Are you saying that if reddit use
> ACL then the developer can go to Tahiti, not taking any call, and Franz
> will support any reddit issue for them?

I believe that Franz works very closely with some customers, and in some
cases may know more about the application than some of the application
programmers. I suppose it depends on the kind of revenue stream that
the application represents. I doubt that a one-time $29.95 shrink-wrap
license fee for an infinite number of redistributed binaries would
get much help, but for $100 per copy (X many copies) it makes sense for
Franz to be helpful.
>
>

>
>
> I thought ACL have something like EXE maker. And that it shakes the
> image and unused function from the distributed image. I never know that
> when ACL build a distributed EXE, it includes all the IDE inside.

If your image uses all of Lisp, that doesn't help. For example, you
might use the IDE, or eval. Or you might use the compiler (e.g. for
dynamic CLOS).
>
> I thought that if I bought MSVC++ and sell a product call
> "start-up-vc.bat" which does nothing but start MSVC. And if I sells it
> for 1$ and distributed MSVC++ with them, I can still get sued from MS.
> And yet MS didn't charge MSVC per distributed app.
>
> To some extent, you are implying that any language that have "eval"
> should be licensed per package sold. As a matter of fact, if a
> interpreter can be written in a language such language implementation
> should be licensed per package sold, too. Right?

If you omitted eval from a copy of Allegro, but then
you wrote an eval that had hooks to every function in Allegro, then
not much could be left out of that Allegro. Just the main line of maybe
a page or two of one program, eval.
>
> If you are afraid of your customer cheat in such a way, why not put it
> in EULA? Get a lawyer to indicate what it means to sell product that,
> in some way, allow one to use devlopment features of ACL. And indicate
> that only such product require per-package distribution fees (well,
> disregarding those Allegro special feature that should reasonably be
> charged per package if used, like AllegroCache).

If there is such an EULA, (I don't know),
I think it would have to be a very elaborate license agreement
to cover every contingency; so elaborate that a potential purchaser
would have difficulty figuring out if it
was right, and maybe would have to re-write particular parts. Maybe
even hire a lawyer :( .

If the current technique is for Franz to try to "partner" with the
application writer, then that cooperation may be at least as useful as the proposal
above. But presumably people at Franz read your note and this one
and can comment.

> You don't have to be afraid that they won't follow the EULA, since if
> they are not willing to, I think they willl cheat in all other ways
> anyway.

There are ways of auditing the sale of products. While MS might not
catch you making an extra copy of Word on your laptop, you can be sure
that if a large organization, even a university, buys software,
(or sells software) it cannot use piracy for long. (Unless it is in
a country that does not respect patents, copyrights, etc.)

So far as I can tell, the pricing model you are suggesting looks
pretty close to what is already in place.

It seems that we are hypothesizing the existence of
a creative lisp programmer who has the following characteristics:
1. he/she avoids lawyers.
2. and yet is willing to assume unknown legal risks to use "free" software;
3. and probably has no financial backing (since investors
(a) have lawyers
(b) do not want to assume unknown legal risks. )
4. so has no money to buy a commercial Lisp.

So Franz is not losing money on this person. :)

More seriously, finding a good way to "not give away the whole thing"
seems to be a key. Maybe distribute a fasl file. Run it on a trial version.

RJF




>
From: Pisin Bootvong on
Richard Fateman wrote:
> Pisin Bootvong wrote:
>
> >
> > And I can be sure Franz did not violate any of them because I never see
> > their source code, right?
>
> And if you did see the source code, you would know? Of course you would
> not know then either! But presumably Franz stands behind their product and
> claims (and will defend the claim that) it owns Allegro Common Lisp.
>

Why would that make me more confident to sign the agreement that my
product, that contains Franz, instead of SBCL, doesn't violate anyone's
copyright? I wouldn't sign any agreement that some product I didn't
write myself doesn't violate any copyright, be it from SBCL, ACL or
Microsoft, would you? And how likely do you think that Linux will
collapse in the next 5-10 year of copyright violation?

> > And nobody that uses Franz in commercial also use CL-PPCRE, CL-PDF,
> > SLIME?
>
> If you are selling a product that relies on BSD code, you are in fact
> assuming a certain risk. If you are selling a product that relies on
> GPL code, you are assuming a certain risk, as well as an obligation,
> in many circumstances, of making your own code free and open source.
> As a business model this has some deficiencies.
>
> And if you use Franz with those BSD/GPL licensed library then
> > you don't have to be worried?
>
> >
> >
> >>the business says, "Oh, about that Tahiti trip, who's your maintenance
> >>person on 24-hour call." you say "Hey man, does a blackberry work in
> >>Tahiti?"
> >>
> >
> >
> > This is quite nonsense. Your client never know, when a problem comes,
> > whether it is your application's problem or the underlying Lisp's
> > problem. So if a customer have a problem with a product (Your product
> > X, not ACL or any Lisp) and cannot contact the you because you went to
> > Tahiti, that is not a problem of using open-source Lisp, it is a
> > problem of choosing the wrong shop.
> Well, when I ask for help from a vendor of Windows-based software, I
> am often told that I should mess around with the operating system, get
> a download of something, etc. The vendors rely on Microsoft maintaining
> a certain presence. We may argue that MS doesn't do everything we want
> it to do, but it usually doesn't require downloading fresh source code
> for the whole operating system, editing a makefile, and recompiling everything.
> That's the kind of advice I get if cygwin is broken.
>

Right, but you still get such advice and support from the product X
first. You don't go directly to MS support and ask why Some X software
doesn't work, do you? Your Tahiti situation was that the product X
vendor is not willing to provide any support and direction. Even if
that irresponsible vendor use ACL, I still wouldn't expect any better
support.

>
> > Are you saying that if reddit use
> > ACL then the developer can go to Tahiti, not taking any call, and Franz
> > will support any reddit issue for them?
>
> I believe that Franz works very closely with some customers, and in some
> cases may know more about the application than some of the application
> programmers. I suppose it depends on the kind of revenue stream that
> the application represents. I doubt that a one-time $29.95 shrink-wrap
> license fee for an infinite number of redistributed binaries would
> get much help, but for $100 per copy (X many copies) it makes sense for
> Franz to be helpful.
> >
> >
>
> >
> >
> > I thought ACL have something like EXE maker. And that it shakes the
> > image and unused function from the distributed image. I never know that
> > when ACL build a distributed EXE, it includes all the IDE inside.
>
> If your image uses all of Lisp, that doesn't help. For example, you
> might use the IDE, or eval. Or you might use the compiler (e.g. for
> dynamic CLOS).
> >
> > I thought that if I bought MSVC++ and sell a product call
> > "start-up-vc.bat" which does nothing but start MSVC. And if I sells it
> > for 1$ and distributed MSVC++ with them, I can still get sued from MS.
> > And yet MS didn't charge MSVC per distributed app.
> >
> > To some extent, you are implying that any language that have "eval"
> > should be licensed per package sold. As a matter of fact, if a
> > interpreter can be written in a language such language implementation
> > should be licensed per package sold, too. Right?
>
> If you omitted eval from a copy of Allegro, but then
> you wrote an eval that had hooks to every function in Allegro, then
> not much could be left out of that Allegro. Just the main line of maybe
> a page or two of one program, eval.
> >
> > If you are afraid of your customer cheat in such a way, why not put it
> > in EULA? Get a lawyer to indicate what it means to sell product that,
> > in some way, allow one to use devlopment features of ACL. And indicate
> > that only such product require per-package distribution fees (well,
> > disregarding those Allegro special feature that should reasonably be
> > charged per package if used, like AllegroCache).
>
> If there is such an EULA, (I don't know),
> I think it would have to be a very elaborate license agreement
> to cover every contingency; so elaborate that a potential purchaser
> would have difficulty figuring out if it
> was right, and maybe would have to re-write particular parts. Maybe
> even hire a lawyer :( .
>
> If the current technique is for Franz to try to "partner" with the
> application writer, then that cooperation may be at least as useful as the proposal
> above. But presumably people at Franz read your note and this one
> and can comment.
>
> > You don't have to be afraid that they won't follow the EULA, since if
> > they are not willing to, I think they willl cheat in all other ways
> > anyway.
>
> There are ways of auditing the sale of products. While MS might not
> catch you making an extra copy of Word on your laptop, you can be sure
> that if a large organization, even a university, buys software,
> (or sells software) it cannot use piracy for long. (Unless it is in
> a country that does not respect patents, copyrights, etc.)
>

Right, that's why you don't have to be thinking of "Product Y that
provide all of ACL and sell for less" kind of situation if you put the
restriction in your EULA.

> So far as I can tell, the pricing model you are suggesting looks
> pretty close to what is already in place.
>
> It seems that we are hypothesizing the existence of
> a creative lisp programmer who has the following characteristics:
> 1. he/she avoids lawyers.
> 2. and yet is willing to assume unknown legal risks to use "free" software;
> 3. and probably has no financial backing (since investors
> (a) have lawyers
> (b) do not want to assume unknown legal risks. )
> 4. so has no money to buy a commercial Lisp.
>
> So Franz is not losing money on this person. :)
>
> More seriously, finding a good way to "not give away the whole thing"
> seems to be a key. Maybe distribute a fasl file. Run it on a trial version.
>
> RJF
>

May be something like VMWare player, say AllegroPlayer.
Just the runtime, where anyone can install for free and use for any
purpose.
Just the pure runtime, no IDE, no interface at all, without ability to
compile file, and must run from fasl. No AllegroCache or any of that
Allegro addon that you think deserve to be charge per copies.

From: Pisin Bootvong on
John Thingstad wrote:
> On Wed, 14 Dec 2005 05:14:46 +0100, Pisin Bootvong <joesb.coe9(a)gmail.com>
> wrote:
>
> > I thought that if I bought MSVC++ and sell a product call
> > "start-up-vc.bat" which does nothing but start MSVC. And if I sells it
> > for 1$ and distributed MSVC++ with them, I can still get sued from MS.
> > And yet MS didn't charge MSVC per distributed app.
> >
>
> This is a bit off topic but MS-VC++ compiler tools are avaliable for
> free from microsoft and have been for a time now.
> It is the IDE they charge for.
>
> --
> Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

Yes i knew and I mean that whole IDE, not just the compiler that I
attached for 1$.

Most development environment out there don't sells the runtime.
They sell the development environment. They sells the convenient of
being able to edit/compile/debug. But they rarely sells the runtime.

Selling each runtime only reduced the chance of spreading the runtime,
therefore reduce the chance of application developed in such runtime of
reaching large user base.
However, Selling the runtime is still better than Franz's model of
price per an application distributed: If I sell my ACL-base-app to
someone who also have purchased another ACL-base-app from somewhere
else, it cost them twice for the "license to run ACL".

Franz's view is that if I deliver the Lisp image:
- My app will be capable of using "Compiler code, Eval". So what? Any
turing complete language is capable of writing an eval/interpreter
anyway. There is nothing special about eval, just a built-in
interpreter. Eval is just a convenient feature. Another of the
arguement is that "If we don't charge for each package someone will
build aproduct which do nothing but invoke the REPL and have all
feature of ACL and compete with us for less price". It is nonsense.
Can't you put in any License agreement to pervent that? Have you ever
seen that happens to LispWorks?
- My app will have all the IDE features. Can't you disable this? You
can build a trial version that can't dump executable image and have
memory limit. But you can't disable IDE class to be built to exe? What
application mainly relied on ACL IDE component?

Instead of requiring fee on special case of the product that use IDE
component, Franz suspect everyone will use it and charge on that
component.

I think Franz target the server class environment anyway, so they don't
care about application that is sold in masses.
I'm not saying how this pricing model is good or bad for Franz, it's
their product after all. They can choose any pricing model they want.
All I'm saying is that I don't think this model will benefit Lisp in
any way.

First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4 5 6 7 8 9 10 11 12
Prev: Newlisp?
Next: cosx - sinx + 3x2