Hi,
I would like to know if there are already plans to make Hello-C a
standalone project in www.common-lisp.net. I am interested in any
development in the (U)FFI field, with aims at making ECL compliant with
this promising new interface. I see there is already code at the Cello
project but I am currently not interested in Cell{o,s}.
Regards,
Juanjo (Maintainer of ECL, http://ecls.sf.net)
> Hello-C mailing list?
There is a mailing list related to Hello-C on this page.
http://www.parenpower.com/cgi-bin/mail.cgi/list/
>I had a look at CFFI and it does not seem to even contain a fraction of
>what UFFI provides. I thought Hello-C was supposed to be a form from
>UFFI, rather than a completely new library written from scratch.
If you are not happy with the direction Hello-C is taking, now would be
a good time to discuss your concerns on the above mailing list. This
page might reassure you, though:
http://www.alphageeksinc.com/cgi-bin/lispnyc.cgi?HelloC
"Hello-C is a recent fork of UFFI, or Universal Foreign Function
Interface. Hello-C adds support for callbacks from C, and is meant to
allow UFFI to move forward, since UFFI's originator needs it to remain
stable for his own work."
Eric Lavigne wrote:
>>Hello-C mailing list?
>
> There is a mailing list related to Hello-C on this page.
> http://www.parenpower.com/cgi-bin/mail.cgi/list/
Yes, there is, and no one is going to use it or any of the other lists.
Google mandated that development take place on something like
SourceForge. In the case of CL, that is common-lis.net. Anyway, projects
at these places come with mail lists, and one wants the legislative
history of a project to be at the same place as the code and doc.
I think the guys who set those lists up simply did not think all that
through.
--
Kenny
Why Lisp? http://lisp.tech.coop/RtL%20Highlight%20Film
"If you plan to enter text which our system might consider to be
obscene, check here to certify that you are old enough to hear the
resulting output." -- Bell Labs text-to-speech interactive Web page
"Juanjo" <····@arrakis.es> writes:
> Hi,
>
> I would like to know if there are already plans to make Hello-C a
> standalone project in www.common-lisp.net. I am interested in any
> development in the (U)FFI field, with aims at making ECL compliant with
> this promising new interface. I see there is already code at the Cello
> project but I am currently not interested in Cell{o,s}.
>
> Regards,
>
> Juanjo (Maintainer of ECL, http://ecls.sf.net)
i'm under the impression[1] that this will be done with the cffi project:
http://common-lisp.net/project/cffi
[1] - i may very well be mistaken.
--
-Marco
Ring the bells that still can ring.
Forget the perfect offering.
There is a crack in everything.
That's how the light gets in.
-Leonard Cohen
I had a look at CFFI and it does not seem to even contain a fraction of
what UFFI provides. I thought Hello-C was supposed to be a form from
UFFI, rather than a completely new library written from scratch.
Juanjo
Juanjo wrote:
> I had a look at CFFI and it does not seem to even contain a fraction of
> what UFFI provides. I thought Hello-C was supposed to be a form from
> UFFI, rather than a completely new library written from scratch.
To find out what is going on and throw in your own ideas, find the
Fetter (FFI autogen) project on common-lisp.net and sign up for the
fetter-devel mailing list. Because of the close relationship between the
two projects, we thought we would discuss them both on the Fetter list.
Hello-C itself is my recent fork of UFFI, mostly adding callbacks from C
to Lisp. We also like some work we see in CFFI, which includes a UFFI
compatibility module (how much I did not notice).
Who knows where this will end up? All projects have friendly licenses
and are free to borrow from each other. Hello-C by name may cease to
exist or may end up with its own project. Meanwhile the CFFI developer
has responded positively to Hello-C extending his work, so the source
will be in a branch of the CFFI CVS tree for now.
Hope that is not too confusing. :)
--
Kenny
Why Lisp? http://lisp.tech.coop/RtL%20Highlight%20Film
"If you plan to enter text which our system might consider to be
obscene, check here to certify that you are old enough to hear the
resulting output." -- Bell Labs text-to-speech interactive Web page
I have subscribed to the mailing list you mentioned.
My concern is that ECL's FFI is yet a growing thing. Formerly it looked
much like GCL's primitives for mixing C and lisp code. Currently it has
evolved to implement all of the UFFI specification and is rather
stable. It is use extensively in ECL itself, with some nice
applications to implement sockets and a win32-bridge to be seen in the
screenshots page
http://sourceforge.net/project/screenshots.php?group_id=30035
However we still do not have the infrastructure for supporting
callbacks in a lispy way -- you can if you mix C and lisp --, and since
we still have to implement it, I would like to fix an API which is
compatible with some popular FFI package.
So then, see you at the mailing list.
Juanjo
On 2005-06-29, Kenny Tilton <·······@nyc.rr.com> wrote:
> [...]
> Who knows where this will end up? All projects have friendly licenses
> and are free to borrow from each other. Hello-C by name may cease to
> exist or may end up with its own project. Meanwhile the CFFI developer
> has responded positively to Hello-C extending his work, so the source
> will be in a branch of the CFFI CVS tree for now.
>
> Hope that is not too confusing. :)
Well, sorry to add to the confusion, but the licensing issue is not
clear, and that is my fault. I originally licensed UFFI under Franz's
Lesser Lisp GPL (LLGPL). However, I subsequently changed the license
to the more liberal BSD license. I updated the LICENSE file with this
information, but neglected to update license header in the individual
source files. I will rectify this discordance.
My inclination is complete the intended transition by change the
license test in the source files to a BSD license. However, that would
prevent UFFI from participating in the "borrowing" process you
mentioned. How would you feel about following the transition to a BSD
license for your UFFI fork?
PS Congratulations on the funding for lispnyc's summer of code
--
Kevin Rosenberg
·····@rosenberg.net
Kevin Rosenberg wrote:
> On 2005-06-29, Kenny Tilton <·······@nyc.rr.com> wrote:
>
>>[...]
>>Who knows where this will end up? All projects have friendly licenses
>>and are free to borrow from each other. Hello-C by name may cease to
>>exist or may end up with its own project. Meanwhile the CFFI developer
>>has responded positively to Hello-C extending his work, so the source
>>will be in a branch of the CFFI CVS tree for now.
>>
>>Hope that is not too confusing. :)
>
>
> Well, sorry to add to the confusion, but the licensing issue is not
> clear, and that is my fault. I originally licensed UFFI under Franz's
> Lesser Lisp GPL (LLGPL). However, I subsequently changed the license
> to the more liberal BSD license. I updated the LICENSE file with this
> information, but neglected to update license header in the individual
> source files. I will rectify this discordance.
>
> My inclination is complete the intended transition by change the
> license test in the source files to a BSD license. However, that would
> prevent UFFI from participating in the "borrowing" process you
> mentioned.
IANAL, but:
(1) I do not see why BSD interferes with borrowing. I just read it
again, nope, do not see a problem. The only constraint is preserving the
copyright notice -- no problem there. If you mean reconciling the
licenses, shucks, looks like we have BSD-MIT-MIT. Should not be too hard
to sort out.
(2) When I forked from UFFI, I think you were LLGPL. Not sure because
all I have is the source and I gather those got out of synch. I guess we
could reconstruct when the fork took place one way or another. This
matters because your change to BSD is not retroactive. Of course, any
new development would fall under the BSD, so sufficient added-value
makes the LLGPL code base unpalatable. Aside from that, the LLGPL-based
release lives on under that license.
> How would you feel about following the transition to a BSD
> license for your UFFI fork?
I have no problem with BSD. I went with MIT on my stuff. CFFI looks like
MIT.
>
> PS Congratulations on the funding for lispnyc's summer of code
>
Thanks. And, as I said long ago when UFFI got me across what seemed like
a vast barrier to my first C library, thanks to you for the uber-contrib
that is UFFI. We will try to live up to that high standard with Fetter
and "UFFI, The Next Generation" (CFFI or Hello-C).
--
Kenny
Why Lisp? http://lisp.tech.coop/RtL%20Highlight%20Film
"If you plan to enter text which our system might consider to be
obscene, check here to certify that you are old enough to hear the
resulting output." -- Bell Labs text-to-speech interactive Web page
On 2005-06-29, Kenny Tilton <·······@nyc.rr.com> wrote:
> (1) I do not see why BSD interferes with borrowing. I just read it
> again, nope, do not see a problem. The only constraint is preserving the
> copyright notice -- no problem there. If you mean reconciling the
> licenses, shucks, looks like we have BSD-MIT-MIT. Should not be too hard
> to sort out.
It's the LLGPL which prevents taking LLGPL code and relicensing under
the more liberal BSD license.
> (2) When I forked from UFFI, I think you were LLGPL. Not sure because
> all I have is the source and I gather those got out of synch. I guess we
> could reconstruct when the fork took place one way or another. This
> matters because your change to BSD is not retroactive. Of course, any
> new development would fall under the BSD, so sufficient added-value
> makes the LLGPL code base unpalatable. Aside from that, the LLGPL-based
> release lives on under that license.
I'd say at the time of the fork, the license of UFFI was interminate
between BSD and LLGPL since both license texts were present. However,
it's fairly easy to take BSD licensed code and add a LLGPL license to
it. It's not possible to do the opposite.
> I have no problem with BSD. I went with MIT on my stuff. CFFI looks like
> MIT.
Very good. I'll complete the transition from LLGPL to BSD in the UFFI
source code. Given my intent (and preconception) was that UFFI was
completely licensed under BSD at the time of the fork and that I'm the
copyright holder, you can safely convert your source file license text
to BSD.
Best,
--
Kevin Rosenberg
·····@rosenberg.net
Kevin Rosenberg wrote:
> Very good. I'll complete the transition from LLGPL to BSD in the UFFI
> source code. Given my intent (and preconception) was that UFFI was
> completely licensed under BSD at the time of the fork and that I'm the
> copyright holder, you can safely convert your source file license text
> to BSD.
OK, great. I will pass that consent on to Luis Oliveira, who is the
developer of <whatever we call it> and took (or will take) the source
from Hello-c before the license gets changed.
--
Kenny
Why Lisp? http://lisp.tech.coop/RtL%20Highlight%20Film
"If you plan to enter text which our system might consider to be
obscene, check here to certify that you are old enough to hear the
resulting output." -- Bell Labs text-to-speech interactive Web page
On 2005-06-30, Kenny Tilton <·······@nyc.rr.com> wrote:
> OK, great. I will pass that consent on to Luis Oliveira, who is the
> developer of <whatever we call it> and took (or will take) the source
> from Hello-c before the license gets changed.
Sounds good to me.
--
Kevin Rosenberg
·····@rosenberg.net