From fred at streamfocus.com  Thu Jan  7 20:33:08 2010
From: fred at streamfocus.com (Fred Gibson)
Date: Thu, 7 Jan 2010 12:33:08 -0800
Subject: [drakma-devel] Drakma - problem with http-request on SBCL 1.0.34
Message-ID: <19290c161001071233x16a4a3ddp86fca6923a4fb7a@mail.gmail.com>

I've found a small problem in Drakma's http-request where in a call to
usocket:socket-connect it sent the key :nodelay which doesn't exist
for sbcl's socket connect version, so I changed that call in drakma to
this:

         (usocket:socket-stream
            (usocket:socket-connect host port
               :element-type 'octet
               #+:openmcl :deadline
               #+:openmcl deadline
               #-:sbcl :nodelay
               #-:sbcl t))

Then it seemed to work just fine.

My best,

Fred Gibson

Founder / Software Developer
http://www.streamfocus.com

(c)2010 Organon Technologies LLC



From ehuels at gmail.com  Thu Jan  7 21:27:38 2010
From: ehuels at gmail.com (Erik Huelsmann)
Date: Thu, 7 Jan 2010 22:27:38 +0100
Subject: [drakma-devel] Drakma - problem with http-request on SBCL 1.0.34
In-Reply-To: <19290c161001071233x16a4a3ddp86fca6923a4fb7a@mail.gmail.com>
References: <19290c161001071233x16a4a3ddp86fca6923a4fb7a@mail.gmail.com>
Message-ID: <aea328ab1001071327s41114855n7721658b84238686@mail.gmail.com>

Did we (usocket developers) miss a function argument?

Chun, shouldn't we at least accept it?

Bye,

Erik.

On Thu, Jan 7, 2010 at 9:33 PM, Fred Gibson <fred at streamfocus.com> wrote:
> I've found a small problem in Drakma's http-request where in a call to
> usocket:socket-connect it sent the key :nodelay which doesn't exist
> for sbcl's socket connect version, so I changed that call in drakma to
> this:
>
> ? ? ? ? (usocket:socket-stream
> ? ? ? ? ? ?(usocket:socket-connect host port
> ? ? ? ? ? ? ? :element-type 'octet
> ? ? ? ? ? ? ? #+:openmcl :deadline
> ? ? ? ? ? ? ? #+:openmcl deadline
> ? ? ? ? ? ? ? #-:sbcl :nodelay
> ? ? ? ? ? ? ? #-:sbcl t))
>
> Then it seemed to work just fine.
>
> My best,
>
> Fred Gibson
>
> Founder / Software Developer
> http://www.streamfocus.com
>
> (c)2010 Organon Technologies LLC
>
> _______________________________________________
> drakma-devel mailing list
> drakma-devel at common-lisp.net
> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel
>



From binghe.lisp at gmail.com  Thu Jan  7 23:24:12 2010
From: binghe.lisp at gmail.com (Chun Tian (binghe))
Date: Fri, 8 Jan 2010 07:24:12 +0800
Subject: [drakma-devel] Drakma - problem with http-request on SBCL 1.0.34
In-Reply-To: <aea328ab1001071327s41114855n7721658b84238686@mail.gmail.com>
References: <19290c161001071233x16a4a3ddp86fca6923a4fb7a@mail.gmail.com>
	<aea328ab1001071327s41114855n7721658b84238686@mail.gmail.com>
Message-ID: <3F6EB825-2E11-490D-9DFB-482FF8952EFF@gmail.com>

Hi, Erik (and Drakma users)

(I'm not in Drakma mailing list until now)

I just check latest Drakma (1.1.0) build with SBCL 1.0.33 and USOCKET (trunk and 0.4.x branch), it works just fine with a test (drakma:http-request "http://www.gnu.org/"). And I checked related USOCKET code, nothing strange found.

I don't know why Fred Gibson think (:nodelay t) are not supported on USOCKET with SBCL. Maybe he's using too old version of either USOCKET or SBCL ... more version details needed.

Regards,

Chun Tian (binghe)

? 2010-1-8?05:27? Erik Huelsmann ???

> Did we (usocket developers) miss a function argument?
> 
> Chun, shouldn't we at least accept it?
> 
> Bye,
> 
> Erik.
> 
> On Thu, Jan 7, 2010 at 9:33 PM, Fred Gibson <fred at streamfocus.com> wrote:
>> I've found a small problem in Drakma's http-request where in a call to
>> usocket:socket-connect it sent the key :nodelay which doesn't exist
>> for sbcl's socket connect version, so I changed that call in drakma to
>> this:
>> 
>>         (usocket:socket-stream
>>            (usocket:socket-connect host port
>>               :element-type 'octet
>>               #+:openmcl :deadline
>>               #+:openmcl deadline
>>               #-:sbcl :nodelay
>>               #-:sbcl t))
>> 
>> Then it seemed to work just fine.
>> 
>> My best,
>> 
>> Fred Gibson
>> 
>> Founder / Software Developer
>> http://www.streamfocus.com
>> 
>> (c)2010 Organon Technologies LLC
>> 
>> _______________________________________________
>> drakma-devel mailing list
>> drakma-devel at common-lisp.net
>> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel
>> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2603 bytes
Desc: not available
URL: <https://mailman.common-lisp.net/pipermail/drakma-devel/attachments/20100108/0711bc0d/attachment.bin>

From fred at streamfocus.com  Fri Jan  8 00:01:18 2010
From: fred at streamfocus.com (Fred Gibson)
Date: Thu, 7 Jan 2010 16:01:18 -0800
Subject: [drakma-devel] Drakma - problem with http-request on SBCL 1.0.34
In-Reply-To: <3F6EB825-2E11-490D-9DFB-482FF8952EFF@gmail.com>
References: <19290c161001071233x16a4a3ddp86fca6923a4fb7a@mail.gmail.com>
	<aea328ab1001071327s41114855n7721658b84238686@mail.gmail.com>
	<3F6EB825-2E11-490D-9DFB-482FF8952EFF@gmail.com>
Message-ID: <19290c161001071601k2444730ck6955568fe2ef8c24@mail.gmail.com>

Hi Chun,

Ah yes, the obvious: I'm using usocket 0.3.5  when the latest release
is 0.4.1 showing many fixes in between, no doubt addressing this
problem.

My best,

Fred

2010/1/7 Chun Tian (binghe) <binghe.lisp at gmail.com>:
> Hi, Erik (and Drakma users)
>
> (I'm not in Drakma mailing list until now)
>
> I just check latest Drakma (1.1.0) build with SBCL 1.0.33 and USOCKET (trunk and 0.4.x branch), it works just fine with a test (drakma:http-request "http://www.gnu.org/"). And I checked related USOCKET code, nothing strange found.
>
> I don't know why Fred Gibson think (:nodelay t) are not supported on USOCKET with SBCL. Maybe he's using too old version of either USOCKET or SBCL ... more version details needed.
>
> Regards,
>
> Chun Tian (binghe)
>
> ? 2010-1-8?05:27? Erik Huelsmann ???
>
>> Did we (usocket developers) miss a function argument?
>>
>> Chun, shouldn't we at least accept it?
>>
>> Bye,
>>
>> Erik.
>>
>> On Thu, Jan 7, 2010 at 9:33 PM, Fred Gibson <fred at streamfocus.com> wrote:
>>> I've found a small problem in Drakma's http-request where in a call to
>>> usocket:socket-connect it sent the key :nodelay which doesn't exist
>>> for sbcl's socket connect version, so I changed that call in drakma to
>>> this:
>>>
>>>         (usocket:socket-stream
>>>            (usocket:socket-connect host port
>>>               :element-type 'octet
>>>               #+:openmcl :deadline
>>>               #+:openmcl deadline
>>>               #-:sbcl :nodelay
>>>               #-:sbcl t))
>>>
>>> Then it seemed to work just fine.
>>>
>>> My best,
>>>
>>> Fred Gibson
>>>
>>> Founder / Software Developer
>>> http://www.streamfocus.com
>>>
>>> (c)2010 Organon Technologies LLC
>>>
>>> _______________________________________________
>>> drakma-devel mailing list
>>> drakma-devel at common-lisp.net
>>> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel
>>>
>
>
> _______________________________________________
> drakma-devel mailing list
> drakma-devel at common-lisp.net
> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel
>
>



-- 
Fred Gibson

Founder / Software Developer
http://www.streamfocus.com

(c)2010 Organon Technologies LLC



From binghe.lisp at gmail.com  Fri Jan  8 00:05:43 2010
From: binghe.lisp at gmail.com (Chun Tian (binghe))
Date: Fri, 8 Jan 2010 08:05:43 +0800
Subject: [drakma-devel] Drakma - problem with http-request on SBCL 1.0.34
In-Reply-To: <19290c161001071601k2444730ck6955568fe2ef8c24@mail.gmail.com>
References: <19290c161001071233x16a4a3ddp86fca6923a4fb7a@mail.gmail.com>
	<aea328ab1001071327s41114855n7721658b84238686@mail.gmail.com>
	<3F6EB825-2E11-490D-9DFB-482FF8952EFF@gmail.com>
	<19290c161001071601k2444730ck6955568fe2ef8c24@mail.gmail.com>
Message-ID: <4F6489EB-5227-4ED3-A6D0-52A1EC2A4CA7@gmail.com>

Hi, Fred

Good to know. The 0.4.x release is first started at Oct, 2008, please do not use 0.3.x any more.

--binghe

? 2010-1-8?08:01? Fred Gibson ???

> Hi Chun,
> 
> Ah yes, the obvious: I'm using usocket 0.3.5  when the latest release
> is 0.4.1 showing many fixes in between, no doubt addressing this
> problem.
> 
> My best,
> 
> Fred
> 
> 2010/1/7 Chun Tian (binghe) <binghe.lisp at gmail.com>:
>> Hi, Erik (and Drakma users)
>> 
>> (I'm not in Drakma mailing list until now)
>> 
>> I just check latest Drakma (1.1.0) build with SBCL 1.0.33 and USOCKET (trunk and 0.4.x branch), it works just fine with a test (drakma:http-request "http://www.gnu.org/"). And I checked related USOCKET code, nothing strange found.
>> 
>> I don't know why Fred Gibson think (:nodelay t) are not supported on USOCKET with SBCL. Maybe he's using too old version of either USOCKET or SBCL ... more version details needed.
>> 
>> Regards,
>> 
>> Chun Tian (binghe)
>> 
>> ? 2010-1-8?05:27? Erik Huelsmann ???
>> 
>>> Did we (usocket developers) miss a function argument?
>>> 
>>> Chun, shouldn't we at least accept it?
>>> 
>>> Bye,
>>> 
>>> Erik.
>>> 
>>> On Thu, Jan 7, 2010 at 9:33 PM, Fred Gibson <fred at streamfocus.com> wrote:
>>>> I've found a small problem in Drakma's http-request where in a call to
>>>> usocket:socket-connect it sent the key :nodelay which doesn't exist
>>>> for sbcl's socket connect version, so I changed that call in drakma to
>>>> this:
>>>> 
>>>>        (usocket:socket-stream
>>>>           (usocket:socket-connect host port
>>>>              :element-type 'octet
>>>>              #+:openmcl :deadline
>>>>              #+:openmcl deadline
>>>>              #-:sbcl :nodelay
>>>>              #-:sbcl t))
>>>> 
>>>> Then it seemed to work just fine.
>>>> 
>>>> My best,
>>>> 
>>>> Fred Gibson
>>>> 
>>>> Founder / Software Developer
>>>> http://www.streamfocus.com
>>>> 
>>>> (c)2010 Organon Technologies LLC
>>>> 
>>>> _______________________________________________
>>>> drakma-devel mailing list
>>>> drakma-devel at common-lisp.net
>>>> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel
>>>> 
>> 
>> 
>> _______________________________________________
>> drakma-devel mailing list
>> drakma-devel at common-lisp.net
>> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel
>> 
>> 
> 
> 
> 
> -- 
> Fred Gibson
> 
> Founder / Software Developer
> http://www.streamfocus.com
> 
> (c)2010 Organon Technologies LLC
> 
> _______________________________________________
> drakma-devel mailing list
> drakma-devel at common-lisp.net
> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2603 bytes
Desc: not available
URL: <https://mailman.common-lisp.net/pipermail/drakma-devel/attachments/20100108/0c6e20cc/attachment.bin>

From binghe.lisp at gmail.com  Fri Jan  8 00:02:33 2010
From: binghe.lisp at gmail.com (Chun Tian (binghe))
Date: Fri, 8 Jan 2010 08:02:33 +0800
Subject: [drakma-devel] Drakma - problem with http-request on SBCL 1.0.34
In-Reply-To: <3F6EB825-2E11-490D-9DFB-482FF8952EFF@gmail.com>
References: <19290c161001071233x16a4a3ddp86fca6923a4fb7a@mail.gmail.com>
	<aea328ab1001071327s41114855n7721658b84238686@mail.gmail.com>
	<3F6EB825-2E11-490D-9DFB-482FF8952EFF@gmail.com>
Message-ID: <88D050A4-58D7-41E4-B223-A6CCC516F4A3@gmail.com>

The same with SBCL-1.0.34 and USOCKET 0.4.1 release.

But I think a new USOCKET 0.4.2 bugfix release is needed, we have reached a another release cycle.

--binghe

? 2010-1-8?07:24? Chun Tian (binghe) ???

> Hi, Erik (and Drakma users)
> 
> (I'm not in Drakma mailing list until now)
> 
> I just check latest Drakma (1.1.0) build with SBCL 1.0.33 and USOCKET (trunk and 0.4.x branch), it works just fine with a test (drakma:http-request "http://www.gnu.org/"). And I checked related USOCKET code, nothing strange found.
> 
> I don't know why Fred Gibson think (:nodelay t) are not supported on USOCKET with SBCL. Maybe he's using too old version of either USOCKET or SBCL ... more version details needed.
> 
> Regards,
> 
> Chun Tian (binghe)
> 
> ? 2010-1-8?05:27? Erik Huelsmann ???
> 
>> Did we (usocket developers) miss a function argument?
>> 
>> Chun, shouldn't we at least accept it?
>> 
>> Bye,
>> 
>> Erik.
>> 
>> On Thu, Jan 7, 2010 at 9:33 PM, Fred Gibson <fred at streamfocus.com> wrote:
>>> I've found a small problem in Drakma's http-request where in a call to
>>> usocket:socket-connect it sent the key :nodelay which doesn't exist
>>> for sbcl's socket connect version, so I changed that call in drakma to
>>> this:
>>> 
>>>        (usocket:socket-stream
>>>           (usocket:socket-connect host port
>>>              :element-type 'octet
>>>              #+:openmcl :deadline
>>>              #+:openmcl deadline
>>>              #-:sbcl :nodelay
>>>              #-:sbcl t))
>>> 
>>> Then it seemed to work just fine.
>>> 
>>> My best,
>>> 
>>> Fred Gibson
>>> 
>>> Founder / Software Developer
>>> http://www.streamfocus.com
>>> 
>>> (c)2010 Organon Technologies LLC
>>> 
>>> _______________________________________________
>>> drakma-devel mailing list
>>> drakma-devel at common-lisp.net
>>> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel
>>> 
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2603 bytes
Desc: not available
URL: <https://mailman.common-lisp.net/pipermail/drakma-devel/attachments/20100108/f9968d06/attachment.bin>

From dodek at dodecki.net  Fri Jan  8 18:13:35 2010
From: dodek at dodecki.net (Adam Dodek Michalik)
Date: Fri, 8 Jan 2010 19:13:35 +0100
Subject: [drakma-devel] post requests and content-type
Message-ID: <407a243c1001081013h3c0893aenc98f8f6cd8486836@mail.gmail.com>

Hey,
I noticed that when using POST method, Drakma completely ignores
content-type provided by user. That's kinda bad for me, since I'd like
to be able to send the content's charset as well - so that
Content-Type header looks for instance like this:

application/x-www-form-urlencoded; charset=utf-8

Modifying drakma so that it uses content-type provided by user is trivial,
but maybe there are other reasons behind this behaviour I'm not aware of.

-- 
Adam Michalik
  vel Dodek Dodecki
<dodek at dodecki.net>



From reddaly at gmail.com  Thu Jan 21 09:57:39 2010
From: reddaly at gmail.com (Red Daly)
Date: Thu, 21 Jan 2010 01:57:39 -0800
Subject: [drakma-devel] Patch - Accept-Encoding: gzip
Message-ID: <f6700f0c1001210157p539fde29n96a8c4615b90eb87@mail.gmail.com>

Dear drakma users,

I made drakma accept gzipped message bodies by default, and also
automatically decode them.  A patch is attached.

Unfortunately this adds a new dependency on gzip-sequence (which in
turn depends on salza2).  See the docstrings for more details.  To sum
it up: if WANT-STREAM is T, It wraps the returned stream in a
flexi-stream that wraps a gzip-input-stream.  Otherwise, it decodes
the sequence before returning from HTTP-REQUEST.

Thanks,
Red
-------------- next part --------------
A non-text attachment was scrubbed...
Name: drakma-gzip-patch.patch
Type: text/x-patch
Size: 9430 bytes
Desc: not available
URL: <https://mailman.common-lisp.net/pipermail/drakma-devel/attachments/20100121/28e82e59/attachment.bin>

From avodonosov at yandex.ru  Fri Jan 22 02:43:26 2010
From: avodonosov at yandex.ru (Anton Vodonosov)
Date: Fri, 22 Jan 2010 04:43:26 +0200
Subject: [drakma-devel] Drakma doesn't work on CLISP because of unsupported
	TCP_NODELAY
Message-ID: <154814915.20100122044326@yandex.ru>

Hello,

The recent Drakma (1.1.0) doesn't work on CLISP because of :NODELAY T
passed to USOCKET:SOCKET-CONNECT.

:NODELAY is not supported on CLISP and usocket signals an error.

IMHO it would be better to remove the :NODELAY T, not only because of
the CLISP support, but also because the Nagle's algorithm disabled
by :NODELAY T seems to be a feature rather than a problem.

Best regards,
- Anton




From edi at agharta.de  Mon Jan 25 20:52:52 2010
From: edi at agharta.de (Edi Weitz)
Date: Mon, 25 Jan 2010 21:52:52 +0100
Subject: [drakma-devel] Patch - Accept-Encoding: gzip
In-Reply-To: <f6700f0c1001210157p539fde29n96a8c4615b90eb87@mail.gmail.com>
References: <f6700f0c1001210157p539fde29n96a8c4615b90eb87@mail.gmail.com>
Message-ID: <b56421a01001251252r58d52e4elbd627e6e8305cfcf@mail.gmail.com>

Hi Red,

Thanks for the patch.  (But note the guidelines at
<http://weitz.de/patches.html>...)

I'm unsure whether I want to accept the patch, though.  I don't really
want to add a dependency on another library which in turn depends on
another library for a pretty rare case.  I'd rather implement a hook
or something so that Drakma users can pre-process the content
externally.

Cheers,
Edi.


On Thu, Jan 21, 2010 at 10:57 AM, Red Daly <reddaly at gmail.com> wrote:
> Dear drakma users,
>
> I made drakma accept gzipped message bodies by default, and also
> automatically decode them. ?A patch is attached.
>
> Unfortunately this adds a new dependency on gzip-sequence (which in
> turn depends on salza2). ?See the docstrings for more details. ?To sum
> it up: if WANT-STREAM is T, It wraps the returned stream in a
> flexi-stream that wraps a gzip-input-stream. ?Otherwise, it decodes
> the sequence before returning from HTTP-REQUEST.
>
> Thanks,
> Red
>
> _______________________________________________
> drakma-devel mailing list
> drakma-devel at common-lisp.net
> http://common-lisp.net/cgi-bin/mailman/listinfo/drakma-devel
>
>