From attila.lendvai at gmail.com  Fri Oct  2 09:39:47 2009
From: attila.lendvai at gmail.com (Attila Lendvai)
Date: Fri, 2 Oct 2009 11:39:47 +0200
Subject: [Bese-devel] CPS-transforming symbol-macros
In-Reply-To: <87ac6kifqf.fsf@lizard.king>
References: <87ac6kifqf.fsf@lizard.king>
Message-ID: <b54b34af0910020239m51446ce9n661562f853214ecc@mail.gmail.com>

> Now everything works. ?Is it a bug in CPS, or am I wrong?

it's a bug in the walker. i believe it's fixed in hu.dwim.walker by
now, maybe in it's origin, in arnesi, too...

-- 
 attila

ps: what are 3 years when you can mark mails? :)



From attila.lendvai at gmail.com  Fri Oct  2 10:21:24 2009
From: attila.lendvai at gmail.com (Attila Lendvai)
Date: Fri, 2 Oct 2009 12:21:24 +0200
Subject: [Bese-devel] Patch to rfc2388 to make IE file uploads work
	properly
In-Reply-To: <m2bqmqs0fk.fsf@bese.it>
References: <befaa7aa0611280541h3240207fnf3baaf76e6abd366@mail.gmail.com> 
	<m2bqmqs0fk.fsf@bese.it>
Message-ID: <b54b34af0910020321k13b1a733u8530474e0617f1ac@mail.gmail.com>

> "Marijn Haverbeke" <marijnh at gmail.com> writes:
>
>> It seems our rfc2388 lib allowed backslashes to escape the next character when
>> reading the attributes to a multipart header. IE passes the whole path of a
>> file as the filename attribute to Content-Disposition, but this library
>> behaviour stripped out these backslashes. After some looking at RFCs and
>> scratching my head, Marco explained to me that this behaviour is wrong, so here
>> is a patch to the library which removes the backslash-escaping behaviour.
>
> can anyone else confirm that our rfc2388 is wrong and ie is right?


any word on this?

this still cleanly applies to the rfc2388-binary repo...

-- 
 attila



From attila.lendvai at gmail.com  Fri Oct  2 10:36:28 2009
From: attila.lendvai at gmail.com (Attila Lendvai)
Date: Fri, 2 Oct 2009 12:36:28 +0200
Subject: [Bese-devel] fix for rfc2388 bug
In-Reply-To: <befaa7aa0612040618k4d81f70aif0c8d8627c7a5ee8@mail.gmail.com>
References: <befaa7aa0612040618k4d81f70aif0c8d8627c7a5ee8@mail.gmail.com>
Message-ID: <b54b34af0910020336m3444f524nf34a66ef18dbd88d@mail.gmail.com>

> Uploading a file containing a CRLF followed by more than two dashes fails
> with the current rfc2388 (it doesn't handle backtracking correctly when
> looking for the boundary string). Attached is a patch which adds test-cases
> for this (you might want to change those a bit -- I'm rather new to the
> world of unit-testing and 5am), and fixes the problem by changing
> read-until-next-boundary.
>
> For some reason it doesn't apply cleanly when my previous patch is also
> applied, but the conflict is very trivial to merge (there is no conflict
> that I can see, but darcs insists on putting V's and ^'s around some code).

any word on this?

i have it applied and merged in my local branch, ready to be pushed.

but should it be pushed? anyone with some rfc insights?

-- 
 attila



From drewc at tech.coop  Sun Oct  4 20:45:12 2009
From: drewc at tech.coop (Drew Crampsie)
Date: Sun, 4 Oct 2009 13:45:12 -0700
Subject: [Bese-devel] fix for rfc2388 bug
In-Reply-To: <b54b34af0910020336m3444f524nf34a66ef18dbd88d@mail.gmail.com>
References: <befaa7aa0612040618k4d81f70aif0c8d8627c7a5ee8@mail.gmail.com>
	<b54b34af0910020336m3444f524nf34a66ef18dbd88d@mail.gmail.com>
Message-ID: <bc10d2190910041345q3d881127u28679945687b66e1@mail.gmail.com>

Hey Attila, Marjin,

>> Uploading a file containing a CRLF followed by more than two dashes fails
>> with the current rfc2388 (it doesn't handle backtracking correctly when
>> looking for the boundary string.

Strangely enough, i just got a bug report from a client regarding this.


> any word on this?
>
> i have it applied and merged in my local branch, ready to be pushed.
>
> but should it be pushed? anyone with some rfc insights?


If you've been using it, and it fixes this problem, then i'm all for
pushing it. Otherwise, if there's any objection.. what's the URL for
your branch? :)

Cheers,

drewc
>
> --
> ?attila
>
> _______________________________________________
> bese-devel mailing list
> bese-devel at common-lisp.net
> http://common-lisp.net/cgi-bin/mailman/listinfo/bese-devel
>



From attila.lendvai at gmail.com  Mon Oct  5 05:32:58 2009
From: attila.lendvai at gmail.com (Attila Lendvai)
Date: Mon, 5 Oct 2009 07:32:58 +0200
Subject: [Bese-devel] fix for rfc2388 bug
In-Reply-To: <bc10d2190910041345q3d881127u28679945687b66e1@mail.gmail.com>
References: <befaa7aa0612040618k4d81f70aif0c8d8627c7a5ee8@mail.gmail.com> 
	<b54b34af0910020336m3444f524nf34a66ef18dbd88d@mail.gmail.com> 
	<bc10d2190910041345q3d881127u28679945687b66e1@mail.gmail.com>
Message-ID: <b54b34af0910042232u5e1cee1em7583bc261231ffd7@mail.gmail.com>

>> any word on this?
>>
>> i have it applied and merged in my local branch, ready to be pushed.
>>
>> but should it be pushed? anyone with some rfc insights?
>
>
> If you've been using it, and it fixes this problem, then i'm all for


well, i don't have any problem regarding this. i was just running
through my starred mails an this patch was one of them...


> pushing it. Otherwise, if there's any objection.. what's the URL for
> your branch? :)


by local branch i've meant my local darcs repo on my laptop.

i've pushed some changes to the official repo, please take a look at
them an report if anything breaks, and do send a darcs patch with the
failing test!

those patches look better than letting them rest in mail archives for
2+ years...

-- 
 attila