Florian Weimer writes ("Re: Mass bugfiling potential: bundled implementation of
md5"):
> > On Thu, Oct 25, 2018 at 02:56:23PM +0800, Yangfl wrote:
> > What exactly are the benefits?
>
> You can use Systemtap or LD_PRELOAD to detect remaining uses of MD5.
> (We
* Adrian Bunk:
> On Thu, Oct 25, 2018 at 02:56:23PM +0800, Yangfl wrote:
>> Hi,
>>
>> Many of packages include bundled Aladdin Enterprises independent
>> implementation of md5. Full list can be seen at
>> https://codesearch.debian.net/search?q=typedef+unsigned+char+md5_byte_t&perpkg=1
>> (100 pac
On Thu, Oct 25, 2018 at 02:56:23PM +0800, Yangfl wrote:
> Hi,
>
> Many of packages include bundled Aladdin Enterprises independent
> implementation of md5. Full list can be seen at
> https://codesearch.debian.net/search?q=typedef+unsigned+char+md5_byte_t&perpkg=1
> (100 packages)
>
> As discussed
On Thu, Oct 25, 2018 at 09:32:57AM +, Holger Levsen wrote:
I dont think delaying fixes is a the best solution.
Especially since that's prioritising backports (support:
maintainer best-effort, as-is basis, use with care) over our main
release (full support)
--
⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Jonathan Dow
Jonas Smedegaard writes ("Re: Mass bugfiling potential: bundled implementation
of md5"):
> Quoting Bas Couwenberg (2018-10-25 09:44:20)
> > I suggest to postpone the MBF until libmd-dev (>= 1.0.1) is available in
> > stable (i.e. after the buster release).
>
>
Quoting Bas Couwenberg (2018-10-25 09:44:20)
> On 2018-10-25 08:56, Yangfl wrote:
> > A potential fix for that might be replace the md5.h header with
> >
> > #define LIBMD_MD5_ALADDIN
> > #include
> >
> > remove building for md5.c, and add
> >
> > Build-Depends: libmd-dev (>= 1.0.1)
>
> That i
On Thu, Oct 25, 2018 at 10:45:46AM +0200, Bas Couwenberg wrote:
> Or:
>
> - Leave the bugreport unresolved until libmd-dev (>= 1.0.1) is available in
> stable
>
> My initial suggestion to wait with the MBF until libmd-dev (>= 1.0.1) is
> available in stable is still the best solution in my opinio
Michael Biebl writes:
> You have two solutions here:
> - backport libmd-dev (>= 1.0.1)
> - revert this particular change in your package if you do a backport
Reasonable.
On 2018-10-25 10:41, Michael Biebl wrote:
Am 25.10.18 um 09:44 schrieb Bas Couwenberg:
On 2018-10-25 08:56, Yangfl wrote:
A potential fix for that might be replace the md5.h header with
#define LIBMD_MD5_ALADDIN
#include
remove building for md5.c, and add
Build-Depends: libmd-dev (>= 1.0.1)
Am 25.10.18 um 09:44 schrieb Bas Couwenberg:
> On 2018-10-25 08:56, Yangfl wrote:
>> A potential fix for that might be replace the md5.h header with
>>
>> #define LIBMD_MD5_ALADDIN
>> #include
>>
>> remove building for md5.c, and add
>>
>> Build-Depends: libmd-dev (>= 1.0.1)
>
> That is not a sol
On 2018-10-25 08:56, Yangfl wrote:
A potential fix for that might be replace the md5.h header with
#define LIBMD_MD5_ALADDIN
#include
remove building for md5.c, and add
Build-Depends: libmd-dev (>= 1.0.1)
That is not a solution for packages that are backported, as libmd-dev
(>= 1.0.1) is o
Hi,
Many of packages include bundled Aladdin Enterprises independent
implementation of md5. Full list can be seen at
https://codesearch.debian.net/search?q=typedef+unsigned+char+md5_byte_t&perpkg=1
(100 packages)
As discussed in #909116, libmd-dev now provides a compatible interface
for that, so
12 matches
Mail list logo
