Re: [PATCH] checksums.ini: correct wrong checksums

[Philip Balister <philip@balister.org>]

[-- Attachment #1: Type: text/plain, Size: 1301 bytes --]

Koen Kooi wrote:
> On 09-11-2008 17:54, Phil Blundell wrote:
>> On Sun, 2008-11-09 at 16:28 +0300, Dmitry Baryshkov wrote:
>>>   
>>> [http://gpephone.linuxtogo.org/download/gpephone/libchenabler-0.1/libchenabler-0.1.tar.gz] 
>>>
>>> -md5=f29e1e63fcc55e3ae4e81f5ee9dd7bdd
>>> -sha256=448a3ab76109664bb6bd8bcc2c62d957c7d44b52e39c37d91b3b73b5cfe35a3f
>>> +md5=866fd13611a8de946428db1df9be8468
>>> +sha256=74def55ba6a61e966e0873a4081fd3dd65089b94b837810d3bc056221d7e41c7
>>
>> How did these wrong checksums get in there?  It seems to defeat the
>> point of checksums.ini if the data isn't trustworthy.
> 
> The gpephone people are in a habit of replacing the tarballs to get 
> bugfixes to people instead of doing a new release. So the problem here 
> is that _upstream_ is making a mess of things.
> This also highlights why OE is now strictly enforcing checksums.

I'd like to see one commit per checksum change with the details in the 
commit message. This we we can easily find when I package checksum 
changed. This way we can easily identifu when the source changed but the 
version number did not.

This might be useful in the future for debugging the package. As in 
feature X used to work, but all of a sudden broke, even though the 
version did not change.

Philip

[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/x-pkcs7-signature, Size: 3303 bytes --]