Packages that change without incrementing the version/release

[Max Bowsher]

Christopher Faylor wrote:
> On Sat, Nov 20, 2004 at 05:39:26PM -0000, Max Bowsher wrote:
>> Christopher Faylor wrote:
>>> setup.ini is always correct.  If the package changed in setup.ini, that
>>> should not trigger a download unless the user specifically asks for it.
>>> I don't see any reason to check md5sums for packages that aren't being
>>> installed.  When a package is specifically requested for reinstallation
>>> an informational message could be displayed indicating that the package
>>> had changed.
>>
>> That works up to a point.
>>
>> But setup cannot tell whether the reason for the mismatch is a corruptly
>> downloaded package, or a silent change - so what should it tell the user?
>
> "Size/MD5 mismatch between local package and internet host. 
> Redownloading."
>
> No need to make a value judgement about whether it's corruption on the
> local system or just a change on the host.  It shouldn't happen that
> often but, even if we forbid the practice, it is going to happen and
> there's no need to concern the user unduly when it does happen.

Fair enough, that's sufficiently neutral.

Hmm.... time to try and decipher the installation code and work out where to 
put the check.

Max.