How to deal with very large sources

[Ricardo Wurmus]

zimoun <address@hidden> writes:

> On Mon, 27 Jan 2020 at 15:45, Pierre Neidhardt <address@hidden> wrote:
>>
>> zimoun <address@hidden> writes:
>>
>> > Well, the questions are:
>> >  - hash?
>> >  - a copy on Berlin as substitute?
>>
>> What are the questions exactly?  I don't understand what you are asking.
>
> Considering "large" data, the questions are:
>
> How to compute the hash?
> Does Berlin have a copy of large data set?

We can mark packages as not substitutable, so the sources would be
downloaded on the users’ machines.  Whoever packages things would have
to compute the hash first, though.

-- 
Ricardo