[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [p2-dev] A typical use-case

A shameless self-promotion :-) I have a tool that lets you browse p2
repositories, see composite structures, IU dependency tree/list, etc. I
can provide prebuilt binaries if anyone is interested (the sources are
at github already).


On 12-02-28 6:08 AM, Gunnar Wagenknecht wrote:
Am 27.02.2012 23:01, schrieb Thomas Hallgren:
Another way to reduce the pain would be to introduce smarter
protocols that would enable p2 queries to be sent and evaluated on a
server rather than copying the "database" in its entirety to the client.

I don't think that downloading the database is a bad idea. Once the database is local, all future queries can happen local. There seems to be an issue with too many database downloads, though. Sometimes p2 does too many downloads for the same database. I whish the caching would be a bit more agressive. Another option might be to ship software packages pre-loaded with local databases.

The simple solution that is feasible short-term is to maintain separate repositories.

I tend to agree. Even for my target platform definitions I tend to be more specific and avoid composites. Luckily, I have shell access and can browse the download servers to find the one I want. But other users just get the 404/403 and will never find out if a repo is a composite or not. I wonder if directory browsing should be enabled for download servers.