On Sat, Jan 17, 2009 at 06:04:13PM +0000, Stephen Gran wrote: > This one time, at band camp, Roger Leigh said: > > Dear all, > > <snip good news> > > > We would be grateful for any comments you might have, particularly > > if you have used the existing wanna-build and can see any defects, > > omissions, or any suggestions for anything we missed or could > > implement in a more optimal fashion. If you are a PostgreSQL guru, > > a review of the PL/pgSQL and PL/Perl functions and triggers used > > for custom constraint checking would be much appreciated. > > I am not a dak hacker, nor an ftpteam or wanna-build guy. That being > said, it looks like a lot of the information you store and the functions > for comparison look an awful lot like the stuff dak currently has. Does > it make sense to consolidate, and expand dak to handle the extra things > you need it to do? I think that's a definite possibility to consider. dak stores (almost?) all the package information we are going to import anyway, so the additional tables we use could just reference those instead. Looking in more detail at the dak schema, a lot of the tables are almost identical, just slightly different implementation details and names. One thing we're lacking there is a debian version type which supports the same comparisons as dpkg --compare-versions. In our schema we are using such a type for indexing, sorting and joining. It wouldn't be hard to update projectb to add such facilities, though. Fundamentally, what we store in the sources and binaries tables is the entire contents of the Sources and Packages files for each suite/component (distribution and section in our schema) and architecture. dist_(binaries|sources) then tells us which versions of which package are available in each suite/component. If this is what dak stores, it's already doing a large part of what we want. Possible differences in what dak stores here are: - information in Packages/Sources not in dak such as Build-Deps (we could use these directly in the DB to determine when to reschedule a build). Is there anything from Packages/Sources deliberately not stored in dak? - we keep a history of old packages and sources; does dak keep these around once they are no longer referenced by (bin|src)_associations? The package state history references these, and would need periodic pruning. - builders/build_jobs/logs are wanna-build-specific. > It could presumably go in a seperate schema with > grants and so on so that permissions can be kept sensible - all of this > would need to be worked out with the ftpteam people. It's just that I > get sad every time I see needless duplication of information in > databases :) I haven't personally used schemas and privileges to any great extent. But if it's technically feasible, and wouldn't cause too much trouble to the ftpteam to have us using the same database, that would be great. One consideration I do have is the ease of setting up a working wanna-build in the absence of a full dak setup. This is very useful if you want to do test builds of the whole archive for example. Currently it's very easy to set up independent setups for testing and development work, and if there's continuing demand for that, I wouldn't want to lose it. However, I think a good first step would for both dak and wanna-build to share as much of their schemas as is reasonably possible. To that end, I'll update my wanna-build schema to use the same naming as is used in projectb as far as is practicable, and then we'll see where we can go from there. It will at the very least make merging the databases possible in the future. Regards, Roger -- .''`. Roger Leigh : :' : Debian GNU/Linux http://people.debian.org/~rleigh/ `. `' Printing on GNU/Linux? http://gutenprint.sourceforge.net/ `- GPG Public Key: 0x25BFB848 Please GPG sign your mail.
Attachment:
signature.asc
Description: Digital signature