[darcs-users] GSoC: network optimisation vs cache?

Eric Kow kowey at darcs.net
Sun Apr 11 09:33:43 UTC 2010


Hi everybody,

I'd like to ask for some feedback on prioritising work for the Google
Summer of Code.

All other things being equal, which of these two projects should have
the higher priority for GSoC?

1. optimising network performance: making darcs get and darcs pull
   much faster over HTTP and perhaps SSH

   - darcs optimize --http (would help all Darcs likely >= 2.6)
   - the smart server

2. cache cleanup: rethinking and rebuilding the Darcs global cache

   - automatically expiring bad cache paths
   - avoiding such paths in the first place

On the one hand the network optimisation would have the most visible
effect, hopefully a very pronounced one.  The first impression people
have of darcs is they (eg) darcs get --lazy http://darcs.haskell.org/ghc
and it takes a couple of minutes.

On the other hand, the cache work is actually quite appropriate to a
summer project because I think it needs to have somebody with a big
block of time to focus on this issue and that it's the kind of work that
could otherwise be perpetually preempted by something more urgent

What do you think?  We'll be discussing other aspects of this choice
off-list among folks listed on http://wiki.darcs.net/DarcsTeam, but I
thought it was important that you all had some input into the technical
choices behind this.

Thanks!

Eric

PS. Perhaps we'll be fortunate this year and successfully make a case
for Darcs having *two* of the precious Haskell.org slots or maybe even
three so we can take on the GUI/library work as well.

-- 
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 195 bytes
Desc: not available
URL: <http://lists.osuosl.org/pipermail/darcs-users/attachments/20100411/6811034d/attachment.pgp>


More information about the darcs-users mailing list