[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnu-arch-users] Preliminary Arch Cache available

From: James Blackwell
Subject: Re: [Gnu-arch-users] Preliminary Arch Cache available
Date: Tue, 14 Sep 2004 02:11:06 -0400

>> DUDE! This sounds neat! Anything holding it up for 1.2.3?
Aaron Bentley wrote:
> Well, there are a few things it should be caching that it isn't yet. 
> Dunno when 1.2.3 will be, so it might be ready in time.

Assuming all goes well, we're looking at 1.2.2 on 24 Sept. Give a couple
weeks for wider use to catch problems, and I should be cutting a
1.2.3rc1 on 8 Oct.

Given Tom's recent pleas for more information, I think that 1.2.3 will
be a bit too fast. But maybe 1.2.4 or 1.2.5... 

>> Instead of reregistering archives with cache:, what if we just do this
>> for all archive if =arch-cache exists, and then for each archive that
>> has a location that doesn't start with path_sep? After all, archives are
>> generally small, and disk space is *much* cheaper than bandwidth.
> I wouldn't oppose that, if people wanted it.  Or we could have a 
>=default-cached configuration option, or something.

Heck. I'd be happy with a front end like
"tla register-archive [-f] --cache (location)" and storing whether or
not an archive is cached off in .arch-params. 

I think what hit me the wrong way was the cache:(location). That just
doesn't feel right to me.

>> I can only imagine how powerful this will be once it has support for
>> disconnected ops, perhaps chattering a "Warning: Original Archive not
>> available; working from cached data"
> Definitely with a warning.  Maybe a config option, not sure.


>> Side question: How does this interrelate with [ar]browse and other 
>> similiar tools that walk archives? 
> I dunno if it's useful to them.  Even if you could get a list of 
> categories from the cache, you'd still want to query the real archive to 
> make sure nothing had been added, so no point querying the cache.

Well, yeah, they want to know about new stuff too, but if you've already
got the first 300 patches in the cache, and there's only 6 new ones,
then doesn't it shave off some hundreds of round trips, since only the
six newest aren't in the cache? 

Originally, I meant the question along the lines of "when somebody
browses, does that cause stuff to get pulled down", but I suspect we're
out of sync on a lower level.

You are going to kept the whole revision, right? checksum, log,
base|patches, cacherevs, etc, right? 

James Blackwell          Try something fun: For the next 24 hours, give
Smile more!              each person you meet a compliment!

GnuPG (ID 06357400) AAE4 8C76 58DA 5902 761D  247A 8A55 DA73 0635 7400

reply via email to

[Prev in Thread] Current Thread [Next in Thread]