Gnus development mailing list
 help / color / mirror / Atom feed
From: Ted Zlatanov <tzz@lifelogs.com>
To: ding@gnus.org
Subject: Re: Trying to enter a group: "registry max-hard size limit reached"
Date: Tue, 10 May 2011 12:45:40 -0500	[thread overview]
Message-ID: <87sjsmxxdn.fsf@lifelogs.com> (raw)
In-Reply-To: <87aaeu26j4.fsf@topper.koldfront.dk>

On Tue, 10 May 2011 12:29:03 +0200 asjo@koldfront.dk (Adam Sjøgren) wrote: 

AS> On Mon, 09 May 2011 12:55:00 -0500, Ted wrote:
AS> I am pretty sure that I set gnus-registry-max-entries because the
AS> registry file got very large and it took a while to load, but I have
AS> never experienced problems with this setting before.

>> The load+save is much faster now that it's entirely managed by EIEIO
>> with a native hashtable, but certainly if you want to save memory and
>> speed Gnus up, it's better to turn the Gnus registry off.

AS> The speed-up is great, but if it makes Gnus stop working when the
AS> registry reaches the size limit, all the speed in the world ain't worth
AS> so much...

I think I've fixed it with the last commits so the pruning will DTRT to
the hard and soft limits.  There is no sort function in gnus-registry
yet so semi-random entries will be removed, but I'll add that too when I
get a chance.

On Tue, 10 May 2011 12:29:15 +0200 asjo@koldfront.dk (Adam Sjøgren) wrote: 

AS> On Mon, 09 May 2011 12:52:02 -0500, Ted wrote:

>> Now the code calls `registry-prune' but will still fail if the registry
>> is full.  I think that's the right thing to do, though I could ask the
>> user if he wants to increase `gnus-registry-max-entries'.

AS> So how does it compare to before the registry rewrite? The way it was
AS> handled then Just Worked(tm) for me for years, even with the small value
AS> I had set...

We're back to that situation now.  But it will always take up memory and
slows down article copy/move/spool operations.  So if you need registry
article marks, spam tracking through the registry (a feature I don't
think anyone needs), or splitting articles to the parent reference's
group, use it, but otherwise don't bother.

Ted




  reply	other threads:[~2011-05-10 17:45 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-05-08 21:04 Adam Sjøgren
2011-05-09  6:37 ` Tassilo Horn
2011-05-09  8:47 ` Adam Sjøgren
2011-05-09 10:54   ` Ted Zlatanov
2011-05-09 15:41     ` Adam Sjøgren
2011-05-09 16:26       ` Ted Zlatanov
2011-05-09 16:50         ` Adam Sjøgren
2011-05-09 17:55           ` Ted Zlatanov
2011-05-10 10:29             ` Adam Sjøgren
2011-05-10 17:45               ` Ted Zlatanov [this message]
2011-05-09 17:52     ` Ted Zlatanov
2011-05-09 19:44       ` Tassilo Horn
2011-05-09 20:03         ` Ted Zlatanov
2011-05-10 10:29           ` Adam Sjøgren
2011-05-10 10:29       ` Adam Sjøgren

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87sjsmxxdn.fsf@lifelogs.com \
    --to=tzz@lifelogs.com \
    --cc=ding@gnus.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).