t of that issues in
jSite... :)
David
-- next part --
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: not available
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/b6f506b7/attachment.pgp>
t attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: not available
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/a3abed5c/attachment.pgp>
ubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: not available
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/c9495183/attachment.pgp>
toad at amphibian.dyndns.org
> >Freenet Project Official Codemonkey - http://freenetproject.org/
> >ICTHUS - Nothing is impossible. Our Boss says so.
>
> Ian Clarke: Co-Founder & Chief Scientist Revver, Inc.
> phone: 323.871.2828 | personal blog - http://locut.us/blog
>
--
Matthew J Toseland - toad at amphibian.dyndns.org
Freenet Project Official Codemonkey - http://freenetproject.org/
ICTHUS - Nothing is impossible. Our Boss says so.
-- next part --
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/11407218/attachment.pgp>
Matthew Toseland wrote:
We must place a moratorium on all but the most trivial "improvements"
to load balancing unless they have been simulated.
That could be another month. Michael has rightly insisted on doing a
fairly low level simulation, which has resulted in us rearchitecting a
lot of the
l Codemonkey - http://freenetproject.org/
ICTHUS - Nothing is impossible. Our Boss says so.
-- next part ------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/f3c8eaa0/attachment.pgp>
appens if 30 seconds runs out whilst the user is filling
> > in
> > the 'Insert File' form at the bottom?
I'm not sure that auto-refreshing the page is a good idea either.
> >
> > Dave
-- next part --
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/fb18b881/attachment.pgp>
odemonkey - http://freenetproject.org/
ICTHUS - Nothing is impossible. Our Boss says so.
-- next part --
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/144e14da/attachment.pgp>
oject Official Codemonkey - http://freenetproject.org/
ICTHUS - Nothing is impossible. Our Boss says so.
-- next part --
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/0c6f1a74/attachment.pgp>
it is terrible slow now, cant get any slower for inserts
PRO: Simple!
CON: Whole network may be dragged down by slow nodes.
___
Devl mailing list
Devl@freenetproject.org
http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
We could try it on the basis that simpler is, all else being equal, better.Ian.On 10 Aug 2006, at 16:19, Matthew Toseland wrote:On Fri, Aug 11, 2006 at 12:08:11AM +0100, Matthew Toseland wrote: Specifically:Metric M, lower is better. M could be proportion of requests rejected,or it could be time fo
othing is impossible. Our Boss says so.
Ian Clarke: Co-Founder & Chief Scientist Revver, Inc.
phone: 323.871.2828 | personal blog - http://locut.us/blog
-- next part --
An HTML attachment was scrubbed...
URL:
<https://emu.freenetproject.org/pipermail/devl/attachme
On Fri, Aug 11, 2006 at 12:08:11AM +0100, Matthew Toseland wrote:
>
> Specifically:
>
> Metric M, lower is better. M could be proportion of requests rejected,
> or it could be time for a successful request.
>
> Distance d = routing distance from target to peer.location.
>
> If peer.M < median.M
On Thu, Aug 10, 2006 at 03:07:07PM -0700, Ian Clarke wrote:
> Ok, lets try it - but unless there is an obvious marked improvement,
> we must *remove* it (rather than just continuing to add complexity).
0. Well, the *simplest* thing would be to remove load balancing
entirely, and rely solely on r
Ok, lets try it - but unless there is an obvious marked improvement, we must *remove* it (rather than just continuing to add complexity).Ian.On 10 Aug 2006, at 14:50, Matthew Toseland wrote:On Thu, Aug 10, 2006 at 11:43:32AM -0700, Ian Clarke wrote: Why use a median, rather than a mean?I'm concerne
d argument to me: If
>>> all
>>> our peers are slow, that DOES NOT mean we shouldn't send requests to
>>> them. Because we have load propagation, it is entirely legitimate to
>>> send requests ANYWAY, and let upstream nodes deal with the
>>> sl
emonkey - http://freenetproject.org/
ICTHUS - Nothing is impossible. Our Boss says so.
-- next part --
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/44d3dfe4/attachment.pgp>
On Thu, Aug 10, 2006 at 11:43:32AM -0700, Ian Clarke wrote:
> Why use a median, rather than a mean?
>
> I'm concerned about this, it increases (yet again) the complexity of
> our load balancing, it hasn't been simulated, and I think there will
> inevitably be unpredictable results. For exampl
On Thursday 10 August 2006 16:01, Matthew Toseland wrote:
> > The HTTP Refresh header isn't part of the HTTP spec, and it doesn't
> > work on Konqueror.
> I thought the right way to do it was with a Location header?
Nope, "Location" is only for redirects. With "pure" HTTP you can not do
a refres
On Thursday 10 August 2006 19:51, Matthew Toseland wrote:
> I agree that it's a good place to add l10n hooks. Unfortunately it
> doesn't have that functionality at the moment.
Hooking in a ResourceBundle would only take minutes... but the majority
of changes would have to be all over the node. T
On Thursday 10 August 2006 13:06, David Sowder (Zothar) wrote:
> It seems to me that you've just re-introduced the array that was
> removed in r10005, a changeset that seems to be the result of
> deciding that generating raw HTML is bad for some reason and
> replacing it with something more compli
On Thursday 10 August 2006 10:24, bombe at freenetproject.org wrote:
> Author: bombe
> Date: 2006-08-10 09:24:28 + (Thu, 10 Aug 2006)
> New Revision: 10018
>
> Modified:
>trunk/freenet/src/freenet/clients/http/QueueToadlet.java
> Log:
> refresh queue page every 30 seconds
>
> Modified: tr
Why use a median, rather than a mean?I'm concerned about this, it increases (yet again) the complexity of our load balancing, it hasn't been simulated, and I think there will inevitably be unpredictable results. For example, what happens if *all* nodes have an AcceptedFrac below the median? This
key - http://freenetproject.org/
> ICTHUS - Nothing is impossible. Our Boss says so.
Ian Clarke: Co-Founder & Chief Scientist Revver, Inc.
phone: 323.871.2828 | personal blog - http://locut.us/blog
-- next part --
An HTML attachment was scrubbed...
URL:
<https://emu.freenet
On Thu, Aug 10, 2006 at 10:27:15AM -0700, Ian Clarke wrote:
> On 10 Aug 2006, at 09:38, Matthew Toseland wrote:
> >On Thu, Aug 10, 2006 at 09:27:32AM -0700, Ian Clarke wrote:
> >>On 10 Aug 2006, at 04:06, David Sowder (Zothar) wrote:
> >>>It seems to me that you've just re-introduced the array that
On 10 Aug 2006, at 09:38, Matthew Toseland wrote:On Thu, Aug 10, 2006 at 09:27:32AM -0700, Ian Clarke wrote: On 10 Aug 2006, at 04:06, David Sowder (Zothar) wrote: It seems to me that you've just re-introduced the array that was removed in r10005, a changeset that seems to be the result of decidi
28 | personal blog - http://locut.us/blog
-- next part --
An HTML attachment was scrubbed...
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20060810/50556b39/attachment.html>
-- next part --
A non-text attachment was scr
On Thu, Aug 10, 2006 at 09:18:06AM -0400, Ed Tomlinson wrote:
> On Wednesday 09 August 2006 10:49, Michael Rogers wrote:
> > Matthew Toseland wrote:
> > > Because we have load propagation, it is entirely legitimate to
> > > send requests ANYWAY, and let upstream nodes deal with the slowness -
> >
On Thu, Aug 10, 2006 at 09:27:32AM -0700, Ian Clarke wrote:
> On 10 Aug 2006, at 04:06, David Sowder (Zothar) wrote:
> >It seems to me that you've just re-introduced the array that was
> >removed in r10005, a changeset that seems to be the result of
> >deciding that generating raw HTML is bad f
On 10 Aug 2006, at 04:06, David Sowder (Zothar) wrote:It seems to me that you've just re-introduced the array that was removed in r10005, a changeset that seems to be the result of deciding that generating raw HTML is bad for some reason and replacing it with something more complicated, but with no
will also help with internationalization.
Ian.
Ian Clarke: Co-Founder & Chief Scientist Revver, Inc.
phone: 323.871.2828 | personal blog - http://locut.us/blog
-- next part --
An HTML attachment was scrubbed...
URL:
<https://emu.freenetproject.org/pipermail/devl/at
On Wednesday 09 August 2006 10:49, Michael Rogers wrote:
> Matthew Toseland wrote:
> > Because we have load propagation, it is entirely legitimate to
> > send requests ANYWAY, and let upstream nodes deal with the slowness -
>
> I'm inclined to agree with this argument, but I'm not sure it follows
* Matthew Toseland <[EMAIL PROTECTED]> [2006-08-10 15:01:06]:
> On Thu, Aug 10, 2006 at 12:35:34PM +0100, Dave Baker wrote:
> > On Thursday 10 August 2006 10:24, [EMAIL PROTECTED] wrote:
> > > Author: bombe
> > > Date: 2006-08-10 09:24:28 + (Thu, 10 Aug 2006)
> > > New Revision: 10018
> > >
>
On Thu, Aug 10, 2006 at 12:35:34PM +0100, Dave Baker wrote:
> On Thursday 10 August 2006 10:24, [EMAIL PROTECTED] wrote:
> > Author: bombe
> > Date: 2006-08-10 09:24:28 + (Thu, 10 Aug 2006)
> > New Revision: 10018
> >
> > Modified:
> >trunk/freenet/src/freenet/clients/http/QueueToadlet.jav
On Wednesday 09 August 2006 10:49, Michael Rogers wrote:
> Matthew Toseland wrote:
> > Because we have load propagation, it is entirely legitimate to
> > send requests ANYWAY, and let upstream nodes deal with the slowness -
>
> I'm inclined to agree with this argument, but I'm not sure it follows
It seems to me that you've just re-introduced the array that was removed
in r10005, a changeset that seems to be the result of deciding that
generating raw HTML is bad for some reason and replacing it with
something more complicated, but with no clear to me advantage.
PeerNodeStatus seems a bit
On Thursday 10 August 2006 10:24, [EMAIL PROTECTED] wrote:
> Author: bombe
> Date: 2006-08-10 09:24:28 + (Thu, 10 Aug 2006)
> New Revision: 10018
>
> Modified:
>trunk/freenet/src/freenet/clients/http/QueueToadlet.java
> Log:
> refresh queue page every 30 seconds
>
> Modified: trunk/freene
It seems to me that you've just re-introduced the array that was removed
in r10005, a changeset that seems to be the result of deciding that
generating raw HTML is bad for some reason and replacing it with
something more complicated, but with no clear to me advantage.
PeerNodeStatus seems a bi
38 matches
Mail list logo
