Re: [freenet-dev] Re: [freenet-support] Usability improvement ideas

2003-10-31 Thread Vitenka
[EMAIL PROTECTED],UUleYfXnBfLThNmkB8dACg part might 
be the bit with which they are having trouble.  freenet:xxx URLs won't 
change that, but they will introduce a world of pain.
Freenet URLs are much more likely to be given to people in hyperlink 
form, in which case the actual form of the URL isn't particularly 
relevant.  The whole freenet:xxx thing is purely cosmetic.
Except that you also have to mandate that everyone uses localhost: 
and not everyone can.  Some people will want to use proxies, others will 
want to use a different port.

Standardising on a freenet: hyperlink is just an obvious thing to do. 
It would also make freesite pages saved to disk work more reliably if 
passed around.

I say go for it, but the way to do it is for the plugins to be done as 
separate projects and their releases included as a part of a freenet 
distribution if a stable one ever happens.

Does anyone have any decent docs on netscape plugin architecture?

___
Support mailing list
[EMAIL PROTECTED]
http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support


Re: [freenet-support] splitfiles. Again

2002-11-23 Thread Vitenka - Zen
I didn't see anything obviously related in the 
logfile.


I'll try repeating it, but 128b lock splitfiles 
take some time...


22/11/02 23:20:18, Matthew Toseland 
[EMAIL PROTECTED] wrote:

On Fri, Nov 22, 2002 at 08:03:14PM -, Vitenka 
- Zen wrote:
 Download failed.  Couldn't download enough 
blocks to reconstruct 
 segment: 0.
 required: 128
 downloaded: 128
 
 Using 537.
It'd be enormously helpful if you found that this 
was repeatable on
demand... try checking the skip local datastore 
flag to make it
redownload everything, does it happen again? If so 
I can give you a
some patched files, or a patched JAR, to try to 
catch more info about
it.

-- 
Matthew Toseland
[EMAIL PROTECTED]
[EMAIL PROTECTED]
Freenet/Coldstore open source hacker.
Employed full time by Freenet Project Inc. from 
11/9/02 to 11/1/03
http://freenetproject.org/





___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Something died on me

2002-11-17 Thread Vitenka - Zen
 java.lang.NullPointerException
  at freenet.node.LoadStats.init(LoadStats.java:74)
  at freenet.node.Main.main(Main.java:645)
What build number?

524.

Killing the lsnodes seems to have fixed it.

And I may have messed up my settings, but now my node only knows about 50 nodes, and 
never seems to 
learn about more.

Can I guage other peoples performance on freenet?
Personally - I can only ever see about four or five of the sites linked from TFE.  And 
very very 
rarely any pages linked off of them - Yodel now shows up, but the 'bankbot' page does 
not, for 
example.

Is this normal?  What settings should I be using?


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] Something died on me

2002-11-16 Thread Vitenka - Zen
Which file do I need to delete to get it back up and running this time?

Killing the rt files didn't help.

17-Nov-2002 00:13:48 (freenet.node.Main, main): starting filesystem
17-Nov-2002 00:13:53 (freenet.node.Main, main): loading data store
17-Nov-2002 00:13:53 (freenet.node.Main, main): loading routing table
17-Nov-2002 00:13:55 (freenet.node.Main, main): Unexpected Exception: 
java.lang.NullPointerException
java.lang.NullPointerException
at freenet.node.LoadStats.init(LoadStats.java:74)
at freenet.node.Main.main(Main.java:645)

java.lang.NullPointerException
at freenet.node.LoadStats.init(LoadStats.java:74)
at freenet.node.Main.main(Main.java:645)



___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] reference suicide

2002-11-11 Thread Vitenka - Zen
 Well, my rt files haven't exploded this time - but the databugs seem to be slowly 
munching through 
 my list of known nodes.
 Are there really only 16 nodes?
 And do we have to remove them (nodes we discard for whatever reason) from the list 
- shouldn't we 
 just put them to one side so that we can get them back?
 
What build number?

534

Gah - now it dd blow up the rt nodes again and fail to boot.

Deleted them, working.

What ARE those files?  And if deleting them is the only way to get the node back up, 
can't it do that on startup?


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] reference suicide

2002-11-09 Thread Vitenka - Zen
Well, my rt files haven't exploded this time - but the databugs seem to be slowly 
munching through 
my list of known nodes.

Are there really only 16 nodes?
And do we have to remove them (nodes we discard for whatever reason) from the list - 
shouldn't we 
just put them to one side so that we can get them back?


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] How big is freenet, right now, you think?

2002-11-08 Thread Vitenka - Zen
yeah, but now you only add routes if they've 
successfully responded to
a request.  To sabotage freenet, the FBI/whoever 
would have to make
requests succeed; making freenet a success.

Nah - they each add each other to the network and 
insert and request files with unique keys that no 
one else cares about.

also the doAnnounce thing...


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] How big is freenet, right now, you think?

2002-11-06 Thread Vitenka - Zen
No.  If other nodes can push references into your routing table, the
whole anonymity aspect of freenet goes up in smoke as the
CIA/FBI/whoever fixes everyone's routing tables to point to them.

So you add the FBI nodes to your current routes and the routes that good nodes are 
sending you.

I don't see how that is different from now.


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Build 530

2002-11-06 Thread Vitenka - Zen
Hmm. Try 531; if this doesn't work... are you in the default
seednodes.ref?

No - I started with that, built up a big list of nodes and have now remembered to save 
it so that when my routing 
table gets eaten (a distressingly common occurence - roughly every other boot) I can 
reseed it.

My node (build 530) now knows about 95 other nodes (down from the seeded 146) and 
seems healthy.

Although thses errors keep showing up:
Nov 6, 2002 6:38:41 PM (freenet.diagnostics.StandardDiagnostics, main): Found file for 
var socketTime type: 3
Nov 6, 2002 6:38:41 PM (freenet.diagnostics.StandardDiagnostics, main): Did NOT find 
file for var socketTime type: 4

 What are those limits, and how do they work?
Freenet contains numerous arguably totally arbitrary limits, partly due
to java's inability to give us any real load statistics.

So, twiddle 'em till it works then tell the list so they can copy me.  Got it :)

 Second problem.  Downloading the jargon file single 
 html file, the FEC error popped up in my browser 
 again - and then non human readable stuff started 
 filling my browser window.
Will look at this.
 Annoyingly the system overloaded my CPU and I had 
 to crash it before I could capture it.
Eh?
I didn't manage to get a copy of the junk it pasted into the browser - but it looked 
like [human ureadable]
java.lang.eception.[human unreadable]

Can the 'FEC download' to browser be made more understandable?  Or, preferably, 
completely transparent?

 Also, had some interesting speculation in the irc 
 channel - why does the graph fror your datastore 
 include an entry for 1gig+ keys?  Has anyone ever 
 insderted such a key - and what would happen if 
 someone inserted 10 20gig keys or so and requested 
 them from elsewhere on the network to forve their 
 propogation?
No, because nodes only cache keys up to 1/200th the size of the
datastore. When you include fields, this means that to cache a 1gb key
you'd need a 210GB node or so - certainly not the common case. If you
want to insert really big files, use fishtools or fproxy to insert them
as FEC splitfiles - and tell us how you get on.

Ok - that makes sense.
A similar question still occurs with more sanely sized keys - insert a large number of 
5 meg keys and simultaneously 
request them from a large number of clients -watch that old data go bye bye.




___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] Build 530

2002-11-05 Thread Vitenka - Zen
Seems vastly improved over 529 and slightly 
improved over 527.

It didn't eat my routing table yet.

Sadly, I left it running for eight hours, and it 
slowly but surely crept up to using 100% CPU time 
again.

What are those limits, and how do they work?



Second problem.  Downloading the jargon file single 
html file, the FEC error popped up in my browser 
again - and then non human readable stuff started 
filling my browser window.

Annoyingly the system overloaded my CPU and I had 
to crash it before I could capture it.


Also, had some interesting speculation in the irc 
channel - why does the graph fror your datastore 
include an entry for 1gig+ keys?  Has anyone ever 
insderted such a key - and what would happen if 
someone inserted 10 20gig keys or so and requested 
them from elsewhere on the network to forve their 
propogation?


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] Yay! All hail 530!

2002-11-05 Thread Vitenka - Zen
Running for over 12 hours now - knows about 143 nodes, 882 requests per hours served, 
30% success 
rate on answering queries.

Still al ot of these errors:
19:57:58
I (freenet.ConnectionHandler@1346e80) wasn't terminated properly! Doing it now..
null

But they suggest that fred is coping with cleaning up behind itself, even if it's 
doing it a little 
belatedly.

Current routingTime:  50ms.  Whatever that means.

Used space in the data store: 961,008 KB  - slowly growing.


Yup, I think that I am, finally, stable.

Thanks whoever it was told me how to save out my noderefs.txt   Hopefully I'll be able 
to recover 
if things go wrong again.


My only SLIGHT concern is the system load.  I'm at 50% - which seems rather high.  
Then again, if 
I'm serving 15 requests per minute, I'd better expect a slightly higher load than my 
apache which 
serves about 15 a day.  And since it's idle priority it should give way to my real 
applications.  
So I guess it's ok.

Now - what compelling reason do I have to upgrade past this version?




___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] windows 0.5.0.4 monolithic installer updated

2002-11-05 Thread Vitenka - Zen
 Secret options are bad :)  I still seem to have
 an un-needed 'temporary directory' setting -
 which it warns me about on every startup.
 Can you expound on this one?

28-Oct-02 17:26:01 (freenet.node.Main, main): 
WORKAROUND: Ignoring obsolete fproxy.* lines in 
freenet.conf/ini. You can remove them.

 Possibly unrelated, but what is the option to
 allow arbitary clients to connect on the client
 port or the web port - or is this incredibly
 unsafe?
 In the windows client, the option is on the
 advanced page in the FCP access box. For
 mainport, there is no option in the config
 program yet.

Not sure I understood that.  All I see is a line 
where you can give a list of addresses which can 
connect.  'must include localhost' it says.  No 
(obvious) way to set up 'allow any'

 Hmm.  If we have dyndns would this make routing
 via dynamic IP a more sensible thing?  In which
 case, might renaming transient mode to 'client
 only' or something similar help?

Was wondering whether, if nodes have a dynamic DNS, 
it might become easier to contact them - less 
chance of sending requests to IP addresses which 
are no longer populated.

 Fproxy will tell you when higher build numbers
 are available.

So it does.

Can I ask where I find a changelog?  I don't need 
or want to stay on the cutting edge, not after what 
529 did to me.

 The problem with that is that sane settings vary
 widely. For example, at home I can upload at
 about 3 times what is safe for me to upload here,
 due to additional congestion, but I can download
 about twice as fast here as at home.

Sure, but if you ask 'cable or 56k modem?' and then 
set 128kbit or 32kbit respectively, you'd get a lot 
more sensible settings straight away and let the 
power users reconfigure it to cope with their own 
environments.

 Also, on windows, offerring an option to set the
 registry keys that frost advises (that set your
 socket limits up for server working) might be a
 good idea.
what?

I'll try and find the freesite again (one problem 
with freenet, URIs are hardly memorable)

There is a problem when you try to run frost on 
windows 9x - and in general a problem with the 
default socket settings in freenet on win9x.  The 
symptom is winsock errors (or in frost: java.socket 
exception: out of buffer space) and it pretty much 
kills your network connection.

There are a couple of registry settings:  One to 
set your MTU and one to set the max number of 
sockets.  Autoconfiguring these may help.  (Though 
give a warning, since gamers especially tend to 
change those things)


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] windows 0.5.0.4 monolithic installer updated

2002-11-05 Thread Vitenka - Zen
Dammit.  System crashed, and the databugs ate my routing table again.

A thousand thousand blessings upon the person who explained how to export my routes :)

Upgraded to 531 (since I crashed, I might as well)


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Nothing but RNFs

2002-11-04 Thread Vitenka - Zen
Now it's nothing but RNFs - can't contact any other servers at all; and at 35% system 
load despite no requests going either 
way!

Nothing in the visible logs.

Spoke too soon, now the load it at 47%

Blegh - my routing table has gotten corrupted AGAIN.



Ok - replies to messages which I can read on the web archive but which I can't see on 
my email server...  Blasted slow 
provider.

Re: Trying to help with the routing table problems.  Yes, I did that, deleted the rt 
files - didn't delete the lp files, not 
sure what they are.  Didn't help  - still getting the no nodes known problem.  
Reseeded - same problem.

Yes - I am using suns jre.

Oddly, even though my node knows about no other nodes - I am still serving up data to 
people.

As to strange ports being open.  Now it seems ok - but they were definately open at 
one poine and it was definately the 
freenet.exe that was listening on them.


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] How big is freenet, right now, you think?

2002-11-04 Thread Vitenka - Zen
Ok, that all makes sense.  I missed the part about 
chooing which host to ask questions about based 
upon lexicographical closeness to the keys it has 
returned before.

Do hosts transmit routing tables to each other 
periodically?  Obviously you wouldn't want to trust 
a routing table you have been sent - but you could 
at least harvest new node addreses from it.

I don't think the 'time until your routing table is 
sensible' is actually my problem though - my whole 
routing table keeps getting eaten.  I'm rolling 
back to build 527.


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Network Error

2002-11-04 Thread Vitenka - Zen
How do I export that list?  --export only gives *my* node id...

I'd like to save it to aid recovery next time my routing table gets eaten by datastore 
bugs :)

John E. Mayorga ([EMAIL PROTECTED]) wrote:

 Attempts were made to contact 5 nodes.
 
  5 were totally unreachable. 
 0 restarted. 
 0 cleanly rejected. 

 Try reseeding your node, and if that doesn't
 work, contact [EMAIL PROTECTED]

Reseeding means to get seednodes.ref from someone else whose node is
working better than yours.  Mine are available at
http://wooledge.org/~greg/seednodes.ref is you'd like to try them.

-- 
Greg Wooledge  |   Truth belongs to everybody.
[EMAIL PROTECTED]  |- The Red Hot Chili Peppers
http://wooledge.org/~greg/ |





___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Lots of java errors in logfile

2002-11-03 Thread Vitenka - Zen
 23:42:48
 Error while receiving message freenet.Message: Accepted 
 @freenet.ConnectionHandler@465c16 @ c9a6523d8573355c in state InsertRequest 
 Pending Transfer @ c9a6523d8573355c
 java.lang.IllegalStateException: null stream from buffer
Hmmm. Can you get a stack trace for this one from the log file?

Uhh... I'm getting a three day turnaround time on this mailing list - I don't appear 
to have an 
entry with this exact time in my log file, and they've grown too big to look through 
manually I'm 
afraid.

I haven't seen that error recently.

I'd also think that the 'no decoder found' erro IS significant - it shows up in the 
browser and 
stops me recieving split files.


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] windows 0.5.0.4 monolithic installer updated

2002-11-03 Thread Vitenka - Zen
Nodeconfig refuses point blank to run on my machine as of main build 529.

* make it so that nodeconf can change every option

Secret options are bad :)  I still seem to have an un-needed 'temporary directory' 
setting - which it warns me about on every 
startup.

* have nodeconf just edit the config in place (effectively, have it treat
%'s and not have to be manually rewritten)

It isn't doing this already?

* fix import/export ref on bunny

Can we have an export all known references option?  At least until the routing table 
stops getting eaten by the datastore 
bugs :)

* popup window when # of threads is changed to equal 0

Is that a bad option to set then?

* make windows installer using windows installer.

Strongly disagree - the nullsoft installer is a fine and accepted piece of work, 
whereas the default windows installer is a 
horrible piece of junk.  Unless you are having some major problems with the nullsoft 
explorer, what is the incentive to 
change?  If you are worried about transparence then most people would probably be 
happy with a zip file and a 'run this 
first' file.

* allow selection of java's if don't want first one found out of the
  registry

Please, yes please.  Give us a list of all of them and let us choose.  Would it be 
possible to list what version each java 
is, since it's not always obvious from the directory.

* have nodeconfig popup a window when using an invalid ip address
(192.168.*.*, 10.*.*.*, phlemball.aol)

Possibly unrelated, but what is the option to allow arbitary lients to connect on the 
client port or the web port - or is 
this incredibly unsafe?

* rewrite in c++ so I don't have to deal with java sh*t.
* automatically register dyndns if they don't have it already (until arks)
(freenetnode.com would probably be setup fairly easily)

A dyndns for frenetnode.com would be kinda nice to have built in.  Make sure the 
explanation of what this is, is 
comprehensible to the average user :)

Hmm.  If we have dyndns would this make routing via dynamic IP a more sensible thing?  
In which case, might renaming 
transient mode to 'client only' or something similar help?

* add messagebox before stopping the node when updating snapshot
* add update option on systray app

How often are updates checked for?  And is it done just by getting a freenet message 
or do we poll or is it via some out of 
band method?

* allow for negative # of threads (add checkbox. Give big warning when set
since FastThreadFactory = evil)

Uh - negative threads?  What would that be?

* fix the file naming of freenet-webinstall.exe to
freenet-webinstall-{$VER}.exe

Uh - surely the webinstaller installs whatever the latest version of freenet there is?


As to using the monolithic installer:  A couple of times downloading via the normal 
update or web update I got an out of date 
version of the software - so I dropped back to the monolithic one, which worked once I 
altered the version number in the 
url...

Right, what else would I request?   Some more sensible defaults would be good - 
perhaps default choke of bandwidth to the 
kinds of levels that make sense on a cable / DSL connection.

Also, on windows, offerring an option to set the registry keys that frost advises 
(that set your socket limits up for server 
working) might be a good idea.


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] version - 139 - what are all thes extra ports?

2002-11-02 Thread Vitenka - Zen
I have listeners (blocked at firewall) on ports 135 (rpcss.exe) 137-139 
(freenet.exe) and 1025 (rpcss again)

What are they?  They aren't being caused by the java code, since that all goes 
through javaw


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] Bugs ate my routing table

2002-11-02 Thread Vitenka - Zen
Win 98, Java 1.4_2, Build 529
Routing rtable corrupt
Please delete all rt files and restart freenet
Do so, freenet starts - all requests cannot make it off the local node.
Reseed node; nothing improves.

The network has lost me completely, I presume.



___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Nothing but DNFs

2002-11-01 Thread Vitenka - Zen
No reply at htl 15.
Trying htl 40  Not found.

But no one can see me either.

Is it posisble that the network has just gotten so 
big (and full of holes) that an insert of 5 and a 
request of 40 simply don't reach far enough across 
the network?

There's definately a lack of flow - most of my 
requests are timing out with '25 hosts queried - 19 
didn't connect, 6 cleanly failed'

But people seem to be using my node, and I can get 
a number of freesites.

Still getting awful performance especially if I 
fire up frost.

I also have a complaint to whoever said that 100 
idling threads is harmless.
Not on windows it isn't, certainly not in java on 
windows.  Threads are rpocesses and have horrible 
overhead.


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] 529 seems to be a step backwards

2002-11-01 Thread Vitenka - Zen
Still running on windows 98, latest jre, attempting to run a permanent node.

One the one paw, I seem to have much better connectivity - many more pages are being 
found.  But I 
think that's mainly due to being (sorta) online for longer now and having 700meg of 
cached datastore.

On the other - HOW can I have *three* connections listening on the same port?  That's 
not even supposed 
to be possible.  We're leaking dead connections - and I suspect that once I leak 
enough I'll be totally 
unable to connect to the net until I reboot.

I'm using the same config files as my pervious version (didn't install new config 
files and then try to 
alter them) - this is because:

NODECONFIG caused an invalid page fault in
module KERNEL32.DLL at 0177:bff9b720.
Registers:
EAX= CS=0177 EIP=bff9b720 EFLGS=00010246
EBX= SS=017f ESP=0064e5f4 EBP=0064e614
ECX= DS=017f ESI=0064f650 FS=f05f
EDX=4d04 ES=017f EDI=0001 GS=
Bytes at CS:EIP:
89 11 89 41 04 8b c7 5f 5e 5b 8b e5 5d c2 10 00 
Stack dump:
0040bc88 0040bd68  0200 9a08 fff6 8000  0064f664 
0040119c  
0064f650 0064f648  0040bc88 0040bd68 



___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] How big is freenet, right now, you think?

2002-11-01 Thread Vitenka - Zen
I mean, how many nodes?  Because I see a big problem.

If I upload a file with an HTL of 5 (10 being the 'max sensible' the client 
recommends)

And you download with an HTL of 15

Then if there are at least 16 sites it becomes posisble for you to fail to 
find my insert - and once there are 50 node it becomes likely.

If peoples routing tables are full, and there are a few thousand nodes... how 
will new content ever be found?  The existing stuff is ok, because it's 
already mirrored.

Do you have to rely upon having five or six logically distant people 
requesting your file and going away unsatisfied each day so that everyone else 
can see it?  Or have I misunderstood the white paper?

{Which has far too few 'node can do this for added privacy' with very little 
expansion on what the impacts of your average node not doing it means}


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Lots of java errors in logfile

2002-11-01 Thread Vitenka - Zen
529 build now:
Guess what, the errors have come back! :)

Any advice at all would be great - 

Should I change jre?  I found some advice on setting my max sockets up for 
frost, let's see if that helps.

From the growth of my datastore, it looks like other people are finding me, 
slowly.

Environment:
JVM Name Java HotSpot(TM) Client VM
JVM Version 1.4.1_01-b01 Operating System
Windows 98 OS Version 4.10 
Total size of the data store 2 GB
Free space in the data store 1,371,712 KB
Used space in the data store 725,440 KB

Logfile:

tipleHttpServletContainer, QThread-92): I/O error in servlet
Nov 1, 2002 3:41:58 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-18): I/O 
error in servlet
Nov 1, 2002 3:43:40 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-104): I/O 
error in servlet
Nov 1, 2002 3:44:42 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-71): I/O 
error in servlet
Nov 1, 2002 3:44:42 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-22): I/O 
error in servlet
Nov 1, 2002 3:45:42 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-110): I/O 
error in servlet
Nov 1, 2002 3:46:12 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-101): I/O 
error in servlet
Nov 1, 2002 3:46:42 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-122): I/O 
error in servlet
Nov 1, 2002 3:46:51 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-60): I/O 
error in servlet
Nov 1, 2002 3:53:20 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-136): I/O 
error in servlet
Nov 1, 2002 3:57:17 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-169): I/O 
error in servlet
Nov 1, 2002 3:59:01 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-173): I/O 
error in servlet
Nov 1, 2002 4:02:19 PM (freenet.client.http.FproxyServlet, QThread-185): 
Error sending data to browser: java.net.SocketException: Connection reset by 
peer: socket write error
Nov 1, 2002 4:02:19 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-185): I/O 
error in servlet
Nov 1, 2002 4:07:20 PM (freenet.client.http.FproxyServlet, QThread-169): 
Error sending data to browser: java.net.SocketException: Connection reset by 
peer: socket write error
Nov 1, 2002 4:07:20 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-169): I/O 
error in servlet
Nov 1, 2002 4:18:56 PM 
(freenet.interfaces.servlet.MultipleHttpServletContainer, QThread-269): I/O 
error in servlet


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] my mp3's

2002-10-31 Thread Vitenka - Zen
Yes it's daft, yes it's silly yes he should have 
backups.

It's STILL a major PR blunder on this softwares 
part.  You can just see the reports in three months 
time, if you are lucky:
From it's beginnings as an aplication which wiped 
out your data and used up all your processor time 
and choked so many sockets that your email server 
couldn't even get the 'delete' message through 
properly, freenet has progressed towards its 
promise of being a truly free-as-in-speech 
network

And as for 'read the docs' - you're the ones who 
made this crossplatform and invited all the windows 
people like me who know that four colour glossy 
help files are NEVER helpful.

Ok, humourous rant aside:  Someone mentioned 
settings in the readme to help the browser getting 
choked off problem?  I see no such mention in my 
readme - does anyone have the text?


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Lots of java errors in logfile

2002-10-30 Thread Vitenka - Zen
Hmmm.

Ok, let's see now.

Routing seems to have settled down - traffic got 
resonable.

I've gotten the temp dir sorted (did you know that 
each install wiupes that field, even though you 
tell it to keep your old ini file?)

I've installed 0.5.2.

I've installed absolutely latest jre

Java errors seem to be gone.

I am STILL getting 'out of winsock resources'
Frost is reporting 'cannot upload message' though 
without the 'buffer full' in the title.

I'm, basically, stumped again.

Can anyone see my test-publish?
freenet:SSK@6-IQgELFYXv~BtumgMbm7shPe5wPAgM/Skulk//


___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



[freenet-support] Lots of java errors in logfile

2002-10-29 Thread Vitenka - Zen
Similar to my previous message, but new errors in 
my logs.


29-Oct-02 17:57:55 
(freenet.node.ds.FSDataStoreElement
$KeyInputStreamImpl, Finalizer): Please close() me 
manually in finalizer: Key: 
9a386f0597dd9d2a9afb7bf35a1d473692d050b50f0203 
Buffer: freenet.fs.dir.NativeFSDirectory
$ExternalNativeBuffer@4d2af2 New: true ( 0 of 2049 
read)
java.lang.IllegalStateException: unclosed
at freenet.node.ds.FSDataStoreElement
$KeyInputStreamImpl.finalize
(FSDataStoreElement.java:310)
at java.lang.ref.Finalizer.invokeFinalizeMethod
(Native Method)
at java.lang.ref.Finalizer.runFinalizer(Unknown 
Source)
at java.lang.ref.Finalizer.access$100(Unknown 
Source)
at java.lang.ref.Finalizer$FinalizerThread.run
(Unknown Source)
===
StateChain started at Tue Oct 29 17:56:46 GMT+00:00 
2002
Current state: Request Done @ a3ad6a2bf140dbaa
===
29-Oct-02 17:57:55 
(freenet.node.ds.FSDataStoreElement
$KeyInputStreamImpl, Finalizer): Please close() me 
manually in finalizer: Key: 
34532ba3369cf57cbc1401ae2a67a0cc043cb6250f0203 
Buffer: freenet.fs.dir.NativeFSDirectory
$ExternalNativeBuffer@48769b New: true ( 0 of 2049 
read)
java.lang.IllegalStateException: unclosed
at freenet.node.ds.FSDataStoreElement
$KeyInputStreamImpl.finalize
(FSDataStoreElement.java:310)
at java.lang.ref.Finalizer.invokeFinalizeMethod
(Native Method)
at java.lang.ref.Finalizer.runFinalizer(Unknown 
Source)
at java.lang.ref.Finalizer.access$100(Unknown 
Source)
at java.lang.ref.Finalizer$FinalizerThread.run
(Unknown Source)
===
StateChain started at Tue Oct 29 17:56:53 GMT+00:00 
2002
Current state: Request Done @ 952765ceb4723f69
===
29-Oct-02 18:20:45 (freenet.ConnectionHandler, 
Finalizer): I (freenet.ConnectionHandler@16) 
wasn't terminated properly! Doing it now..
 End of forwarded message 

 



___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support



Re: [freenet-support] Connections running amok

2002-10-29 Thread Vitenka - Zen
Similar problem - I am seeing a vast number of open sockets - all listening 
on bound to 0.0.0.0:8481 and listening for connections - but not seeing any 
traffic.

Surely there should only be one socket open listening on that port?

I also seem to be seeing extremely low transfer rates on all of the actual 
traffic coming in - 17 bytes per minute kind of transfer rates.  I have 
turned the timeouts down, but this does not seem to have helped.

Are there a vast number of ill behaved clients out there or have I 
misunderstood how this is supposed to work?

29/10/02 13:47:42, Doug Bostrom [EMAIL PROTECTED] wrote:

Greetings,

Checking my node this morning, (after waiting for top to come up, swap 
swap swap) I found I had some 155 java 
processes going. Any way to control this? Right now I can either run a 
node or work, not both. Maxnodeconnections 
seems to have no effect on the number of connections the node tries to 
support. Nor does bandwidth control seem 
to have any effect. Obviously there's going to be a storm of activity when 
a public announcement is broadcast 
about Freenet but just now it seems there are too few persistent nodes to 
keep up. We use our net connection to 
make a living here, so this comes down to a hard choice. If folks like me 
have to continue taking nodes down in 
order to work it will exacerbate the shortage of persistent nodes. Thanks 
for any hints you can provide.

--
There's no telling how many wars it will take to secure freedom in the 
homeland.
- George W. Bush



___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support






___
support mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/support