Messages

As far as I can tell there weren't any errors reported, so far so good...

great!

Would this be a suitable addition to test with: http://wiki.linuxmce.org/index.php/Logitech_QuickCam_Sphere?

i don't care for our first tests any device template will be fine.

best regards,
Hari

Hi Guys,

i have a repo for testing online.
I need two or three helpful souls to test the following:

Alice has a fresh database
Bob does a checkin of some stuff (e.g. a dce template) with given credentials
Hari approves the batch
Alice pulls the update and verifies the dce template
Charlie modifies template wand commits with other credentials

anyone?

the lazy guys can do a sqlCVS diff via web interface to see differences between the local database and the initial repository.

sqlCVS://62.99.143.187:3999

regards,
Hari

I apologise coming off snappish... just built up frustrations..

no prob, we are all human, are we?

anyway.. thanks for the suggestions.

hope this helps.

best regards,
Hari

Just for the record, I'm using a brand new Linksys 802.11N router/AP

Just for the record, i prefer the aironet line from this vendor. As you can read in my post: some vendors even don't manage it today.
I don't know if this model is affected. But cheapo hardware often doesn't get the trick..

and yes, I am an open source developer. But I've also been in situations where communities have inherited LARGE code-bases and have been ineffective in maintaining or extending them. This is my fear with this code-base.

I don't share your fears. But don't let this go off topic. Did you even try to disable power management with iwconfig? Maybe it's lunch time for trolls?

Step one: man iwconfig:
---8<---
Examples :
                   iwconfig eth0 power period 2
                   iwconfig eth0 power 500m unicast
                   iwconfig eth0 power timeout 300u all
                   iwconfig eth0 power off
                   iwconfig eth0 power min period 2 power max period 4
--->8---
Step two: iwconfig <interface> power off
Step three: test with ping (both sides), tcpdump and the debug settings recommended in this post

best regards,
Hari

Yes, I am being a bit snappy, it's the cause of much frustration built up over trying to use this damned thing as an orbiter...and all I've ever gotten from any of you, is "Gee, try another AP." THAT'S NOT A REAL ANSWER, and I really honestly, don't buy it... but I'll try....

I mean, really.. this version of the orbiter is somewhat hobbled, are we just waiting for Pluto to fix it? or is someone actively making changes to the code?

Thom,

all people involved spend their leasure time to improve LinuxMCE. Use the source, Luke, and stop bit****.

The problem is the following: early 802.11b APs (and others, some vendors even don't manage it today) have big bugs in the power management routines. The AP should cache packets for the client while it is asleep for short periods. Bad APs didn't cache properly or simply sent the packets out while the client does not listen on the air.
So yes, the source of your flaky connection with the N770 is your AP. THIS IS THE ANSWER.

Step one: Deactivate power management and be stunned by the improvements
Step two: look at the source and improve the code
Step three: ?
Step four: profit

best regards,
Hari

i would prefer a generic upnp av server and control point implementation. I thought about using mediathomb and writing a dce device with perl and Net::UPnP, but its missing the control point features. The only free software i found is java based: http://www.cidero.com/.

So one could control the DLNA features of the PS3 or any other UPnP media renderer with any orbiter. That would be cool.

regards,
Hari

Awesome work hari, and thank you very much for expanding the wiki page! 

that was easy after your preparations. What i don't get is how to list the pending batches on the server. The LinuxMCE Admin interface doesn't print the id after checkin. Maybe it will help to use the "View batches" scripts from the admin page on the server.

We need a work flow for updates:
1.) Contact the maintainer before submission
2.) Checkin changes and report batch id to maintainer (or maintainer gets batch id from repo)
3.) Maintainer approves batch

With a test system and two or three people submitting changes we should figure out proper row ownership for the repository. I wonder if we should assign ownership of new rows to a default user. Maybe i should further investigate the permission model of sqlCVS.

I think we should have a developement and a production repository. We could synchronize them with sqlCVS (or maybe with the sqlCVS diff admin interface

@all: does anybody have information about the sqlCVS workflow @ pluto or contacts to pluto regarding sqlCVS?

@tschak909: do you have the slice ready? My public ip pool is empty, need a few days to free another ip and assign more cores and ram.

best regards,
hari

I'm one of those that did not have a backup of the sql db and know my core is dead.

Havent managed to find a way to get the db loaded from scratch.

you can use sqlCVS to fix the database. Instructions are in the Wiki.

regards,
Hari

The procedure is documented in the wiki:
http://wiki.linuxmce.org/index.php/SqlCVS#How_to_set_up_a_sqlCVS_Server

The LinuxMCE webinterface for sqlCVS works fine. If there is demand i would host a master database for testing.

regards,
Hari

Well I got a little further, I copied /usr/pluto/database/dce.sqlcvs from an existing installation, and this seems to have been the missing file.  I've dropped the database to start fresh again first.

based on your findings i set up a sqlCVS server. The procedure is documented in the wiki:
http://wiki.linuxmce.org/index.php/SqlCVS#How_to_set_up_a_sqlCVS_Server

regards,
Hari

Hi Folks,

has anybody tried the new ACT Z-Wave USB Interface ZCU[1,2]01 (http://www.act-solutions.com/HomeProProductGrid.htm)?
They seem to pop up in shops now.

best regards,
Hari

Are you trying to use  a Squeezebox as an IR blaster?

yes, the IR signal is transmitted by the Squeezebox (optional IR transmitter needed).

That would be a nice feature to have.  I think the slimserver streamer object is currently the only one that accesses the CLI for slimserver, so that may be what needs to be modified, as opposed to writing a GSD interface from scratch.

i see no problem in extending "./src/SlimServer_PlugIn/SlimServer_PlugIn.cpp". My first idea is to put the call in "SlimServer_PlugIn::ReceivedCommandForChild". So i can get the player-mac from "*pDeviceData_Impl", can't i?
But how to configure the core to use the players as IR Transmitters?

regards,
Hari

I agree, that we should discuss the matter of the right repository - espacially if it turns out that SVN does not work as expected. But please, don't discuss everything until it's dead :/

so true..

Hi Folks,

i investigated the DHCP PNP process and had the idea of using arpwatch for pnp detection. It produces log entries like this:
Oct 11 13:38:18 dcerouter arpwatch: new station 10.1.1.17 de:ad:be:ef:0:4 eth0

derive from Dhcpd-Plugin-Parser.sh:

dcerouter_37298:~# diff -u  /usr/pluto/bin/Dhcpd-Plugin-Parser.sh /usr/pluto/bin/Arpwatch-Plugin-Parser.sh
--- /usr/pluto/bin/Dhcpd-Plugin-Parser.sh       2007-03-01 04:15:34.000000000 +0100
+++ /usr/pluto/bin/Arpwatch-Plugin-Parser.sh    2007-10-11 13:40:04.000000000 +0200
@@ -23,7 +23,7 @@
                log) Type="00" ;;
        esac

-       printf "%s\n" "$Type $(date -R) $Data" >>/var/log/pluto/dhcp_pnp.log
+       printf "%s\n" "$Type $(date -R) $Data" >>/var/log/pluto/arpwatch_pnp.log
}

parse()
@@ -57,16 +57,16 @@
        line=$(parse "$inline")
        tag=$(extract_field 5 "$line")
        op=$(extract_field 6 "$line")
-       mac_found=$(extract_field 10 "$line")
+       mac_found=$(extract_field 9 "$line")
        ip_sent=$(extract_field 8 "$line")

-       if [[ "$tag" != "dhcpd:" || -z "$mac_found" ]]; then
+       if [[ "$tag" != "arpwatch:" || -z "$mac_found" ]]; then
                continue
        fi

        case "$op" in
-               "DHCPACK")
-                       log_plugin log "DHCP : MAC '$mac_found' IP '$ip_sent'"
+               "new")
+                       log_plugin log "ARPWATCH : MAC '$mac_found' IP '$ip_sent'"

                        Q="SELECT PK_Device FROM Device WHERE MACaddress='$mac_found' AND IPaddress='$ip_sent'"
                        R=$(RunSQL "$Q")

seems to work:

+ read mac ip garbage
+ RetrySend de:ad:be:ef:0:4 10.1.1.17
+ local MAC=de:ad:be:ef:0:4 IP=10.1.1.17
+ /usr/pluto/bin/MessageSend localhost 1 -1001 2 65 52 3 53 2 5 de:ad:be:ef:0:4 28 10.1.1.17

Caveats: boot options for media directors and stuff have to be specified in the standalone DHCP server.

I think this would be a nice fallback for PNP when the internal dhcp server is deactivated. What do you think about it?

regards,
Hari