http://talk.jeelabs.net/forum/2 Discussion about Jee Labs projects en-US Sat, 11 Sep 2010 02:55:21 +0000 http://bbpress.org/?v=1.0.2 q http://talk.jeelabs.net/search.php http://talk.jeelabs.net/topic/486#post-2722 Sat, 11 Sep 2010 00:16:21 +0000 David 2722@http://talk.jeelabs.net/ <p>I'd love to take you up on your kind offer, but I need to first resolve an issue; the first JLv2 I received had a bad RF section but the (8Mbit!) DF seems to be OK. The replacement JL that Paul@Modern Device sent me (I think it has the new 4Mbit memory) may have bad memory but the RF12 is OK. I've not sent the first bad JL back to Paul yet - I'm tempted to try to swap the memory chip over but I'm not up to working on SMD chips.</p> http://talk.jeelabs.net/topic/486#post-2720 Fri, 10 Sep 2010 21:21:34 +0000 Jean-Claude 2720@http://talk.jeelabs.net/ <blockquote> <p>Why was the 1Mb flash dropped?</p> </blockquote> <p>It's no longer produced. I wasn't too happy about that either. I have 4 chips left - I can retrofit one for you, if you send me your JL, but that's about it...</p> http://talk.jeelabs.net/topic/486#post-2718 Fri, 10 Sep 2010 21:08:55 +0000 David 2718@http://talk.jeelabs.net/ <p>This isn't working for me. I'm not sure if I have a bad JLv2, but when I do Serial.print (info); within df_initialize(), I get "0" instead of what was expected in the definition of DF_DEVICE_ID.</p> <p>Why was the 1Mb flash dropped? I could really have used the additional space.</p> http://talk.jeelabs.net/topic/486#post-2715 Fri, 10 Sep 2010 19:48:52 +0000 Jean-Claude 2715@http://talk.jeelabs.net/ <p>Code has been updated - <a href="http://code.jeelabs.org/libraries/RF12/examples/RF12demo/RF12demo.pde">http://code.jeelabs.org/libraries/RF12/examples/RF12demo/RF12demo.pde</a></p> <p>Weblog post about the data-logging built into RF12demo when used with a JeeLink will go out in about two hours...</p> http://talk.jeelabs.net/topic/486#post-2714 Thu, 09 Sep 2010 12:57:53 +0000 Jean-Claude 2714@http://talk.jeelabs.net/ <p>Ah - good timing, I'm currently debugging the change needed to make RF12demo recognize the 4 MBit chips. Will post on the weblog when everything works - tomorrow or the day after.</p> http://talk.jeelabs.net/topic/486#post-2713 Thu, 09 Sep 2010 05:28:21 +0000 David 2713@http://talk.jeelabs.net/ <p>I should have been clearer - it appears to break the DF part of RF12demo.</p> http://talk.jeelabs.net/topic/486#post-2712 Thu, 09 Sep 2010 05:27:01 +0000 David 2712@http://talk.jeelabs.net/ <p>I just received a Jeelink v2 from Modern Devices. I understand that it has new/different flash memory installed - and it appears to break RF12demo.4</p> <p>How can I get this fixed?</p> http://talk.jeelabs.net/topic/456/page/4#post-2710 Tue, 07 Sep 2010 12:05:11 +0000 Jean-Claude 2710@http://talk.jeelabs.net/ <p>No - the EtherNodePull code in demo-ethernodepull.tcl will remain the same.</p> <p>But there will be a chain of filters / transformers, to process an obscure "OK 56 243 0 200 0" packet of bytes and turn it into a set of meaningful readings, regardless whether they came via an EtherNode or via a directly connected JeeNode/JeeLink. Transport independence - this isn't doable with the current logic in JeeMon.</p> <p>Basic idea:</p> <ul> <li>"OK 56 ..." - oh, that's on 868 MHz net group 5 ... let's look it up: aha, a room node in the kitchen</li> <li>"ROOM-RF12 56 ..." - ah, we need to run the "ROOM-RF12" packet decoder and generate a "ROOM" message</li> <li>"ROOM ..." - now run the "ROOM" handler, and interpret those 4 values as kitchen light/motion/temp/humid</li> </ul> <p>Incoming packets generate events, which get tagged with info about the device that picked them up (ethernode ip address, JeeLink band/group, etc). The "OK" event then triggers an "OK" handler, which generates another event (or more than one, or none). These events in turn trigger new handlers, etc - until the process stops, producing a set of "readings".</p> <p>This isn't JeeNode-specific, I'm just using that as example. Anything that can be connected and has a little driver for it to turn incoming data into events will be able to use the same mechanism. There's a lot more going on behind the scenes, email me if you'd like to look at some early docs and code.</p> http://talk.jeelabs.net/topic/456/page/4#post-2709 Tue, 07 Sep 2010 09:18:59 +0000 JoopH 2709@http://talk.jeelabs.net/ <p>"Redoing the way messages get collected and routed" inclusive collecting over ethernet??</p> http://talk.jeelabs.net/topic/456/page/4#post-2705 Sun, 05 Sep 2010 20:18:15 +0000 Jean-Claude 2705@http://talk.jeelabs.net/ <p>I'm still using the 2009 version of JeeMon for collecting my own meter data, so can't help you with this right now.</p> <p>Am in the process of completely redoing the way messages get collected and routed. The puzzle I'm trying to solve alongside that, is how to auto-detect new devices and make it possible to add / replace drivers for different devices without restarting the system. It's most of all a modeling/design task, i.e. handling direct connections, devices that collect packets from other devices, as well as devices which consists of a group of separate sensors and data sources (e.g. ookRelay). Once the basic mechanism works, I'll start bringing over all the currently working bits &amp; pieces.</p> <p>It's all taking longer than planned, and blocking much more other work than I had anticipated. The sort-of-good news is that I've been making some substantial progress these last weeks.</p>