<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0"><channel><title>Disqus - Latest Comments for deepspawn</title><link xmlns="http://www.w3.org/2005/Atom" rel="http://api.friendfeed.com/2008/03#sup" href="http://disqus.com/sup/all.sup#usercomments-42b4cca8" type="application/json"/><link>http://disqus.com/people/deepspawn/</link><description></description><language>en</language><lastBuildDate>Wed, 10 Jun 2009 19:47:21 -0000</lastBuildDate><item><title>Re: Uzbl Browser</title><link>http://blog.valkertown.org/2009/06/09/uzbl_browser/#comment-10723072</link><description>Good, I'm going to corred the PID on the post.&lt;br&gt;&lt;br&gt;The patch needs work since the button it's hardcoded and I think it would be against the uzbl spirit of doing this stuff.&lt;br&gt;&lt;br&gt;I agree that the scripts render awfully on the blog, I'm going to move them there and link from here.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Wed, 10 Jun 2009 19:47:21 -0000</pubDate></item><item><title>Re: adventures in systems administration</title><link>http://www.tychoish.com/2009/05/adventures-in-systems-administration/#comment-9967938</link><description>Ejabberd is quite heavy on the memory usage part, I don't know how much ram you got but basic linode tends to choke on ejabberd's default config. I guess it can go lower but it will require some sort of tweaking. Transports and stuff like that doesn't eat too much ram. &lt;br&gt;&lt;br&gt;Mail is such a pain to setup that I ended using google hosted mail, since it requires very few steps to get it working. But, well you handle them all your mail. &lt;br&gt;&lt;br&gt;I'm willing to get out of there for some of the domains I handle but for others the users are quite happy with their google hosted accounts. I need to find a more libre solution that copes with my lazyness of dealing with a smtp+smap filter+ssl+pop/imap+webmail setup.&lt;br&gt;&lt;br&gt;On the system administration part of your post. I have a mixed feeling about it, some times I like it and most of the time it feels rather "uncreative" job.  I'm not well suited for that kind of task.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Tue, 26 May 2009 16:54:34 -0000</pubDate></item><item><title>Re: Setting up Redmine with archlinux</title><link>http://blog.valkertown.org/2008/12/21/setting_up_redmine_with_archlinux/#comment-8070719</link><description>Oh I'm sorry been way to busy doing some other random stuff but since you want it, I'm going to finish it, thank you for bringing this back to my mind#</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Fri, 10 Apr 2009 21:53:49 -0000</pubDate></item><item><title>Re: Tape Your Stuff to the Web</title><link>http://metajack.im/2009/04/07/tape-your-stuff-to-the-web/#comment-7961864</link><description>Good, I was needing this badly, even with tramp on emacs it's not that easy. I even setup apache but that approach is a lot more troublesome than this</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Tue, 07 Apr 2009 22:10:36 -0000</pubDate></item><item><title>Re: Writting Circular Lists in Erlang</title><link>http://blog.valkertown.org/2009/03/18/writting_circular_lists_in_erlang/#comment-7573404</link><description>Thanks, I'll give it a try and do a benchmark to see if It's better to&lt;br&gt;the other solution in the above comments.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Fri, 27 Mar 2009 17:45:21 -0000</pubDate></item><item><title>Re: Writting Circular Lists in Erlang</title><link>http://blog.valkertown.org/2009/03/18/writting_circular_lists_in_erlang/#comment-7363124</link><description>Nice, I'm just repeating what I have read so far but&lt;br&gt;&lt;br&gt; [Push] ++ drop_last(List)&lt;br&gt;&lt;br&gt;Wouldn't it be better written as:&lt;br&gt; &lt;br&gt; [ Push | drop_last(List)]&lt;br&gt;&lt;br&gt;Anyway, ++ is discouraged since it makes a copy of the left operator and with long lists it would be highly inefficient, in this case it wouldn't be much of a problem but I still like more how the second one looks.&lt;br&gt;&lt;br&gt;Thank you a lot for this idea, I'm guessing it's probably the best way to write it on erlang.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Thu, 19 Mar 2009 21:24:26 -0000</pubDate></item><item><title>Re: Emacs Denting mode for laconicas (incliding identi.ca)</title><link>http://blog.valkertown.org/2009/02/16/emacs_denting_mode_for_laconicas_28incliding_identica29/#comment-6615310</link><description>Hello, I had the very same problem but Christian Cheng sent me a patch which solves the issue. I did my work without proper search first so I duplicated a lot of your efforts and I did a merge with your code few days ago so denting-mode has most of the features you did implement. I don't know if you updated the code recently but I will check again the diffs to merge your work in.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Wed, 25 Feb 2009 12:12:53 -0000</pubDate></item><item><title>Re: Pimp my Shell Parte 2</title><link>http://blog.valkertown.org/2005/12/21/pimp-my-shell-parte-2/#comment-6600758</link><description>delete3</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Wed, 25 Feb 2009 08:57:14 -0000</pubDate></item><item><title>Re: Emacs Denting mode for laconicas (incliding identi.ca)</title><link>http://blog.valkertown.org/2009/02/16/emacs_denting_mode_for_laconicas_28incliding_identica29/#comment-6392483</link><description>Sorry for taking so long, well It's not so hard&lt;br&gt;&lt;br&gt;line 661 of current denting-mode.el&lt;br&gt;&lt;br&gt;648 (defun denting-http-post-default-sentinel (proc stat &amp;optional suc-msg)&lt;br&gt;649&lt;br&gt;650  (condition-case err-signal&lt;br&gt;651      (let ((header (denting-get-response-header))&lt;br&gt;652	    ;; (body (denting-get-response-body)) not used now.&lt;br&gt;653	    (status nil))&lt;br&gt;654	(string-match "HTTP/1\.1 \\([a-z0-9 ]+\\)\r?\n" header)&lt;br&gt;655	(setq status (match-string-no-properties 1 header))&lt;br&gt;656	(case-string status&lt;br&gt;657		     (("200 OK")&lt;br&gt;658		      (message (if suc-msg suc-msg "Success: Post")))&lt;br&gt;659		     (t (message status)))&lt;br&gt;660	)&lt;br&gt;661    (error (message (prin1-to-string err-signal))))&lt;br&gt;662  )&lt;br&gt;&lt;br&gt;You can replace it with:&lt;br&gt;&lt;br&gt; (error (message "Failed")))&lt;br&gt;&lt;br&gt;You will replace the annoying message for a brief "Failed" message, it will not solve the issue but at least will not bug you with the full error. I will keep the full error printing for now, since maybe I can do a proper fix of the problem.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Wed, 18 Feb 2009 23:16:56 -0000</pubDate></item><item><title>Re: Emacs Denting mode for laconicas (incliding identi.ca)</title><link>http://blog.valkertown.org/2009/02/16/emacs_denting_mode_for_laconicas_28incliding_identica29/#comment-6376710</link><description>Yes I have the same problem, it's a bit annoying but totally harmless, it's just the xml parser failing from time to time to parse the laconica response. For a quick fix I only can think about shutting the hell out of it by capturing the error. On the proper way to solve the problem, it would require to figure out why the twitter/identica serves bad xml and correct it. My bets are on the api limit(so maybe just setting up the update interval to a larger number might fix it)&lt;br&gt;&lt;br&gt;Also you can avoid the wrapper function by using the identi.cas twitter bridge on your account settings. If you really want that function I think it would be something like this:&lt;br&gt;&lt;br&gt;(defun multi-update-status-from-minibuffer (&amp;optional init-str)&lt;br&gt;  (if (null init-str) (setq init-str ""))&lt;br&gt;  (let ((status init-str))&lt;br&gt;  (setq status (read-from-minibuffer "status: " status nil nil nil nil t))&lt;br&gt;  (denting-update-status-if-not-blank status)&lt;br&gt;  (twittering-update-status-if-not-blank status)))</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Wed, 18 Feb 2009 11:58:55 -0000</pubDate></item><item><title>Re: Arreglando un robot de Planeta Linux para identi.ca</title><link>http://blog.valkertown.org/2008/12/08/arreglando_un_robot_de_planeta_linux_para_identica/#comment-4278264</link><description>Ehhh pues me hubiese gustado saberlo(invesitgado, realmente) antes, voy a agregar esto pronto, para que el programa me quede más organizado.&lt;br&gt;&lt;br&gt;Gracias!, aff no quedo como respuesta del otro comentario, tengo que aprender a usar este disqus</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Mon, 08 Dec 2008 19:59:13 -0000</pubDate></item><item><title>Re: Awesome-WM settings and workplace setup</title><link>http://blog.valkertown.org/2008/12/06/awesome-wm_settings_and_workplace_setup/#comment-4225001</link><description>Check again, I have it shared just below the screenshot ;)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Sat, 06 Dec 2008 12:32:09 -0000</pubDate></item><item><title>Re: Disqus comments now on Valkertown</title><link>http://blog.valkertown.org/2008/12/03/disqus_comments_now_on_valkertown/#comment-4164830</link><description>Well all my old comments that were imported wrongly to disqus will have to stay like that.&lt;br&gt;&lt;br&gt;If disqus could do a real deleteion of comments I could run import again and that should fix it. But even when they are deleted, disqus keeps them and when I try to import them again it doesn't update the posts or create new ones.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Thu, 04 Dec 2008 08:45:56 -0000</pubDate></item><item><title>Re: Disqus comments now on Valkertown</title><link>http://blog.valkertown.org/2008/12/03/disqus_comments_now_on_valkertown/#comment-4162732</link><description>Managed to fix it, now I have to see how disqus handles it since it first imported the wrongly encoded plugins and now it has to import the properly encoed. Will it create duplicates or will it update them properly.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Thu, 04 Dec 2008 02:51:36 -0000</pubDate></item><item><title>Re: Disqus comments now on Valkertown</title><link>http://blog.valkertown.org/2008/12/03/disqus_comments_now_on_valkertown/#comment-4162208</link><description>Testing now disqus, and regreting I have just wasted all my encoding, maybe if I process again the database I can get the stuff right.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Thu, 04 Dec 2008 01:45:18 -0000</pubDate></item><item><title>Re: Setting my desktop backgrounds from Flickr</title><link>http://blog.valkertown.org/2008/12/02/setting_my_desktop_backgrounds_from_flickr/#comment-4162164</link><description>Sure let me google a bit about it.&lt;br&gt;&lt;br&gt;It seems that for ubuntu the simplest way would be something like this:&lt;br&gt;&lt;br&gt;&lt;br&gt;gconftool-2 --type string --set /desktop/gnome/background/picture_filename "/path/to/image.ext"&lt;br&gt;&lt;br&gt;So in the script it would be replace line 13 with&lt;br&gt;&lt;br&gt;gconftool-2 --type string --set /desktop/gnome/background/picture_filename "$BG1"&lt;br&gt;&lt;br&gt;and line 23 with&lt;br&gt;&lt;br&gt;gconftool-2 --type string --set /desktop/gnome/background/picture_filename /tmp/cur_bg.bmp&lt;br&gt;&lt;br&gt;Try it and tell me if it works</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Tue, 02 Dec 2008 14:44:44 -0000</pubDate></item><item><title>Re: Setting my desktop backgrounds from Flickr</title><link>http://blog.valkertown.org/2008/12/02/setting_my_desktop_backgrounds_from_flickr/#comment-4162162</link><description>Well you can install Eterm it most likely will install also Esetroot.&lt;br&gt;&lt;br&gt;I use it since it works with the transparency from rxvt-unicode but you could replace it with some other command that lets you set the background from the command line.&lt;br&gt;&lt;br&gt;There was a little mistake that now is corrected on the post, on line 9 of the python script it should be "tagmode" and not "tagsmode", otherwise you will end with a very narrow set of images.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Tue, 02 Dec 2008 06:37:58 -0000</pubDate></item><item><title>Re: Goodbye Pownce, shutting down Jance</title><link>http://blog.valkertown.org/2008/12/01/goodbye_pownce2c_shutting_down_jance/#comment-4162159</link><description>Hello Evan, it was an xmpp bot to interact with Pownce.&lt;br&gt;&lt;br&gt; I implemented most of the pownce API  and added some very interesting  regexpes track/blacklist capabilites. &lt;br&gt;&lt;br&gt;I don't know if it would be of much use for laconica since most of the problems it solves are now pretty much solved now for laconica, Yet there may be one or two ideas that might be interesting to see on xmpp bots from identi.ca</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Tue, 02 Dec 2008 00:31:45 -0000</pubDate></item><item><title>Re: DS1M12 usb osciloscope in linux and python</title><link>http://blog.valkertown.org/2008/10/01/ds1m12_usb_osciloscope_in_linux_and_python/#comment-4162156</link><description>They are already published, check the &lt;a href="http://hg.valkertown.org" rel="nofollow"&gt;hg.valkertown.org&lt;/a&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Thu, 27 Nov 2008 19:06:43 -0000</pubDate></item><item><title>Re: DS1M12 usb osciloscope in linux and python</title><link>http://blog.valkertown.org/2008/10/01/ds1m12_usb_osciloscope_in_linux_and_python/#comment-4162154</link><description>I'm going to publish them now, I was waiting to see if there was any problem from my current employer if I do that, but seems that there's no problem.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Wed, 26 Nov 2008 07:54:38 -0000</pubDate></item><item><title>Re: I Have a Dream. In Common Lisp and Twitter.</title><link>http://zvoase.tumblr.com/post/59243688#comment-3695886</link><description>I have to agree with you with erlang syntax and that I also prefer lisp to some extend and that's why I like to program in Python so much.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Tue, 11 Nov 2008 21:18:06 -0000</pubDate></item><item><title>Re: Concurso mejores trabajos de grado</title><link>http://blog.valkertown.org/2008/10/31/concurso_mejores_trabajos_de_grado/#comment-4162157</link><description>La mediocridad de la Universidad Nacional cada vez esta peor, y el orgullo y soberbia nos ha caracterizado por demasiado tiempo. Desafortunadamente la relevacia de la Universidad y su participaciÃ³n con el sector privado y productivo se  encuentra en serio detrimento.&lt;br&gt;&lt;br&gt;Es gracioso como se ubica en este correo un trabajo de grado contra una pasantia como algo mucho mÃ¡s productivo para la "humanidad".  Realmente pienso que es hora de dejar la soberbia y empezar a trabajar con los demÃ¡s, la universidad no puede ser un ente tan ajeno al sector privado, colaborar con las iniciativas de otras universidades.&lt;br&gt;&lt;br&gt;Las tesis que he visto en los Ãºltimos aÃ±os han sido muy probres, con muy poco trabajo y casi sin continuidad, se tratan mÃ¡s de juguetes que de serias contribuciones a la humanidad. Y que decir de los trabajos de grado en maestrias?&lt;br&gt;&lt;br&gt;La observaciÃ³n que realizamos algunos de los que trabajamos en el Centro, donde durante muchos aÃ±os y con mucha lucha hemos tratado de vincular el desarrollo y la investigaciÃ±Ã³n con la industria Colombiana. &lt;br&gt;&lt;br&gt;Dejo asÃ­ por ahora, pero es hora de volver a la realidad y seguir trabajando.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Fri, 31 Oct 2008 10:46:38 -0000</pubDate></item><item><title>Re: Simulador de Robot con chip ADN simulado</title><link>http://blog.valkertown.org/2006/12/09/simulador-de-robot-con-chip-adn-simulado/#comment-4162125</link><description>Hola Alejandro, pues el simulador fue creado solo con estos dos lenguajes VHDL para realizar una simulaciÃ³n conscisa del chip ADN, eventualmente ese chip se realizo en una FPGA junto con el robot.&lt;br&gt;&lt;br&gt;Las entradas/salidas del chip ADN al tratarse de seÃ±ales digitales realizar la interacciÃ³n con python fue realmente simple pero requiere de una estrategia asincronica para permitir al simulador de VHDL evolucionar en el tiempo, o generar las seÃ±al del reloj principal del chip desde python.&lt;br&gt;&lt;br&gt;En python se realizo el resto, basicamente son 3 componentes, visualizaciÃ³n, simuladores y control.&lt;br&gt;&lt;br&gt;El proceso de visualizaciÃ³n trabaja completamente independiente de la simulaciÃ³n y su Ãºnica labor es mostrar el estado en el que encuentre el sistema en el momento que decider actualizar la ventana, aquÃ­ escogi realizar una mimica de proceso y pues al tratarse del trabajo final de un curso no implemente mÃ¡s visualizaciones, es mÃ¡s interesante ver los resultados de los componentes internos, seÃ±ales de los sensores, motores, modelo mecÃ¡nico, etc. Este componente se construyo utilizando pygame&lt;br&gt;&lt;br&gt;El componente de control es quien interactua con el usuario y cambia las condiciones de la simulaciÃ³n y establece en este caso las condiciones iniciales, tales como posiciÃ³n y tambien permite escoger diferentes pistas que el robot debe reconocer.&lt;br&gt;&lt;br&gt;La parte mÃ¡s sofisticada del asunto resulto ser el esquema de simulaciÃ³n pues de alguna forma cada cada mÃ³dulo tiene que tener su base de tiempo independiente y el sistema no resulta estable en consecuencia de la simulaciÃ³n nÃºmerica de los sistemas si los tiempos de resoluciÃ³n de los componentes. En este punto en particular aÃºn hoy en dÃ­a trabajo en simulaciones de arquitecturas  de hardware mÃ¡s sofisticadas y un compaÃ±ero del centro se inspiro en este trabajo para realizar en su trabajo de grado un estudio de diferentes alternativas para comunicar estos procesos de simulacion en un esquema de n procesadores. Este trabajo se realizo en python y pues realmente inicio con un esquema round robin que preguntaba por el estado de cada modulo de simulacion y de su resultado actualizaba el estado pertienente a los otros modulos de la simulaciÃ³n. Es el esquema mÃ¡s sencillo que se me ocurrio pero luego lo complique rapidamente para hacerlo mÃ¡s dinamico y concentrarme en otros problemas.&lt;br&gt;&lt;br&gt;Con el esquema de simulacion definido, se podian interconectar y mejorar la simulaciÃ³n rapidamente, una parte era el chip ADN, pero otras partes eran la simulaciÃ³n de los motores actuadores, los sensores de proximidad con defectos dinÃ¡micos y estÃ¡ticos(esto es tan sencillo como implementar una funcion de transferencia como un modulo de simulacion). El esquema de movimiento era dos motores y una rueda libre, a partir de las ecuaciones diferenciales implemente un nuevo modulo de simulaciÃ³n y asÃ­ hasta que tuve un esquema mÃ¡s complejo. &lt;br&gt;&lt;br&gt;La verdad todo suena mÃ¡s complicado de lo que realmente es, especialmente en python, pero mÃ¡s o menos esa fue la arquitectura que me invente para este simulador, que fue impulsado principalmente por la falta de tiempo para construir un robot real como lo exigia el curso y pues utilizar un simulador libre como los que ya existen no era permitido por el maestro de la asignatura.&lt;br&gt;&lt;br&gt;Como anecdota, este trabajo recibio una nota de 3.0/5.0 apenas aprovatoria en la Universidad, sin embargo en general es uno de los trabajos que mÃ¡s orgulloso me siento y ha tenido bastante impacto tanto en mi trabajo como en algunas personas con las que he podio compartirlo.&lt;br&gt;&lt;br&gt;Se me fue algo larga la respuesta Alejandro, quizÃ¡ construir el simulador desde el principio no es la mejor opciÃ³n para algunos ya que requiere de mucho esfuerzo y para lograr los mejores resultados un sistema a simular muy especifico, existen en la red otros simuladores libres y de proposito mÃ¡s general que pueden acortar mucho el trabajo de diseÃ±o y evaluaciÃ³n. Esto lo digo por que implementar esquemas de simulacion de algunos sistemas fÃ­sicos con todos los defectos y no linealidades requiere esfuerzo y tiene propositos diferentes a utlizar la herramienta de simulaciÃ³n en diseÃ±o o evaluaciÃ³n.&lt;br&gt;&lt;br&gt;&lt;br&gt;Espero haberle dado una buena respuesta y en cualquier caso sigo atento a cualuiqer otra pregunta.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Wed, 22 Oct 2008 12:37:17 -0000</pubDate></item><item><title>Re: USB y Tiempo Real</title><link>http://blog.valkertown.org/2007/05/17/usb_y_tiempo_real/#comment-4162139</link><description>Acabo de enviar el cÃ³digo, y algunas notas.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Mon, 29 Sep 2008 14:27:01 -0000</pubDate></item><item><title>Re: USB y Tiempo Real</title><link>http://blog.valkertown.org/2007/05/17/usb_y_tiempo_real/#comment-4162136</link><description>Una Ãºltima cosa, el cÃ³digo que tengo esta escrito en espacio del kernel como un modulo que crea un dispositivo de espacio de usuario, podria facilitarle parte del cÃ³digo. PodrÃ­a resultar algo intimidante si no se esta acostumbrado al cÃ³digo del kernel de Linux, de otra forma es solo un modÃºlo comÃºn y corriente.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">deepspawn</dc:creator><pubDate>Thu, 11 Sep 2008 17:29:05 -0000</pubDate></item></channel></rss>