blog/archive/2016/08/08.rss

   1 <?xml version="1.0" encoding="ISO-8859-1"?>
   2 <rss version='2.0' xmlns:lj='http://www.livejournal.org/rss/lj/1.0/'>
   3         <channel>
   4                 <title>Petter Reinholdtsen - Entries from August 2016</title>
   5                 <description>Entries from August 2016</description>
   6                 <link>http://people.skolelinux.org/pere/blog/</link>
   7
   8
   9         <item>
  10                 <title>Coz can help you find bottlenecks in multi-threaded software - nice free software</title>
  11                 <link>http://people.skolelinux.org/pere/blog/Coz_can_help_you_find_bottlenecks_in_multi_threaded_software___nice_free_software.html</link>
  12                 <guid isPermaLink="true">http://people.skolelinux.org/pere/blog/Coz_can_help_you_find_bottlenecks_in_multi_threaded_software___nice_free_software.html</guid>
  13                 <pubDate>Thu, 11 Aug 2016 12:00:00 +0200</pubDate>
  14                 <description>&lt;p&gt;This summer, I read a great article
  15 &quot;&lt;a href=&quot;https://www.usenix.org/publications/login/summer2016/curtsinger&quot;&gt;coz:
  16 This Is the Profiler You&#39;re Looking For&lt;/a&gt;&quot; in USENIX ;login: about
  17 how to profile multi-threaded programs.  It presented a system for
  18 profiling software by running experiences in the running program,
  19 testing how run time performance is affected by &quot;speeding up&quot; parts of
  20 the code to various degrees compared to a normal run.  It does this by
  21 slowing down parallel threads while the &quot;faster up&quot; code is running
  22 and measure how this affect processing time.  The processing time is
  23 measured using probes inserted into the code, either using progress
  24 counters (COZ_PROGRESS) or as latency meters (COZ_BEGIN/COZ_END).  It
  25 can also measure unmodified code by measuring complete the program
  26 runtime and running the program several times instead.&lt;/p&gt;
  27
  28 &lt;p&gt;The project and presentation was so inspiring that I would like to
  29 get the system into Debian.  I
  30 &lt;a href=&quot;https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=830708&quot;&gt;created
  31 a WNPP request for it&lt;/a&gt; and contacted upstream to try to make the
  32 system ready for Debian by sending patches.  The build process need to
  33 be changed a bit to avoid running &#39;git clone&#39; to get dependencies, and
  34 to include the JavaScript web page used to visualize the collected
  35 profiling information included in the source package.
  36 But I expect that should work out fairly soon.&lt;/p&gt;
  37
  38 &lt;p&gt;The way the system work is fairly simple.  To run an coz experiment
  39 on a binary with debug symbols available, start the program like this:
  40
  41 &lt;p&gt;&lt;blockquote&gt;&lt;pre&gt;
  42 coz run --- program-to-run
  43 &lt;/pre&gt;&lt;/blockquote&gt;&lt;/p&gt;
  44
  45 &lt;p&gt;This will create a text file profile.coz with the instrumentation
  46 information.  To show what part of the code affect the performance
  47 most, use a web browser and either point it to
  48 &lt;a href=&quot;http://plasma-umass.github.io/coz/&quot;&gt;http://plasma-umass.github.io/coz/&lt;/a&gt;
  49 or use the copy from git (in the gh-pages branch).  Check out this web
  50 site to have a look at several example profiling runs and get an idea what the end result from the profile runs look like.  To make the
  51 profiling more useful you include &amp;lt;coz.h&amp;gt; and insert the
  52 COZ_PROGRESS or COZ_BEGIN and COZ_END at appropriate places in the
  53 code, rebuild and run the profiler.  This allow coz to do more
  54 targeted experiments.&lt;/p&gt;
  55
  56 &lt;p&gt;A video published by ACM
  57 &lt;a href=&quot;https://www.youtube.com/watch?v=jE0V-p1odPg&quot;&gt;presenting the
  58 Coz profiler&lt;/a&gt; is available from Youtube.  There is also a paper
  59 from the 25th Symposium on Operating Systems Principles available
  60 titled
  61 &lt;a href=&quot;https://www.usenix.org/conference/atc16/technical-sessions/presentation/curtsinger&quot;&gt;Coz:
  62 finding code that counts with causal profiling&lt;/a&gt;.&lt;/p&gt;
  63
  64 &lt;p&gt;&lt;a href=&quot;https://github.com/plasma-umass/coz&quot;&gt;The source code&lt;/a&gt;
  65 for Coz is available from github.  It will only build with clang
  66 because it uses a
  67 &lt;a href=&quot;https://gcc.gnu.org/bugzilla/show_bug.cgi?id=55606&quot;&gt;C++
  68 feature missing in GCC&lt;/a&gt;, but I&#39;ve submitted
  69 &lt;a href=&quot;https://github.com/plasma-umass/coz/pull/67&quot;&gt;a patch to solve
  70 it&lt;/a&gt; and hope it will be included in the upstream source soon.&lt;/p&gt;
  71
  72 &lt;p&gt;Please get in touch if you, like me, would like to see this piece
  73 of software in Debian.  I would very much like some help with the
  74 packaging effort, as I lack the in depth knowledge on how to package
  75 C++ libraries.&lt;/p&gt;
  76 </description>
  77         </item>
  78
  79         <item>
  80                 <title>Sales number for the Free Culture translation, first half of 2016</title>
  81                 <link>http://people.skolelinux.org/pere/blog/Sales_number_for_the_Free_Culture_translation__first_half_of_2016.html</link>
  82                 <guid isPermaLink="true">http://people.skolelinux.org/pere/blog/Sales_number_for_the_Free_Culture_translation__first_half_of_2016.html</guid>
  83                 <pubDate>Fri, 5 Aug 2016 22:45:00 +0200</pubDate>
  84                 <description>&lt;p&gt;As my regular readers probably remember, the last year I published
  85 a French and Norwegian translation of the classic
  86 &lt;a href=&quot;http://www.free-culture.cc/&quot;&gt;Free Culture book&lt;/a&gt; by the
  87 founder of the Creative Commons movement, Lawrence Lessig.  A bit less
  88 known is the fact that due to the way I created the translations,
  89 using docbook and po4a, I also recreated the English original.  And
  90 because I already had created a new the PDF edition, I published it
  91 too.  The revenue from the books are sent to the Creative Commons
  92 Corporation.  In other words, I do not earn any money from this
  93 project, I just earn the warm fuzzy feeling that the text is available
  94 for a wider audience and more people can learn why the Creative
  95 Commons is needed.&lt;/p&gt;
  96
  97 &lt;p&gt;Today, just for fun, I had a look at the sales number over at
  98 Lulu.com, which take care of payment, printing and shipping.  Much to
  99 my surprise, the English edition is selling better than both the
 100 French and Norwegian edition, despite the fact that it has been
 101 available in English since it was first published.  In total, 24 paper
 102 books was sold for USD $19.99 between 2016-01-01 and 2016-07-31:&lt;/p&gt;
 103
 104 &lt;table border=&quot;0&quot;&gt;
 105 &lt;tr&gt;&lt;th&gt;Title / language&lt;/th&gt;&lt;th&gt;Quantity&lt;/th&gt;&lt;/tr&gt;
 106 &lt;tr&gt;&lt;td&gt;&lt;a href=&quot;http://www.lulu.com/shop/lawrence-lessig/culture-libre/paperback/product-22645082.html&quot;&gt;Culture Libre / French&lt;/a&gt;&lt;/td&gt;&lt;td align=&quot;right&quot;&gt;3&lt;/td&gt;&lt;/tr&gt;
 107 &lt;tr&gt;&lt;td&gt;&lt;a href=&quot;http://www.lulu.com/shop/lawrence-lessig/fri-kultur/paperback/product-22441576.html&quot;&gt;Fri kultur / Norwegian&lt;/a&gt;&lt;/td&gt;&lt;td align=&quot;right&quot;&gt;7&lt;/td&gt;&lt;/tr&gt;
 108 &lt;tr&gt;&lt;td&gt;&lt;a href=&quot;http://www.lulu.com/shop/lawrence-lessig/free-culture/paperback/product-22440520.html&quot;&gt;Free Culture / English&lt;/a&gt;&lt;/td&gt;&lt;td align=&quot;right&quot;&gt;14&lt;/td&gt;&lt;/tr&gt;
 109 &lt;/table&gt;
 110
 111 &lt;p&gt;The books are available both from Lulu.com and from large book
 112 stores like Amazon and Barnes&amp;Noble.  Most revenue, around $10 per
 113 book, is sent to the Creative Commons project when the book is sold
 114 directly by Lulu.com.  The other channels give less revenue.  The
 115 summary from Lulu tell me 10 books was sold via the Amazon channel, 10
 116 via Ingram (what is this?) and 4 directly by Lulu.  And Lulu.com tells
 117 me that the revenue sent so far this year is USD $101.42.  No idea
 118 what kind of sales numbers to expect, so I do not know if that is a
 119 good amount of sales for a 10 year old book or not.  But it make me
 120 happy that the buyers find the book, and I hope they enjoy reading it
 121 as much as I did.&lt;/p&gt;
 122
 123 &lt;p&gt;The ebook edition is available for free from
 124 &lt;a href=&quot;https://github.com/petterreinholdtsen/free-culture-lessig&quot;&gt;Github&lt;/a&gt;.&lt;/p&gt;
 125
 126 &lt;p&gt;If you would like to translate and publish the book in your native
 127 language, I would be happy to help make it happen.  Please get in
 128 touch.&lt;/p&gt;
 129 </description>
 130         </item>
 131
 132         <item>
 133                 <title>Vitenskapen tar som vanlig feil igjen - relativt feil</title>
 134                 <link>http://people.skolelinux.org/pere/blog/Vitenskapen_tar_som_vanlig_feil_igjen___relativt_feil.html</link>
 135                 <guid isPermaLink="true">http://people.skolelinux.org/pere/blog/Vitenskapen_tar_som_vanlig_feil_igjen___relativt_feil.html</guid>
 136                 <pubDate>Mon, 1 Aug 2016 16:00:00 +0200</pubDate>
 137                 <description>&lt;p&gt;For mange år siden leste jeg en klassisk tekst som gjorde såpass
 138 inntrykk på meg at jeg husker den fortsatt, flere år senere, og bruker
 139 argumentene fra den stadig vekk.  Teksten var «The Relativity of
 140 Wrong» som Isaac Asimov publiserte i Skeptical Inquirer i 1989.  Den
 141 gir litt perspektiv rundt formidlingen av vitenskapelige resultater.
 142 Jeg har hatt lyst til å kunne dele den også med folk som ikke
 143 behersker engelsk så godt, som barn og noen av mine eldre slektninger,
 144 og har savnet å ha den tilgjengelig på norsk.  For to uker siden tok
 145 jeg meg sammen og kontaktet Asbjørn Dyrendal i foreningen Skepsis om
 146 de var interessert i å publisere en norsk utgave på bloggen sin, og da
 147 han var positiv tok jeg kontakt med Skeptical Inquirer og spurte om
 148 det var greit for dem.  I løpet av noen dager fikk vi tilbakemelding
 149 fra Barry Karr hos The Skeptical Inquirer som hadde sjekket og fått OK
 150 fra Robyn Asimov som representerte arvingene i Asmiov-familien og gikk
 151 igang med oversettingen.&lt;/p&gt;
 152
 153 &lt;p&gt;Resultatet, &lt;a href=&quot;http://www.skepsis.no/?p=1617&quot;&gt;«Relativt
 154 feil»&lt;/a&gt;, ble publisert på skepsis-bloggen for noen minutter siden.
 155 Jeg anbefaler deg på det varmeste å lese denne teksten og dele den med
 156 dine venner.&lt;/p&gt;
 157
 158 &lt;p&gt;For å håndtere oversettelsen og sikre at original og oversettelse
 159 var i sync brukte vi git, po4a, GNU make og Transifex.  Det hele
 160 fungerte utmerket og gjorde det enkelt å dele tekstene og jobbe sammen
 161 om finpuss på formuleringene.  Hadde hosted.weblate.org latt meg
 162 opprette nye prosjekter selv i stedet for å måtte kontakte
 163 administratoren der, så hadde jeg brukt weblate i stedet.&lt;/p&gt;
 164 </description>
 165         </item>
 166
 167         <item>
 168                 <title>Techno TV broadcasting live across Norway and the Internet (#debconf16, #nuug) on @frikanalen</title>
 169                 <link>http://people.skolelinux.org/pere/blog/Techno_TV_broadcasting_live_across_Norway_and_the_Internet___debconf16___nuug__on__frikanalen.html</link>
 170                 <guid isPermaLink="true">http://people.skolelinux.org/pere/blog/Techno_TV_broadcasting_live_across_Norway_and_the_Internet___debconf16___nuug__on__frikanalen.html</guid>
 171                 <pubDate>Mon, 1 Aug 2016 10:30:00 +0200</pubDate>
 172                 <description>&lt;p&gt;Did you know there is a TV channel broadcasting talks from DebConf
 173 16 across an entire country?  Or that there is a TV channel
 174 broadcasting talks by or about
 175 &lt;a href=&quot;http://beta.frikanalen.no/video/625529/&quot;&gt;Linus Torvalds&lt;/a&gt;,
 176 &lt;a href=&quot;http://beta.frikanalen.no/video/625599/&quot;&gt;Tor&lt;/a&gt;,
 177 &lt;a href=&quot;http://beta.frikanalen.no/video/624019/&quot;&gt;OpenID&lt;/A&gt;,
 178 &lt;a href=&quot;http://beta.frikanalen.no/video/625624/&quot;&gt;Common Lisp&lt;/a&gt;,
 179 &lt;a href=&quot;http://beta.frikanalen.no/video/625446/&quot;&gt;Civic Tech&lt;/a&gt;,
 180 &lt;a href=&quot;http://beta.frikanalen.no/video/625090/&quot;&gt;EFF founder John Barlow&lt;/a&gt;,
 181 &lt;a href=&quot;http://beta.frikanalen.no/video/625432/&quot;&gt;how to make 3D
 182 printer electronics&lt;/a&gt; and many more fascinating topics?  It works
 183 using only free software (all of it
 184 &lt;a href=&quot;http://github.com/Frikanalen&quot;&gt;available from Github&lt;/a&gt;), and
 185 is administrated using a web browser and a web API.&lt;/p&gt;
 186
 187 &lt;p&gt;The TV channel is the Norwegian open channel
 188 &lt;a href=&quot;http://www.frikanalen.no/&quot;&gt;Frikanalen&lt;/a&gt;, and I am involved
 189 via &lt;a href=&quot;https://www.nuug.no/&quot;&gt;the NUUG member association&lt;/a&gt; in
 190 running and developing the software for the channel.  The channel is
 191 organised as a member organisation where its members can upload and
 192 broadcast what they want (think of it as Youtube for national
 193 broadcasting television).  Individuals can broadcast too.  The time
 194 slots are handled on a first come, first serve basis.  Because the
 195 channel have almost no viewers and very few active members, we can
 196 experiment with TV technology without too much flack when we make
 197 mistakes.  And thanks to the few active members, most of the slots on
 198 the schedule are free.  I see this as an opportunity to spread
 199 knowledge about technology and free software, and have a script I run
 200 regularly to fill up all the open slots the next few days with
 201 technology related video.  The end result is a channel I like to
 202 describe as Techno TV - filled with interesting talks and
 203 presentations.&lt;/p&gt;
 204
 205 &lt;p&gt;It is available on channel 50 on the Norwegian national digital TV
 206 network (RiksTV).  It is also available as a multicast stream on
 207 Uninett.  And finally, it is available as
 208 &lt;a href=&quot;http://beta.frikanalen.no/&quot;&gt;a WebM unicast stream&lt;/a&gt; from
 209 Frikanalen and NUUG.  Check it out. :)&lt;/p&gt;
 210 </description>
 211         </item>
 212
 213         </channel>
 214 </rss>