[ramblings
tom@goto10**20070411230149] {
hunk ./ramblings.txt 856
-what about:
-- hard linked pool? and sync only the pool?
+what about hard linked pool? and sync only the pool? hard links are
+better because they are not directional.
hunk ./ramblings.txt 868
+this is for another time. don't have enough context in my head for
+it.. would be a nice opportunity to give scheme shell a try though.
+
+
+
+Entry: dvd ripping
+Date: Thu Mar 22 13:17:42 EDT 2007
+
+dvdbackup -i /dev/dvd -M -o where_to_save
+
+for naked lunch i had to "mplayer -aid 128". by default it took 129
+which was a voice over narrative.
+
+probably best to have -o be the toplevel rip directory (dir containing
+ripped cds) and set the title with -n in case of multiple disks (naked
+lunch was named the same) something like
+
+dvdbackup -i /dev/dvd -M -o /data/dvdbackup -n <title>
+
+
+Entry: bidir udp
+Date: Sat Mar 24 18:40:25 EDT 2007
+
+2 ends / symmetric
+
+socat - UDP4:localhost:11111,sourceport=22222
+socat - UDP4:localhost:22222,sourceport=11111
+
+
+that seems to work ok, now using firewalls
+
+A  FW_A   FW_B  B
+
+using firewall doesnt work.. the thing to know is how ports are
+mapped. assuming this is just copied apparently doesnt work.. if it is
+known then A -> FW_B and B -> FW_A
+
+see
+http://www.gotroot.com/tiki-view_blog_post.php?blogId=2&postId=4
+
+
+
+Entry: scheme shell on openwrt
+Date: Tue Mar 27 17:42:50 EDT 2007
+
+i managed to finally compile it, using a native compiler from
+http://uclibc.org/FAQ.html#dev_systems
+
+apparently, scsh/scheme48 doesn't like cross compilation, since it
+uses the generated image to perform further compilation. the debian
+uclibc-toolchain wrapper seems to be broken too. i tried this on
+mipsel debian (and i386).
+
+the only problem i have left is that the versions don't
+match. currently i just use another root fs, but that's not very
+elegant.
+
+
+Entry: mencoder
+Date: Mon Apr  2 12:04:46 EDT 2007
+
+cd /data/mencoder
+ln -s .../*.VOB .
+
+then delete all unwanted VOBS
+
+mencoder -vf crop=:460 -ovc frameno -oac mp3lame -lameopts vbr=3 -o frameno.avi -mpegopts format=dvd VTS_01*VOB 2>&1
+
+->
+
+ODML: Aspect information not (yet?) available or unspecified, not writing vprp header.
+Recommended video bitrate for 650MB CD: 903
+Recommended video bitrate for 700MB CD: 981
+Recommended video bitrate for 800MB CD: 1138
+Recommended video bitrate for 2 x 650MB CD: 1923
+Recommended video bitrate for 2 x 700MB CD: 2080
+Recommended video bitrate for 2 x 800MB CD: 2394
+
+Video stream:    0.767 kbit/s  (95 bps)  size: 512744 bytes  5344.845 secs  128188 frames
+
+Audio stream:  102.225 kbit/s  (12778 bps)  size: 68303400 bytes  5345.352 secs
+
+
+ mencoder -vf crop=:460 -oac copy -ovc lavc -lavcopts vcodec=mpeg4:vpass=1:vbitrate=22000 -o movie.avi *.VOB
+
+ODML: Aspect information not (yet?) available or unspecified, not writing vprp header.
+
+Video stream:  118.936 kbit/s  (14867 bps)  size: 79458950 bytes  5344.645 secs  128188 frames
+
+Audio stream:  102.228 kbit/s  (12778 bps)  size: 68302344 bytes  5345.088 secs
+
+
+oops. changed bitrate to 981
+
+// last try, something wrong with aspect, so setting manually
+mencoder -vf crop=:460 -oac copy -ovc lavc -lavcopts vcodec=mpeg4:vpass=1:aspect=16/9:vbitrate=980 -o movie.avi *.VOB
+
+
+-- again
+
+// crop if necessary
+mencoder -ovc frameno -oac mp3lame -lameopts vbr=3 -o frameno.avi -mpegopts format=dvd *.VOB
+
+
+
+Entry: scsh -- the pain of unix
+Date: Thu Apr  5 13:08:52 EDT 2007
+
+the pain of unix is really impedance matching. it's nice to have all
+these small utilities output data tables in their own adhoc way, but
+it sucks to parse them == match them to other things.
+
+i guess, the 'worse is better' motto applies here. it's good to not
+have standards since it would slow dev, but it's good to have
+standards to allow interopration. looks like we just need to fill the
+gap with something that does the trick, and perl is really not the
+answer..
+
+scsh to the rescue, i hope. i'd like to get rid of sh,grep,awk or
+perl, and do it properly: parse output into an s-expression, then work
+with that.
+
+problem is i can't get scsh to run on the routers. cross compilation
+doesn't seem to work, and it seems to be a bit too big too.
+
+so what about tethering? all i really need is ssh and some basic
+'machine language' which could be the shell.
+
+
+
+Entry: fred and sam
+Date: Fri Apr  6 10:29:30 EDT 2007
+
+fix openvpn when network is switched.
+wow that thing is borked!
+
+it looks like there are 4 copies of the wds-ifup script spawned, so i
+had to add some sync code (rm /tmp/ifup-wds || exit)
+
+seems to work now. firewall is decent too.
+
+
+Entry: plt scheme XML
+Date: Tue Apr 10 17:59:51 EDT 2007
+
+see /usr/lib/plt/collects/xml/private/xexpr.ss
+
+(require (lib "xml.ss" "xml"))
+(xml->xexpr (document-element (read-xml (open-input-file "test.xml"))))
+
+
+Entry: how to simplify xml
+Date: Wed Apr 11 14:53:20 EDT 2007
+
+using the xexpr representation mentioned above, with spaces
+stripped. still this is too complicated a structure. the idea is to
+make it just a nested s-expression, with only name -> value pairs, and
+use attributes for things that are not text or nested expressions.
+
+(tag
+	((a1 v1) (a2 v2))
+	"text1"
+	(tag1 ((a11 v11) (a12 v12)) (tag11 ..) (tag12))
+	"text2"
+	(tag2 <attrib> <elements>))
+
+what i want is just this:
+
+(tag
+	(a1 . v1)
+	(a2 . v2)
+	(string . "text1")
+	(tag1 ((a11 . v11) (a12 . v12) (tag11 . (...))))
+	(string . "text2")
+	(tag2 . (...))
+	...)
+	
+this means the fist simplification is to not allow text floating
+inbetween nested things: it needs to be wrapped in a 'string' tag.
+
}