Hacks

xrayspx's picture

Haiku OS

Music: 

Dr. Octagon - Blue Flowers

I've been playing recently with different unix OS's, trying to resurrect some old hardware and see if I can't make some of these old machines useful again. Like at least have one as an XScreensaver-only machine to keep Natalie amused.

xrayspx's picture

Lattice of Convenience - MP3 Playlists

Music: 

Underworld - Kittens

Hopefully everyone can live in the future someday.

We do a lot with MP3 playlists. I run Airsonic for streaming around the house and in the car, and we have a playlist-based FM transmitter setup, etc. So I have scripts which run every night and generate playlists based on star ratings and other things (GET THE LED OUT ANYONE?).

Previously what I've done is dump the contents of a bunch of Smart Playlists in Clementine to a file and use those files to generate the randomized 200 track daily playlists. The downside to that is that every time I add music or change star ratings, I'd have to refresh these "base" files like some kind of animal. I had base playlists for "3+ stars", "4+ Stars" and "5 Stars", among others.

Today I decided to fix all that. Clementine uses a SQLite3 database, so now I'm just querying it instead, and it seems to be working well. For example, my "5 star" playlist in Clementine results in 10800 or so tracks. The same one built from the DB ends up with a couple hundred more tracks, but is pretty close. I'm not entirely sure what the difference there is just yet, but "close enough". What it looks like to me is I probably need to enable Samba case sensitivity.

The DB records ratings as decimal numbers from 0.0 (Zero stars) through 1.0 (5 Stars). So to build a "4-Star +" playlist, searching for rating >= "0.8", you get ratings like this:

1
1.10000002384186
0.800000011920929
0.800000011920929
0.800000011920929
1.10000002384186
1.10000002384186
0.800000011920929
0.800000011920929
1
1
1
1

! Caveat: Prior to Clementine 1.4.0rc1-533-gf4e70face there was a bug where it was possible to give a song a higher than 5 star rating (higher than 1.0 in the DB) as you can see above, so know that if you have Clementine from the repositories, it's likely you have that bug. For instance in the UI, if you want to show all 5 star songs, use "Rating is Greater Than 4.5 Stars" rather than "Rating is Equal to 5 Stars".

Now I can just have a cron job to copy the master Clementine DB once a day to my server and drop it in next to the playlist generation scripts.

The downside to all this is speed. When using the Clementine-Generated base playlists, I could be sure all the files actually exist on disk. However while Clementine will only show you files that exist in the UI, it doesn't seem to do a very good job of cleaning the database of stale files which no longer exist. So if you move or rename files, the old DB entries stick around unless you purge it completely and start over from scratch. That means I have to test every single file as I add it to the playlist, which takes time. It takes about 5-8 seconds to generate my 200 track 5-Star M3U file.

The 5-Star.sh script is below if you'd like to play along at home:

  


#!/bin/bash

rm /Volumes/Filestore/CDs/playlists/5\ Stars.m3u

i=1

while [ $i -le 200 ]
do
 file=$(sqlite3 ./clementine.db "select filename from songs where rating > "0.9" order by random() limit 1;" | awk -F "file://" '{print $2}')

 ### Clementine data encodes special characters and accent marks and stuff so I'm using
 ### Joel Parker Henderson's urldecode.sh to undo that: https://gist.github.com/cdown/1163649
 
 data=$(urldecode.sh "$file")
 if [ -f "$data" ]
 then
  ### Have to escape leading brackets because grep treated it as a range and would allow duplicates ###
  ### Can't do that in "data" because \[ isn't in the filename so they'll fail ###

  escaped=$(echo "$data" | sed 's/\[/\\[/g')
  #echo "$escaped"

  ### Avoid duplicates
  match=$(grep -i "$escaped" /Volumes/Filestore/CDs/playlists/5\ Stars.m3u)
  if [ -z "$match" ]
  then
   echo "$data" >> /Volumes/Filestore/CDs/playlists/5\ Stars.m3u
   ((i++))
  fi
 fi
done

For the 3+ and 4+ lists, I repeat this main block, but instead each rating dumps into a text file that I randomize into an .m3u at the end. So for the 3-Star + script below, I collect 130 5-star tracks, 45 4-star, and 25 3-star, push them out to a temp file and then cat temp.m3u | sort -R > "./3 Star +.m3u". I could do all this by creating a new table in the database and stuffing tracks into that, but this was faster for me to write and it works well enough:


#!/bin/bash

rm /Volumes/Filestore/CDs/playlists/3\ Stars\ +.m3u

i=1

while [ $i -le 130 ]
do
 file=$(sqlite3 ./clementine.db "select filename from songs where rating > "0.9" order by random() limit 1;" | awk -F "file://" '{print $2}')

 ### Clementine data encodes special characters and accent marks and stuff so I'm using
 ### Joel Parker Henderson's urldecode.sh to undo that: https://gist.github.com/cdown/1163649
 
 data=$(urldecode.sh "$file")
 if [ -f "$data" ]
 then
  ### Have to escape leading brackets because grep treated it as a range and would allow duplicates ###
  ### Can't do that in "data" because \[ isn't in the filename so they'll fail ###

  escaped=$(echo "$data" | sed 's/\[/\\[/g')
  #echo "$escaped"

  ### Avoid duplicates
  match=$(grep -i "$escaped" ./3-star-tmp.m3u)
  if [ -z "$match" ]
  then
   echo "$data" >> ./3-star-tmp.m3u
   ((i++))
  fi
 fi
done

i=1

while [ $i -le 45 ]
do
  file=$(sqlite3 ./clementine.db "select filename from songs where rating >= "0.8" and rating

  ### Clementine data encodes special characters and accent marks and stuff so I'm using
  ### Joel Parker Henderson's urldecode.sh to undo that: https://gist.github.com/cdown/1163649

  data=$(urldecode.sh "$file")
  if [ -f "$data" ]
  then
   ### Have to escape leading brackets because grep treated it as a range and would allow duplicates ###
   ### Can't do that in "data" because \[ isn't in the filename so they'll fail ###

   escaped=$(echo "$data" | sed 's/\[/\\[/g')
   #echo "$escaped"

   ### Avoid duplicates
   match=$(grep -i "$escaped" ./3-star-tmp.m3u)
   if [ -z "$match" ]
   then
    echo "$data" >> ./3-star-tmp.m3u
    ((i++))
   fi
  fi
done

i=1

while [ $i -le 25 ]
do
  file=$(sqlite3 ./clementine.db "select filename from songs where rating >= "0.6" and rating

  ### Clementine data encodes special characters and accent marks and stuff so I'm using
  ### Joel Parker Henderson's urldecode.sh to undo that: https://gist.github.com/cdown/1163649

  data=$(urldecode.sh "$file")
  if [ -f "$data" ]
  then
   ### Have to escape leading brackets because grep treated it as a range and would allow duplicates ###
   ### Can't do that in "data" because \[ isn't in the filename so they'll fail ###

   escaped=$(echo "$data" | sed 's/\[/\\[/g')
   #echo "$escaped"

   ### Avoid duplicates
   match=$(grep -i "$escaped" ./3-star-tmp.m3u)
   if [ -z "$match" ]
   then
    echo "$data" >> ./3-star-tmp.m3u
    ((i++))
   fi
  fi
done

cat ./3-star-tmp.m3u | sort -R > /Volumes/Filestore/CDs/playlists/3\ Stars\ +.m3u

rm ./3-star-tmp.m3u

xrayspx's picture

Dr Pepper Clock

Music: 

We just got a really great Dr Pepper restaurant clock, as featured in the arcade scene of Wargames. We think it's meant to be modular and can be attached to lighted menu-boards. It's got a pass-through 110V plug so you can daisy chain stuff off of it, and the side trim is easily removable to clip in a sigh board on each side.

However, while it is beautiful, it doesn't need to be on all the time, so I added a switch. The wiring is very simple with just a couple of wire nuts tying the inbound power to the AC plug and fluorescent light. I found some suitably retro-ey looking switches on Amazon at 6 for $9.

We decided to put the switch on the bottom just in case I cracked the plastic or something similarly horrible happened. All in all, very straight forward, but one pro-tip was in drilling the hole. I started with a small drill bit and worked my way up, but when I got to the larger drill bits I ran the drill in reverse to just use friction to make the hole rather than trying to hog out a hole in very thin and brittle material with a way-too-big bit. This worked /great/, thanks Dave.

Wiring before:

Switch:

Wiring After:

Switch external:

Fixed Tags:
xrayspx's picture

Caching Password passer

Similar to the RDP Launcher, I occasionally need to grab passwords that I use all the time from KeePassXC to paste into various forms or prompts. Basically anything I use more than once per day, I have defined in this script for quick access. I don't want to be able to remember these, and I also don't want to have to interact with the password manager UI if I'm in a shell.

xrayspx's picture

Sharp Boombox Repair

Music: 

Tom Petty - Freefallin'

RDPLauncher

TL;DR: Here's the Link:
RDPLauncher

I use RDP a lot and had some scripts to let me launch lots of RDP sessions without having to enter my random-generated passwords over and over. I wasn't happy with how I was handling those passwords so I've made it more secure using gpg and KeePassXC. Last night I made it compatible with Windows and MSTSC which will be uploaded here shortly once it's cleaned up a bit.

Basically I'll click a shortcut for whatever host, which runs my launcher. I get prompted for my GPG passphrase, which reads from an encrypted file containing my KeePassXC passphrase, which is then used to retrieve the user password for launching the RDP session.

Gpg-agent uses a cache-TTL to "hold the door open" for 10 minutes by default, so I can launch a bunch of sessions and only type my passphrase once.

Requirements:

- gpg client and running gpg-agent (gpg4win, etc) with a private key set up, etc.
- cygwin if you're running Windows
- KeePassXC (or some other key-store that has a command-line interface
to query the database. In the beginning I was just using the gpg file
with user/password pairs, so that works too)

The tool has a few neat features:

- If run from the command line with no arguments, it will prompt for user/pass/host/domain, good for one-off sessions to machines I won't log into much. That's great since I spend all my time in terminal windows and this stops me having to go back and forth to the mouse and keyboard while entering credentials.

- If launched with -b, it prompts you for information for a one-off connection, but will also build a new shortcut launcher from a template. So like for the first connection to a machine you know you're going to use a lot. (Linux/Mac only)

- Automatically tunnel sessions over ssh. This means I can launch RDP sessions on my Mac and they'll seamlessly proxy through my work laptop to the VPN.

For tunneling, I am taking an arbitrary range of 200 ports and incrementing them based on what's currently listening. If there's already a process listening on port 6201, then try 6202 etc until there's an open one. So I can easily open 20-30 ssh tunneled sessions each with its own ssh process which will close down when the RDP window closes. 200 is "probably overkill", which means it might just be barely enough in the real world.

The launcher shortcut mechanics are a bit different on my Linux and Mac machines so I split the -b script builder piece out based on OS. On Linux, I use KDE/Plasma, and so I generate these as KDE desktop files which look like this:

#!/usr/bin/env xdg-open
[Desktop Entry]
Comment[en_US]=
Comment=
Exec=/home/xrayspx/bin/rdplauncher.sh -h it-host.xrayspx.com -d xdomainx -u xrayspx
GenericName[en_US]=
GenericName=host.xrayspx.com
Icon=remmina
MimeType=
Name[en_US]=
Name=host.xrayspx.com
Path=
StartupNotify=true
Terminal=false
TerminalOptions=
Type=Application
X-DBUS-ServiceName=host.xrayspx.com
X-DBUS-StartupType=
X-KDE-SubstituteUID=false
X-KDE-Username=

On the Mac side, I use shell scripts with the extension .rdp (which conflicts with Microsoft's client, but I don't care since I never use their client anyway). Those just launch using Terminal, so it does pop a terminal for a fraction of a second, but I really don't have a problem with that. To get the Terminal window to close (and I do associate these files with Terminal.app specifically rather than iTerm2), open Terminal.app, go to the Terminal menu -> Settings -> Profiles (tab) -> "Basic" or whatever profile is your default -> Shell (tab). Choose what action to take when the shell exits. I have it set to "Close if the shell exited cleanly" and "ask before closing" set to "only if there are processes other than the login shell..."

The launcher for that looks like:

#! /bin/bash
rdplauncher.sh -h host.xrayspx.com -d xdomainx -u xrayspx &

I generate those from the KDE .desktop files with a command like this:

for host in $(ls | grep "\.desktop$" | awk -F ".desktop" '{print $1}'); do cmd=$(grep Exec $host.desktop | awk -F "xrayspx/bin/" '{print $2}'); echo "\!#/bin/zsh" >> $host.rdp; echo "$cmd &" >> $host.rdp; done

That creates .rdp files in the same directory as the .desktop files, so now they can be moved around, have chmod set, etc.

If I call it with AppleScript or Automator instead of a bash script as above, none of the password retrieval process works. I think it short circuits and sends the output back to the AppleScript rather than the bash script which ran the command. If I can get that working that would be ideal.

The mechanics on Windows are similar to the Mac method. a .bat file which launches the bash script via Cygwin:

C:\cygwin64\bin\mintty.exe -w hide -e /bin/bash -l -c '/home/user/bin/rdplauncher.sh -h host -u username -d domain'

On Windows at least the Cygwin window it creates is hidden from the user, so that's nice.

xrayspx's picture

Lots of RDP

Music: 

Annie Lennox - Why?

Do you do lots of RDP? Like lots and lots? I do, and even with password management it's annoying. I tend to use generated passwords for all my normal user, Domain Admin user and obviously Administrator accounts. That means lots of workarounds to deal with those passwords while doing bulk RDP sessions.

A typical use case for me is to RDP to 20 machines at a time, run a thing, wait, and log out. I've always scripted this, but not always in strictly the safest way. Plaintext passwords stored in a script, or read off disk. The philosophy is "if someone can read this script, I've already lost the game anyway", but still it's ugly and sick, and so I fixed it. In my defense, the Red Team never did pop my laptop...

I already use gpg-agent to facilitate unpacking of log files. On my syslog servers I roll logs over hourly, gzip them and then gpg encrypt them to my key. Then I can download a bunch of them, run my logunpack script, enter my passphrase once and since gpg-agent caches that credential for a period of time, decrypt all my files in one go.

What I wanted here was basically a way to have keepassxc.cli "hold the door open" and cache the passphrase like gpg-agent does. So what I've done is to use gpg-agent itself for that purpose. I have a GPG encrypted file containing my KeePass-XC passphrase, and I open it using gpg-agent, so it can be reused until gpg-cache-ttl expires.

I've also always had slightly different copies of this script for use cases of "Fullscreen on my laptop" and "fullscreen on larger displays", so I have a switch here for "resolution" as well. "fs" for fullscreen or "fsbm" for "big monitors". Since I'll never go to my office again, that's pretty much never going to get used. The default for the $res value will remain 1280x960. Reasonable enough.

I also added prompts so that it'll ask for host, domain, user and password if you run the script with no prompts from a shell. So /that/ will be super useful to me when I have to do a one-off connection to some remote host but don't need a whole launcher for it. While I'm at it, I put in the -b switch so that you can have it generate a launcher based on that input. That saves me hand editing a template when I add a new RDP host.

I use Linux, but this should work with minimal-if-any changes on Mac and Windows/Cygwin, both of which can run xfreerdp and gpg-agent. I have a good automated ssh-tunneled RDP setup for my Mac, so I might try using that with this so I can use a 4k display for those "busy RDP days".

Being that I do run Linux, here's how I launch this. KDE desktop files like this:


xrayspx@dummyhost:~/rdps$ cat windowsmachine
#!/usr/bin/env xdg-open
[Desktop Entry]
Comment[en_US]=
Comment=
Exec=/home/xrayspx/bin/rdplauncher.sh -h windowsmachine -d domain -u xrayspx
GenericName[en_US]=
GenericName=windowsmachine
Icon=remmina
MimeType=
Name[en_US]=
Name=windowsmachine
Path=
StartupNotify=true
Terminal=false
TerminalOptions=
Type=Application
X-DBUS-ServiceName=windowsmachine
X-DBUS-StartupType=
X-KDE-SubstituteUID=false
X-KDE-Username=

So anyway, here's the thing: RDPLauncher

xrayspx's picture

Rippin' DVDs

Music: 

Dana Carvey - Choppin' Broccoli

Today in Lattice of Convenience news, here's how to rip DVDs.

I barely understand the mencoder command that is the backbone of this thing, and there are many better ways to do lots of the stuff in this script, in fact I know several of those better ways, and looking at it fresh, I see some redundant stuff that cancels out other stuff. But it runs, and I use it, so here goes.

Ripping DVDs isn't fun, the disk labels are iffy at best, even within a single box set you might go from the Gold Standard "TV Show - S1D1" to "DVD_VIDEO" as a disk label. So it can get kind of ugly. To mitigate that I create an output folder based on the DVD disk label + a timestamp. If you get a run of disks with the same name, at least they're not overwriting each others files because the timestamp will shift. I currently have a dvdrip-output directory with the following DVDs in it:

...
DVD_VIDEO-090720202337
DVD_VIDEO-090820201025
DVD_VIDEO-090820201027
DVD_VIDEO-090820201142
I_LOVE_LUCY_S2_D1-090520202354
I_LOVE_LUCY_S2_D3-090620201047
LUCY_S1D1-090520201043
LUCY_S1D2-090520201043
LUCY_S1D3-090520201359
...

Those are all from the same box set. So that's 3 naming conventions from one series. To be fair I think that while it's the same company producing them they probably came as separate "season" boxes rather than one big set. Still. Come on. Jesus.

Another big gotcha I've hit, again mainly with TV series box sets, a single show might exist on the disk as many as THREE times. Once as a "standalone episode", once as "episode with commentary track" and once as part of a massive concatenated file of all the episodes on that disk. In the case of the commentary track, that audio seems to be separate, so the actual episode rips to exactly the same filesize, the commentary track seems not to be something I have access to, so you just get two identical files at the end.

So as you're ripping, that's going to triple the rip time.

The way I'm trying to fix that is to rip the first 30 seconds of every Title on the disk, then do a SHA sum on those ripped sample files. As a Title rips, when it's done I'll drop its clip checksum into a "rippedchecksums" file. The next TItle starts the first thing it does is check to see if its checksum has already been ripped. If it has, skip it. It seems to catch 100% of repeated Titles, and probably 70% of the "Big Concatenated File" cases will match the sum for Title 1. Saves a shitload of time.

In this case, Title 1 is a standalone episode, and Title 21 is the Big Concatenated File of all the episodes on the disk. Title 21 will be skipped. Since I get about 70 or 80 FPS on my Mac Pro, that probably saved 90 minutes of rip time or so with 3 hours of video on the disk:

763b6035c4bf239b4425fb8f484018387574baca /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/1-sample.avi
59cca1b18759647e13e3e1b6a4facace0520fc06 /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/10-sample.avi
125add4181b9dc6eee57c32c07568765b8e4483b /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/11-sample.avi
4daae35d014032964fe57e70e2cc3450f7dac4e5 /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/12-sample.avi
a942f31a9ee42c5839772f733b2c666195397ad5 /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/13-sample.avi
8c9473a940a9bc685d84e0ac29c66f53efa6667d /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/14-sample.avi
29d2200d8c46ac11417119b4b7179e4b526d99cf /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/15-sample.avi
466860b79bba6d132fcc97d6dc7c0c3a20dd771c /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/16-sample.avi
f4ae11cca0752956c4d6025a8760a260a59fe79b /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/17-sample.avi
00753d529f4bbf4081f647056cf44db7c630c198 /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/18-sample.avi
b7f9c9087fed6b00d22de5033c153f9ffb3cd3b1 /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/19-sample.avi
14efcb6164f1424b894cc28200ab621ec805ecd0 /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/2-sample.avi
6c411c8869f1e6bc9a6ec298ba9b6a5c9eefc9ae /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/20-sample.avi
763b6035c4bf239b4425fb8f484018387574baca /Volumes/Filestore/dvdrip-output/DVD_VIDEO-090720202337/21-sample.avi

At the end of it, I still end up with just a directory full of files labeled 1 through whatever.avi. I have to take a few seconds per file to get it to "TV Show - S01E01.avi". But from there FileBot can mass-rename them with episode titles.

So here's the full ugliness. You'll want to adjust all the paths. I should have made variables, but I don't care, I maybe have 3 or 4 ripping trays running at a time on various machines, so I don't mind just changing the paths for each host. Works on OSX and Linux, and probably Windows with Cygwin, but I don't care about Windows so I'm not going to test it.


#! /bin/bash

timestamp=`date +%m%d%Y%H%M`

id=$(drutil status |grep -m1 -o '/dev/disk[0-9]*')

if [ -z "$id" ]; then
echo "No Media Inserted"
else
name=`df | grep "$id" |grep -o /Volumes.* | awk -F "Volumes\/" '{print $2}' | sed 's/ /_/g'`

fi
name=`df | grep "$id" |grep -o /Volumes.* | awk -F "Volumes\/" '{print $2}' | sed 's/ /_/g'`
echo $name
dir="$name-$timestamp"
mkdir /Volumes/Filestore/dvdrip-output/$dir

maxtitle=`/Applications/mencoder dvd://100 -o bob | grep "titles on this DVD" | awk '{print $3}'`

for title in {1..100}
do
if [ $title -le $maxtitle ]
then
/Applications/mencoder dvd://$title -alang en -ovc lavc -lavcopts vcodec=mpeg4:vhq:vbitrate="1200" -vf scale -zoom -xy 720 -oac mp3lame -lameopts br=128 -endpos 30 -o /Volumes/Filestore/dvdrip-output/$dir/$title-sample.avi
shasum /Volumes/Filestore/dvdrip-output/$dir/$title-sample.avi > /Volumes/Filestore/dvdrip-output/$dir/$title-checksum
touch /Volumes/Filestore/dvdrip-output/$dir/rippedchecksums.txt
fi
done

cat /Volumes/Filestore/dvdrip-output/$dir/*checksum >> /Volumes/Filestore/dvdrip-output/$dir/allchecksums.txt

for title in {1..100}
do
if [ $title -gt $maxtitle ]
then
chmod -R 775 /Volumes/Filestore/dvdrip-output/$dir
sleep 3
drutil tray eject
exit 0
fi
sum=`cat /Volumes/Filestore/dvdrip-output/$dir/$title-checksum | awk '{print $1}'`
match=`grep $sum /Volumes/Filestore/dvdrip-output/$dir/rippedchecksums.txt`
if [ -z $match ]
then
echo "CURRENTLY RIPPING TITLE #$title"
/Applications/mencoder dvd://$title -alang en -ovc lavc -lavcopts vcodec=mpeg4:vhq:vbitrate="1200" -vf scale -zoom -xy 720 -oac mp3lame -lameopts br=128 -o /Volumes/Filestore/dvdrip-output/$dir/$title.avi
echo $sum >> /Volumes/Filestore/dvdrip-output/$dir/rippedchecksums.txt
rm /Volumes/Filestore/dvdrip-output/$dir/$title-checksum
rm /Volumes/Filestore/dvdrip-output/$dir/$title-sample.avi
fi
done
chmod -R 775 /Volumes/Filestore/dvdrip-output/$dir

xrayspx's picture

Bouncing from Kodi to EmulationStation, and back

Music: 

Ninety-Nine And A Half (Won't Do) - Wilson PIckett

Update:

----
As pointed out on the RetroPie forum, just add the loop in autostart.sh, duh: I searched for a while before writing this thing and if I'd seen anyone mention that I'd have just done that instead.

while :
do
kodi
emulationstation
done

I also think it makes a more sensible default for RetroPie to implement. That's all I actually wanted at the start.

However...

Now I've added Features. I can hijack my loop and add one-off commands.

So now there's a Desktop button in my Kodi main menu that will touch a file to cause the loop to gracefully exit Kodi and send me to a desktop session. When I leave the desktop session, it takes me back to Kodi. So that's pretty goddamn convenient.

-----

Because if there's one thing I love, it's having to sysadmin my TV.

Like most reasonable people I use a Kodi mediacenter to run my TV. Lately this has been on a Raspberry Pi 4 running RetroPie. Generally people boot RetroPie into EmulationStation and use it as an emulator, such as on an arcade cabinet. I'm also one of those people.

But in this case I primarily use the TV to watch TV shows and movies, but also want to run console games, so I upgraded to a better RPi and migrated from LibreElec to RetroPie.

RetroPie lets you choose whether to boot into EmulationStation or Kodi, which is fine, and the idea is that if you quit Kodi, it loads ES so you can play games. That works fine. Once. The trouble is in going the other way. If you quit EmulationStation, you exit to a shell. If you run Kodi from within the Ports menu in EmulationStation, well, now you're running both ES and Kodi. This also changes the behavior the next time you quit Kodi to play a game. You end up back in the Ports menu with Kodi highlighted, because ES never quit.

So, that's what I fixed.

The way the RetroPie tool works is they create a script at /opt/retropie/configs/all/autostart.sh. If you have Kodi booting first, it will have two lines:

kodi-standalone
emulationstation.

That script gets run at login time for the pi user. Basically it runs Kodi, and autostart.sh is still running. When Kodi exits, it runs ES and autostart.sh exits. If you wanted to you could just put 1000 lines of:

kodi-standalone
emulationstation
kodi-standalone
emulationstation
kodi-standalone
...

However that's ugly, so I kind of daemon-fied it with a bash script of my own that I wanged together in like 10 minutes, and then I launch that through their autostart.sh. I didn't want to replace their script with mine because the RetroPie one could get regenerated with an upgrade or if I hit something in RetroPie-config. It's safer to have their script call mine.

So what I do is I start with whichever application is passed to me in the command line:

autolaunch.sh -f kodi

Then I start an infinite loop and, based on what application the script is called with, it will start the first application. When that app exits, I change the value of the variable so that the next time it loops, it runs the other one:


#! /bin/bash

while getopts f: name
do
  case $name in
    f) fval="$OPTARG";;
    ?) printf "Usage %s: [-f application to start] args\n" $0
    exit 2;;
  esac
done

while :
do
  if [ $fval = kodi ]
  then
    kodi-standalone
    fval="emulationstation"
  elif [ $fval = emulationstation ]
  then
     emulationstation
    fval="kodi"
  fi
done

Downsides and ToDo's:

Obvious downside is that this makes it difficult to get a shell at the console of the machine. However, I can count on one hand the number of times I've had to do that in the last 6 years or so of running my TV from a Raspberry Pi, so I really don't care.

A definite ToDo is to add some level of process control and general safety so I don't somehow end up running a bunch of instances of Kodi and ES. I did test with "Restart Emulationstation", so it would pick up new games, and it seemed to work as expected. It didn't launch another instance of Kodi or anything.

My main ToDo is to have the ability to use more launchers. Basically right now I have a "Games" menu item in my Kodi main menu, I hit it, it just runs the Kodi "Quit" command, which causes ES to start. Same thing in ES, though I'm just quitting it using the context menu at the moment.

I'd like to be able to add a "Desktop Session" button to quit Kodi or ES and launch a desktop with a browser for those very rare times I want a browser on my TV. This would also solve the "can't get a local shell" problem, at least mostly. I could add a "quit to shell" in this way obviously as well. I think the best way to do this is to stop the script as I exit Kodi and restart it with a new starting value, like -f startx. Kind of like if it were a real system daemon.

However I think in my case, since I'm not a very good programmer, I'm going to just bang this out with a file in /var/tmp or somewhere which carries the "Next Command", so rather than update $fval as I am now, I'd check that file and have it read in each loop to set fval. That would allow me to hijack it from outside the loop.

So I'm in Kodi, if I quit, it's going to set $fval to "emulationstation" and load ES. However, if I run a shell script, and /then/ quit or killall kodi-standalone, that shell script can populate /var/tmp/nextcommand or whatever with "startx".

Then, when Kodi quits, it sets $fval to ES, the next loop comes, but instead of just launching ES, we check to see if there's a value in nextcommand. If there is, set $fval to that and run it instead.

Then you'll start an X session, and when that quits, it should take me back to Kodi.

I seem to recall Kodi's internal tools are pretty good, and I can combine "run this external command" with "run this internal 'quit' command" and assign that to a menu "Action". Just need to remember where all that stuff is.

xrayspx's picture

Hello Woodgrain My Old Friend

Music: 

A couple of weeks ago Natalie bought an Atari 2600 for me as a present. It came with 60 games, two joysticks and paddles, so it was a pretty good deal. She did ask the seller if it worked, and she said it did.

When Natalie got it home, I took the 2600 apart to make sure nothing had leaked all over the place or was otherwise obviously broken or loose and it was SUPER clean, which was encouraging. Just some minor nicotine film everywhere, but it was even dust-free inside. So I hacked together a super crappy cable to convert from the Atari's RCA RF cable to coax to get it hooked up to the TV. Unfortunately it didn't seem to work after all. Natalie was pretty bummed out, but I decided that she'd tried to get me a present, and instead got us a project, and so I convinced her not to bother the seller or leave negative feedback or whatever. After all, as far as the seller knew, the last time she saw it hooked up, it probably did work.




As it was, it did the same thing whether there was a cartridge in or not, just various interference patterns:

The 2600 is like the VW Bug of electronics. There's not that much in there to break, what can break is relatively easy to repair even for rank amateurs like us, and they sold 30 million of them, so there are loads of parts. We'd make it work, and we'd learn A Thing. And if it didn't work at the end, just dip in and grab another console and try again.

To that end, we bought a $5 RCA -> Coax converter, a 2600 re-cap kit, and went way overkill on a swell Hakko solder station to replace the garbage iron I had and hated using. I also grabbed some junk electronics to show Natalie what we're trying to do. We're gonna heat this stuff up, get these components out, get all the old solder out and put new components in. I thought of doing a composite video conversion at the same time, but I kind of wanted to take it one step at a time.

There are only like 5 caps in here to replace, and they also sent a replace voltage regulator. I had previously tested the voltage regulator and it was fine, (near enough) 12V on the input and 5V output, so I opted not to replace it. We split the re-cap duties so both of us could get in some practice on the new soldering iron. It worked great, 0 -> 600 degrees in seconds. We pulled the 3 socketed chips out and hosed the sockets with contact cleaner along with all the switches and the cartridge port.

After that work, I hooked it back up again and....Definite Progress:

A staticky display, but definitely playable. Joysticks work, paddles work (after a hosing out with the brake cleaner). I decided to stop touching anything until the RCA->coax converter showed up, since I was 90% sure the problem was probably that janky as fuck cable I gooped together. Natalie had tracking on it, and they said it was delivered "Today" to our same street address, but 3 towns away. It turns out they'd sent the wrong tracking number, however the correct tracking number showed it as having been delivered a week ago. A quick harried rummage through a week's worth of opened shipping envelopes and we found it. Hooked it up and, yeah, I'd say this is pretty much handled:

I see no point in doing a composite conversion. This is 100x better than any image you'd ever get off of one of these when they were new. If the RF modulator starts to fail, it's always an option later on.

So now the 2600 is in its place along side the arcade cabinet. Of course all of these roms are available in emulation, and I've got an archive of nearly all 2600 games. If we really want to play some serious Fishing Derby, we can do so there. But it's really nice to have a functional 2600 with original games. Natalie is totally enamored with this whole thing and is already stacking up her favorite games. It's totally likely that we'll flip the whole game-time layout and put the arcade cabinet on music video duty while we play Atari on the main TV.

There are still games that don't work, and we often have to switch the console off and on once or twice to get a game to start. Some are also pretty susceptible to any nudge to the console, so we'll probably get it back apart, clean the cartridge port properly, make sure all the contacts are good and re-tin them if they aren't, and see if that helps. But really, this instability is par for the course for 8 year old me. It's part of the charm


--


I have a very close long-time friend who's been trying to mentor me into doing electronics projects since...1988 or so? I never got any good at it. I only ever had one of those garbage un-regulated plug-it-in-and-wait-10-minutes soldering irons that never really got hot enough. I couldn't be bothered at the end of the day. I was too busy playing games to care. I think he was pretty happy to see me tackle even such a simple project and dip our toes in.

At the end of it, his only comment was "Working with electronics takes on a different dimension if you have a scope. Fun to see what is happening in a circuit.".

So I guess we gotta get project-ing.

Fixed Tags:

Pages

Subscribe to RSS - Hacks