dead .conf

Viewing 3 posts - 1 through 3 (of 3 total)
  • Author
  • #68

    I just grabbed HEAD and rebuilt and ran. Everything worked fine until I tried to run it again (I run it in the foreground and quit with ^C). This time the .conf file looks like it’s been converted to a new format, only all of its nodes are (null). The last time I updated/ran was a few weeks ago, so perhaps I missed something in the forums. Has the .conf been upgraded?

    # mt-daapd.conf
    # Edited: Sun Jun 12 18:48:08 2005
    # By: joe
    web_root (null)
    port (null)
    admin_pw (null)
    mp3_dir (null)
    servername (null)
    runas (null)
    playlist (null)
    extensions (null)
    ssc_extensions (null)
    ssc_prog (null)
    db_dir (null)
    rescan_interval (null)
    scan_type (null)
    process_m3u (null)
    compress (null)

    Ahh, nevermind about the .conf upgrade biz. I see what happened: I used the status html pages during the first run and triggered a bug in config_handler() that clobbered the .conf file.

    What I did was:
    1. access status.html
    2. click ‘rescan’
    3. strip the ‘?rescan’ from the url and reload the page

    This was dumb, since the resulting url was ‘/config-update.html’ and not ‘/status.html’. I wasn’t paying attention. This causes the code to redisplay the status.html page (correct), but also fall through a case where the config file gets rewritten (with (null)) (incorrect). Dumb user mistake on my part. Not even really worth fixing, but I’ll take a stab and send Ron a patch.

    These aren’t the droids you’re looking for.


    Ah, but that’s kind of a nice feature, the fact that a simple mistake renders your config file useless.

    Perhaps that’s a bug. :/

    — Ron

Viewing 3 posts - 1 through 3 (of 3 total)
  • The forum ‘General Discussion’ is closed to new topics and replies.