ivman is dead, long live halevt

It’s been a while since ivman stopped working on my Gentoo box but I never had the time nor the willingness to take a look into it. It appears that ivman is incompatible with some newer versions of hal and dbus. The good thing is that there’s an alternative, it’s called halevt and as far as I’ve taken a look into it the configuration options look quite straightforward.
For Gentoo, there are ebuilds for halevt on Gentoo bugzilla, which install just fine.

In my point of view there’s an issue here for Gentoo. Latest ivman (sys-apps/ivman-0.6.14) compiles just fine against all of its dependencies, but then it does nothing at all when a deviced is plugged in. If the devices are present when ivman starts then it can detect and mount them, if you plug the devices after ivman is started, then ivman does nothing at all. I think ivman is broken since hal 0.5.9.X versions. Gentoo developers stll keep ivman in the stable tree though. I find no real logic to this decision. Ivman is buggy with current stable hal and dbus. I would prefer a de-stabilization of ivman or even a package mask for it. What’s the point in keeping a package (ivman) in the stable tree since it requires not the latest stable but an older version of another package (hal) ? IMHO, since they correctly decided to stabilize hal 0.5.11-r8, which subsequently rendered ivman useless, ivman should be wiped from the stable tree.
Some bugs on ivman reported on Gentoo Bugzilla: http://bugs.gentoo.org/buglist.cgi?quicksearch=ivman

I once used ivman with a couple of custom scripts to create/remove icons of automounted devices on my ROX desktop. I think I can make these scripts work again with halevt…I am in the process of rewriting them. More on that in the following days…

7 Responses to “ivman is dead, long live halevt”

  1. May 28th, 2009 | 14:47
    Using Opera Opera 9.64 on Linux Linux

    Just open a bug to inform developers that stable ivman + stable hal is a b0rked combination. They might not be aware of this because they might not using ivman

  2. June 21st, 2009 | 13:49
    Using Mozilla Firefox Mozilla Firefox 2.0.0.14 on Fedora Linux Fedora Linux

    Hi! We are going to start the same task for our little project http://minino.galpon.org

    Googling for information about this topic, we have found your plans. Would you like we exchange ideas about rox auto-mount using halevt?

  3. June 22nd, 2009 | 11:28
    Using Mozilla Firefox Mozilla Firefox 3.5b99 on Linux Linux

    Hello Antonio,

    I have already created a script to mount devices using halevt and making them appear on ROX desktop. I will post it very soon on this blog!

    I’d be glad to exchange ideas on any topic you wish, you can find my contact address at: http://www.void.gr/kargig/blog/contact/

  4. Gary
    October 3rd, 2009 | 18:10
    Using Mozilla Firefox Mozilla Firefox 3.5.3 on Gentoo Linux Gentoo Linux

    Thank You!

    I’ve been wracking my brain around why ivman only detects stuff when it is started.

  5. dborca
    October 15th, 2009 | 05:35
    Using Mozilla SeaMonkey Mozilla SeaMonkey 1.1.17 on Linux Linux

    @Gary:
    The breakage is because ivman calls xmlCleanupParser() each time it finishes with a certain xml document; which is incorrect: http://xmlsoft.org/html/libxml-parser.html#xmlCleanupParser

    @kargig
    I’d like to take a look at halevt, but I really abhor having to install yet another dependency (BoolStuff) 🙂

  6. October 18th, 2009 | 13:29
    Using Mozilla Firefox Mozilla Firefox 3.5.3 on Linux Linux

    Well, I don’t like having another dependency installed as well but there is no other alternative since ivman simply does not work as it should.
    Thanks for the information about xmlCleanupParser() though 🙂

  7. February 10th, 2010 | 15:21
    Using Mozilla Firefox Mozilla Firefox 3.5.7 on SuSE Linux SuSE Linux

    Patch for fixing ivman available at http://sourceforge.net/tracker/?func=detail&aid=2945321&group_id=109696&atid=654289 (two working versions 1) move up the g_thread_init()/dbus_g_thread_init() calls, 2) remove xmlCleanupParser calls).

Leave a reply