For my primary desktop is KDE which is running the Plasma Desktop plasmoids are a powerful option to add helper widgets to this desktop. Looking for a small, space efficient and simple tool to show cpu usage, memory usage etc. there was none I liked on kde-apps.

Enter CPU and System Viewer

The CPU and System Viewer plasmoid provides a simple and rather minimalistic view onto the current CPU usage of your system and can be added to the panel. If wanted, CPU, memory and swap usage are shown in a graph, too. Also, a popup gives more advanced information about the current state of the system.

This plasmoid is written in python and is derived from the Simple CPU Meter plasmoid. Also, ideas from the System Load Viewer plasmoid shipped with KDE SC were used.

And hence my first foray into KDE Plasma hacking commences.

No, I don’t like social media stuff as it delivers to much noise and wastes to much time with to little outcome. Yet, there are some – rare – occasions where a twitter feed of a certain person or group might offer valuable pieces of information that I can not access otherwise.

For this purpose – normally – I use RSS, so that any news automatically pops up in my mail programm and feed reader. This way, there is no need for an extra communication channel and yet another portal, yap, to check regularly.

Unfortunately, twitter hates RSS so that the former option to subscribe to a feed is no longer shown. Yet, there are workaround, see here

Thus, twitter hates me. Yet, I do have my feed.

Ever wondered how to schedule an automatic file system check on the next reboot of your linux server? This is how you do it.

Following the information given by everflux create a forcefsck file in your root directory:

 touch /forcefsck

And if you have a hetzner server, please be advised that this work only if you have changed your pass statements for the relevant partitions from 0 to 1 (root partition) or 2 (other). Otherwise, your request for a filesystem check will be ignored.

PS: Personal note, of course 🙂

Personal Note: If you want to mass rename a collection of files via the shell in linux, do as shown here or here. I. e. feed some regular expressions to the rename command followed by a filter which selects the files. Just like so:

rename oldname newname *.files

Or in my use case a follows:

rename s/JPG/jpg/ *.JPG
rename s/DSC/DSD/ *.jpg

While I am not completely happy with Thunderbird – the old and quite slow UI can be quite cumbersome – it is a nice application to offer an aggregated view onto all my mails, news and feeds. That’s something I like 🙂 And with Memotoo syncing to my Nokia Phone is easy and works flawlessly. Which I like even better 🙂

With one exception…. on both of my news accounts I do experience incorrect values for the counter showing the number of unread news. More precisely, I do end up with all visible mails being shown as unread yet the counter still showing unread messages. Well, not only me…

And if you are using shortcuts to jump to the next unread message like N you end up in these folders. Folders, where there is nothing to read for you.

Looking for a solution I found some comments which required the manual editing of some config files within the thunderbird profile. Doable, yes. But quite a nuisance if you have to do this every n-th day. And then I found the following python script on Reinouts site which does this automagically.

Python Skript:

 CONFIG_FILE = ('/home/reinout/.thunderbird-3.0/0gahkqdp.default/' +
               'News/newsrc-news.gmane.org')

 def fix_thunderbird():
    """Fix the thunderbird newsrc settings

    The settings sometimes contain lines like::

      gmane.comp.python.distutils.devel: 1-12428,12431-12446

    There's a two-article 'hole' in there that shows up as two unread
    messages.  This script removes the holes.

    """

    lines = [line.strip() for line in open(CONFIG_FILE).readlines()]
    print "======= OLD ======="
    for line in lines:
        print line
    print "======= NEW ======="
    outfile = open(CONFIG_FILE, 'w')
    for line in lines:
        if not line:
            continue
        newsgroup, messages = line.split(': ')
        parts = messages.split('-')
        new = '%s: %s-%s\n' % (newsgroup,
                               parts[0],
                               parts[-1])
        print new,
        outfile.write(new)
    outfile.close()

 if __name__ == '__main__':
    fix_thunderbird()

What happens if you have a hanging server with next to no recognisable computation going on and with plenty of free RAM to spare? You start exploring….

Yes, this is exactly what happened recently. In addition to vmstat and iostat mentioned quite a while ago hereby I discovered iotop.

iotop is a quite nifty tool which shows you exactly which process on your machine is using up which amount of IO resources. Nice to have if you are looking for your resource hogging application. Also, some information is given about iowait, i. e. the amount of time the complete system is actually waiting for a write operation to complete. Hence, you can determine whether IO is actually a bottleneck or not.

And yea, yea, …., yea: This is a personal note 🙂

Another personal note….

Thunderbird does not adhere to any of the time or date format settings defined via KDE system settings. So, if you want to go for a 24 hour clock as used in Europe, a German keyboard and yet stick to the general English language you need to change the global locale settings of your machine, as e.g. described here.

Still easier, log into GNOME -> System -> Administration -> Language Support and set the format to English (United Kingdom) or English (Denmark) in the Text tab. Finally, use the Apply System-Wide button.