Nijel's weblog | blog.cihar.com

It is finally out :-). Only minor changes in docs and build system to fix win32 build.

Some time ago, I switched python-gammu and Wammu RSS from Mantis generated to my weblog. However there are still many people using old locations which are redirected to new place. The redirection won't stay there forever, please subscribe to actual locations, you can find them on project pages or direcly on blog pages in all RSS and Atom variants.

As it turned out that my previous blog post is first returned by Google while searching for this company, I got some feedback about them. They seems to send spam, attempt to steal site keywords to get before them and do other simmilar silly things. If you want to know more, just try more Googling or read discussion.

If you want to fight them, either block them by iptables:

iptables -A INPUT -s 208.66.195.0/28 -j DROP

Or deny access to webserver (this example is for lighttpd I use):

$HTTP["remoteip"] == "208.66.195.0/28" {
    url.access-deny = ( "" )
}

Yes I know there is gammu 1.07.00 which should be packaged, yes I know there is python transition ongoing which needs changes on my python packages, yes I know that there is new libbluetooth2 in archives which need gammu rebuild. Please be patient I will return back to packaging work as soon as I will have internet connection at home.

By that time you can also expect new python-gammu release which will be just minor update to fix some urgent things.

In order to allow inclusion in distros like Gentoo or Arch, I decided to add "stable" builds of dictionary. Those will stay in place forever (well at least long enough) so you can safely reference it in build recipes. You can find them in separate directory on download server.

I have no clue whether python-support or python-central is better solution as new python packaging helper. I follow debian-python list too shortly to know all backgrouds between their authors. Despite of those current situation is really bad.

There were filed many bugs and tools and policy still change every day. Will it use Python-Standards-Version or debian/pycompat? Will there be XS-Python-Version or debian/pyversions? Now it looks that both. I thought that new policy shoud simplify things and not making them even more complex. Calm down people and try to find one solution.

New beast called Internet Explorer 7 is not yet released, but it seems to show some of new features. They broke something in CSS and because of that text won't display in phpMyAdmin textareas. According to them it is problem in our CSS, but they won't tell where exactly. Why the hell our CSS works fine in all browsers (including MSIE 5.5 and 6) and now suddenly it is broken?

See bug report in phpMyAdmin's tracker for more info.

I again noticed that multimedia repositories don't work for a while. Today I finally spent few minutes with Google to find new place :-). It looks that these packages got own domain - debian-multimedia.org. So if you look for these packages, change your sources.

I just did large cleanup of data in Gammu phone database. So there should be no duplicates and all phones should be in correct category. Please take care on selected manufacturer when you add new phones as otherwise people lookig for phone from specific manufacturer are confused.

I also fixed case of some letters in names, however I'm definitely not sure whether new variants are 100% correct as I don't know all phones :-).

As I live in another part of city since last Wednesday, I also have to find new ways for my bicycle. I know how poor situation in Prague is, but I hoped that marked way I noticed while moving will be good for travel to work.

But it looks like expected too much. Maybe somebody forgot to tell people marking this way that most people don't use bicycles on stairs. I don't know why does this way in about 3 kilometers I used it go over three stairs.

If you want to go through Prague on bike, forgot to use marked ways, you will most likely end with troubles. Better find some local people who will guide you.

During yesterday afternoon I moved most of my things to new flat. I didn't expect I have so much things :-). However it went smoothly, I only feel really tired today.

There is still lot of work there, but we will hopefully manage it soon. One of first thing is internet connection, which is must for me. Until it is ready don't expect me to react quickly on emails and or bugs. Anyway I'll try to handle regullar mails during day at work.

On request I enabled generation of package lists for all architectures in my repository. So you can now install startict-english-czech on all arches directly using apt without need to manually download this package.

And not this does not mean I will build binaries for other arches, they will be still i386 and amd64 only, others will contain just Arch: all packages.

Yesterday Google released beta version of Google Earth which runs natively on Linux. As it is great toy I wanted to immediatelly test it :-). However it ran damn slow and libgl was complaining about wrong structure sizes:

ERROR! sizeof(I830DRIRec) does not match passed size from device driver

After some Googling I found bug #359328 which tells to use CVS snapshot of Mesa from experimental. Any yay, it works then. Now back to work let's contiue with playing later.

It looks like somebody in company Digital Infinity Ltd loves to read my blog. Maybe they should consider setting up personal mirror as if every computer in their address space is going to download whole blog as they did it today, it will cost them lot of bandwidht and I will get really upset and iptables will get new rule about 208.66.195.0 - 208.66.195.15.

So it's up to you, you have been warned.

I just finished make distclean in my CDs. In past years I collected many data CDs which I will hardly use in future. Many of them contains various software for Windows I used in past, rest are free applications which were given with some computer magazines. They are anyway obsolete by now even if I'd use Windows.

So about hundred of CDs travels to waste bin for a long jurney to landfill :-).

When I noticed that something like indexing services for blogs exists, I hacked simple dog command to download ping page. This worked more or less, but it didn't gave me any feedback. Today I decided it's time to switch to something more sophisticated.

As most (if not all) services are using XML-RPC with same functions, implementing this in python using xmlrpclib was quite easy. After few minutes reading documentation I had working code which can be attached to publish command in nanoblogger:

import xmlrpclib
import sys

def do_ping(url):
    print "Pinging %s ... " % url,
    sys.stdout.flush()
    try:
        s = xmlrpclib.ServerProxy(url)
        r = s.weblogUpdates.ping("Nijel's Weblog", "http://blog.cihar.com/")
        if r['flerror']:
            print "FAILED"
            print " error: %s" % r['message']
        else:
            print "ok"
        sys.stdout.flush()
    except Exception,v:
        print "FAILED"
        print " error: %s" % str(v)

do_ping("http://ping.blo.gs/")

I'm only affraid that ping services don't like me much now as during testing I produced lots of pings :-).

More than month after starting this blog, it's time to summarise experiences with nanoblogger. I enjoy blogging this way much more than previously used web based interface on abclinuxu.cz. Partly this is caused by change in editing (vim is much better than textarea in browser), another reason definitely is markdown which allows me to forget about HTML in most cases.

Anyway this way leads to creating many short posts, while previously I usually posted longer stories. I don't know which way is better and readers do prefer :-). So I will try to keep rate on approximately one (short) post per day.

And back to nanoblogger. It's not perfect system, however I was not able to find better suited for me. Generating blog is quote slow, but I don't care much. I had to switch to CVS version because some URLs were wrong in rc. And I had to modify even this CVS version to fit my needs, but those were only small changes. If you want text editor driven blogging system without comment support, nanoblogger might be good choice for you, otherwise you probably need something else.

I today spent quite a lot of time while trying to setup new television OVP CTV 2991 RF DVB-T. The only reason why we choose this one is because it is cheap and has integrated DVB-T tuner. Set top box was not an option because it would be too complicated for grandma who will be the only user.

The DVB-T tunes in this TV seems to be crap. It requires much stronger signal than my Hauppauge Nova-USB. Why the hell they can not make simmilar quality receiver as in cheap USB adapter? And don't trust anybody who will tell you that DVB-T can be received on anything.

Upon several requests I added another demo which is configured to use config authentication, so you can again directly link subpages without need to enter login information.

While working with that I also removed 2.8.0 branch from demo and CVS snapshots as it is not maintained and updated anymore.

After quite long searching we finally managed to choose new place to live. Looking was hard due to requirements on low price and good quality, however I thing result is quite good. It's located at the end of Prague near underground station Opatov.

Moving will start next week, hurray.

Have you ever need to lookup something in source code? Most developers probably did so. I needed to lookup something in Gammu sources and I decided to setup search engine over it. I decided to use Gonzui as it looks good and was easy to setup. So if you want to find something in sources, just point your browser to gonzui.cihar.com.

Later I added also other my code (phpMyAdmin, Wammu, python-gammu, polld), so you probably want to use advanced search to limit results.

Finally recent Gammu version (1.06.00) has been migrated to testing. It took so long as it needed manual hint because of both python-gammu/gammu breakage loops. Unfortunately there is no other way to resolve unstable Gammu ABI than using so strict dependencies.

Now I finally undestand why Gnokii is using that complex way for passing parameters. It is painful to use it, but it keeps ABI stable enough. We need to find better solution for this.

I never knew there are few special zones in DNS to grab infromation about DNS servers. Today while browsing PowerDNS recursor configuration, I noticed several config options defining replies on those special queries. So if you want to know which server is on the other side, you can try querying version.bind, version.pdns or id.server (CHAOS type, TXT class):

$ dig version.bind chaos txt

; <<>> DiG 9.3.2 <<>> version.bind chaos txt
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 56291
;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;version.bind.                  CH      TXT

;; ANSWER SECTION:
version.bind.           86400   CH      TXT     "PowerDNS Recursor 3.1.1 $Id: pdns_recursor.cc 838 2006-05-19 14:35:27Z ahu $"

;; Query time: 0 msec
;; SERVER: 127.0.0.1#53(127.0.0.1)
;; WHEN: Fri Jun  2 15:30:06 2006
;; MSG SIZE  rcvd: 119

Debian package for recent mpdscribble is ready. For now it's available only on usual location, upload into official archives will need sponsor as current one is not available. I hope to find one soon.

The idea is simple, but I haven't seen this before. Place bot on IRC channel, collect URL and generate web page which shows linked images. This is exactly what irc.tnet.no does. Big brother is becoming reality. Maybe sometimes we will have also analysis what people really talk about, but images can tell quite a lot.

Anyway if you want to know what people search for, Slovak search engine provides you this possibility under URL containing well known year 1984 :-).

I just wondered, whether it is possible to receive more ham than spam emails. After 12 hours server outage, I can say that it is possible. However it takes only one hour and then spam again wins. At least it is good that spam bots are still not clever enough to retry sendins, they simply drop everything what could not be delivered. I only hope that none regullar mail server does it also, otherwise I'd lost that message.

I sometimes random page on internet, where somebody in blog, wiki or whatever complains that something in Wammu or Gammu is broken. Last place where I noticed it was phone database. Why don't you report such problems to bug tracker, where you have much higher chance to make your problem fixed?

I know it is more work to collect all needed information, but if you want to use the software, it should be worth of it for you.