Yesterday I have uploaded the second release candidate of GnuTLS 2.8 (i.e. version 2.7.12) to experimental. Apart from the usual shiny new features there are two changes to the infrastructure.

• libgnutls(-extra)-config and the respective m4 test AM_PATH_LIBGNUTLS have been removed. This breaks building of any packages relying on these scripts. Since gnutls ships .pc files, fixing this by switching to pkg-config is simple. Using rebuildd I have already done testbuilds of all packages in sid/main which build-depend on libgnutls-dev and have submitted bugreports where necessary. Due to the number of involved packages I have not done that for all binary packages that link against libgnutls but do not Build-Depend on it. If you are a maintainer of such a package please make a test build.
• Symbol versioning has changed. libgnutls provides a number of symbols that are not part of the API but are needed by the "sister" libraries libgnutls-extra and libgnutls-openssl. These symbols are now versioned PRIVATE instead of GNUTLS_1_4. This should not break any applications, but please give your gnutls-using packages a short testrun with the library in experimental. Thank you.

Due to the later date of easter this year we had a rather long season this year. I started snowboarding on 2008-12-06 and stopped today 2009-04-19. Following up on previous entries here is the balance sheet:

	2005/06	2006/07	2007/08	2008/09
number of (partial) days	25	17	29	37
Dam ls	10	10	5	10
Diedamskopf	15	4	24	23
Warth/Schr cken	0	3	0	4
total meters of altitude	124634	74096	219936	226774
highscore	10247m	8321m	12108m	11272m
# of runs	309	189	503	551

In the second half of the season we actually had a lot of snow, up two 130 cm down at 800m above sealevel. Complete snow cover lasted from December until mid of April. Riding was good, I did not have any major injury, my skills did not increase though. The addition of a snowboard, a SG Speed 173 to my quiver worked out well. Loads of fun and less danger in slushy conditions (nose-in) compared to the shorter slalom board. Now the waiting starts. ;-)

Having some fun on snowshoes:

... early. yesterday:
today:

Good news everyone, after long time of inactivity and huge development over past month, new stable Gammu release is out. Since 1.20.94, there are only minor fixes in handling some corner cases in AT parser, but full list of changes since last stable 1.20.0 is quite long:

• Handle properly empty SMSC (bug #722).
• Disable AT/OBEX for Motorola K1v (bug #721).
• Automatically switch to nodtr connection if DTR/RTS setting fails.
• More verbose debug messages in API.
• Free phone data after testing model to avoid memory leaks.
• Properly handle if IMEI is too long (bug #709).
• Try to handle situation when CMGL does not list all messages.
• Fix typo in MySQL file (bug #716).
• Better parsing of USSD replies (bug #718).
• Properly restore caller groups on Nokia 2630 (bug #706).
• Fix listing of used calendar locations on Nokia (bug #710).
• Properly encode class and coding to SMS (bug #714).
• SMSD for MySQL no longer fails when sending message (bug #715).
• SMSD stores signal and battery state in PostgreSQL.
• Properly handle multiline firmware information.
• Fix crash when decoding EMS pictures (bug #549).
• Motorola phones sometimes reply in UCS2 instead of HEX (bug #712).
• Use native functions sleep/usleep if available.
• Improve FBUS initialization sequence.
• Fixed reading for ringtones from Nokia 5510 (bug #529).
• Do not crash if ringtone name is too long (bug #529).
• Fix possible crash while reading of messages using AT.
• Increase timeout when writing SMS message (bug #596).
• Cleaned up issues with reading from Bluez (bug #670).
• Fixed Windows binary (define properly MSVCRT to use) (bug #683).
• Fixed compilation with CURL on Windows.
• Windows binary now has only MySQL and CURL support (bug #683).
• Fill in timezone information (bug #696).
• Detect reserver message type (bug #691).
• Increase timeout for switching to OBEX (bug #690).
• Disable postal address for Nokia 2600 (bug #659).
• Encode phone number only with phones which need it (bug #654).
• Increase timeout for writing memory entry (bug #654).
• Handle properly when phoned does not tell SMSC (bug #616).
• Initial support for waiting calls in AT engine (bug #551).
• Fixed dialing on some AT phones (bug #471).
• Support for skipping location when reading (bug #467).
• Improved detection of serial port Bluetooth channel.
• SMSD reconnects to MySQL (thanks to Wouter D'Haeseleer).
• SMSD stores signal and battery state in MySQL (thanks to Wouter D'Haeseleer).
• Initial support for reading messages in simple text mode.
• Report if monitor failed because of an error.
• Handle not found errors on OBEX connect (bug #705).
• Report more verbosely that some feature has been disabled on compile time.
• Add ID for Nokia 5200 (thanks to Andreas Metzler).
• Add two IDs for Nokia 2600 (bug #659).
• Add ID for lot's of Sony Ericsson phones.
• Use AT parser for AT+CPBS? reply.
• Use all possible ways to get memory status (bug #687).
• Detect more error codes from Nokia phones (bug #659).
• Fix decoding of text using iconv (bug #687).
• Fix linker flags check, caused failures on MacOS X.
• Add second ID for Nokia 2610 (thanks to Nicolas Forget).
• Add ID for Nokia 6060.
• Try to use OPP and PBAP Bluetooth profiles also (not tested!).
• DumpMessage is not printing last char (bug #692).
• Support for favorite messaging number (bug #693).
• Add ID for Nokia 2630 (bug #693).
• First attempt for Bluetooth support on OS X.
• Do not open debug file for second time when using global debug.
• Fix DC memory name printing (bug #698).
• Add API function GSM_MemoryTypeToString.
• Report correct size of DC/RC/MC memory on DCT4 phones (bug #699).
• Store favorite messaging as phone number, not only as ID (bug #693).
• Avoid empty line at beginning of vCard backup (bug #701).
• Add ID for Nokia 2760.
• Catch not acceptable errors in OBEX (bug #702).
• Operation not allowed is not an unknown problem (bug #678).
• Add ID for Nokia 5000 (bug #677).
• Increase limit of files for DCT4 and check array size (bug #676).
• Allow to enter PIN code from stdin (bug #633).
• Handle HEX reply instead of UCS-2 from Motorola (bug #697).
• Avoid decoding of @ in email address (bug #697).
• ABI change, because of increasing some limits.
• Config file option rsslevel no longer works, superseded by 'gammu checkversion'.
• Dropped GSM_ReadHTTPFile from API, gammu now uses CURL instead.
• Support for longer notes on Nokia phones (bug #607).
• Add deleteallmemory command.
• Ignore trailing spaces in AT commands.
• Fix crash on long lines in vCards.
• Compatibility with broken iWOW responses (bug #650).
• Do not use hard delete flag for updating entries (bug #634).
• Identify Nokia 1200 (bug #649).
• Fix parsing of date/time reply on Sony Ericsson G900 (bug #647).
• Increased limit for folder name (bug #646).
• Fix parsing of empty date (bug #643).
• Recognize general number on Nokia 6230i (bug #655).
• Fix writing of phone number using HEX charset (bug #512).
• Recognize more CME errors, thanks to Paolo oopla@users.sf.net.

You can download from usual place: http://cihar.com/gammu/, Binary packages will be available soon.

I am back from a really nice holiday, driving around in the Provence. Best stuff:

• Driving. If you are not street smart roundabouts work a lot better than crossings with traffic-lights and separate lanes. Instead of having to (try to) change lanes at the last moment, you just enter the roundabout and take the correct exit. If you are slow just drive around twice. Also French drivers seemed to be less stressed out than Austrian ones.
• Driving in the Alpes Maritimes. The really sharp hairpin turns are fun with a small car. (We had a Citroën C1.) The absolute highlight was the road from St-Saveur-sur-Tinée to Roubion. A narrow street in an impressive gorge. (I am grateful we had almost no opposing traffic.)
  
• Fromage. (We even stayed in Banon for two nights.)
• Parc National du Mercantour. Great hiking, and we saw eagles in the wild.
  
• Scenery: Loads of it. Gorges du Verdon was as impressive as I could hope for, from Mont Ventoux (basically a huge heap of gravel) we saw up to Mont Blanc, then there was the rocky coast and the sea at the Cote d'Azur, etc.
  

New testing version of Gammu was just born. It brings quite a lot of fixes -- I simply release it because changelog got too big :-). Full list of changes:

• Report more verbosely that some feature has been disabled on compile time.
• Add ID for Nokia 5200 (thanks to Andreas Metzler).
• Add two IDs for Nokia 2600 (bug #659).
• Add ID for lot's of Sony Ericsson phones.
• Use AT parser for AT+CPBS? reply.
• Use all possible ways to get memory status (bug #687).
• Detect more error codes from Nokia phones (bug #659).
• Fix decoding of text using iconv (bug #687).
• Fix linker flags check, caused failures on MacOS X.
• Add second ID for Nokia 2610 (thanks to Nicolas Forget).
• Add ID for Nokia 6060.
• Try to use OPP and PBAP Bluetooth profiles also (not tested!).
• DumpMessage is not printing last char (bug #692).
• Support for favorite messaging number (bug #693).
• Add ID for Nokia 2630 (bug #693).
• First attempt for Bluetooth support on OS X.
• Do not open debug file for second time when using global debug.
• Fix DC memory name printing (bug #698).
• Add API function GSM_MemoryTypeToString.
• Report correct size of DC/RC/MC memory on DCT4 phones (bug #699).
• Store favorite messaging as phone number, not only as ID (bug #693).
• Avoid empty line at beginning of vCard backup (bug #701).
• Add ID for Nokia 2760.
• Catch not acceptable errors in OBEX (bug #702).
• Operation not allowed is not an unknown problem (bug #678).
• Add ID for Nokia 5000 (bug #677).
• Increase limit of files for DCT4 and check array size (bug #676).
• Allow to enter PIN code from stdin (bug #633).
• Handle HEX reply instead of UCS-2 from Motorola (bug #697).
• Avoid decoding of @ in email address (bug #697).

You can download from usual place: http://cihar.com/gammu/, Debian users will find packages in experimental soon. PS: I still have about 50 unread mails about Gammu, so please be patient.

GnuTLS will soon release 2.4.0. The 4th release candidate 2.3.15 should be available in experimental after the next dinstall run/mirror push. The upgrade and propagation to testing should be smooth, since the soname was not bumped. (The only tricky stuff was moving OpenPGP from gnutls-extra to gnutls, but since usage of gnutls-extra also requires direct linking to gnutls a soname bump could be averted.) Preliminary release notes are here. Anyway, I would appreciate if you could give 2.3.15-1 a short test. Add experimental to your sources.list and install libgnutls26=2.3.15-1 libgnutls-dev=2.3.15-1. FWIW I did not find any (new) problems when rebuilding the packages build-depending on libgnutls-dev. Now I only need to doublecheck with -release when an upload to unstable would work for them.

This teaser/trailer almost makes me wish it was winter again. Shiny!

On sid dist-upgrading requires some handholding currently: (SID)root@argenau:/# DEBIAN_FRONTEND=noninteractive apt-get --purge dist-upgrade
[...]
Calculating upgrade... Failed
The following packages have unmet dependencies:
e2fsprogs: PreDepends: libuuid1 (>= 1.34-1) but it is not going to be installed
mount: PreDepends: libuuid1 but it is not going to be installed
E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages. The reason is the recently started upgrade to perl 5.10. This makes many packages temporarily uninstallable, (including lintian) but perhaps you need to upgrade to perl-5.10 now nevertheless. apt seems to stumble over debconf-i18n,
apt-get --purge install debconf-english
lets dist-upgrade succeed.

Today was the last day the local ski resorts were open. I had a rather strange last day with a lot of freshly fallen snow, running off-piste almost the whole day. Following up on previous entries here is my balance sheet:

	2005/06	2006/07	2007/08
number of (partial) days	25	17	29
Damüls	10	10	5
Diedamskopf	15	4	24
Warth/Schröcken	0	3	0
total meters of altitude	124634	74096	219936
highscore	10247m	8321m	12108m
# of runs	309	189	503

New highscore! Still puny, compared to what other people manage on one day. I doubt it is possible to do 24500m (81000 ft) in the local resorts, you would need steeper slopes for that. ;-) Weather/climate was strange. We had an early first snow in the middle of November (about 60cm). Up to start of march no big snowfall followed, but we had (basically the same) snow continuously even down in the valley. Then there was some thawing, and finally more snow. Now that the ski resorts are closing they probably have got more snow up on the mountain than ever this season.

I have been pondering buying an entry level color laser printer, mainly for my dad (Vista) but of course at least bw-text should work with Linux, too. As expected, the situation is slightly worse than I expected.

• The printers come with toner cartridges that only contain a fracture of a real cartridge (33% to 70% depending on the model). And of course the real toner is rather expensive. HP's LJ1600 seems to be the worst offender, at EUR 160 it contains "start toner" worth EUR 151, a full original refill is EUR 318.
• Every single company still has their own home grown version of protocol for host-based printers. And it still changes.
• Most of them have free drivers, however Samsung's color laser printer's use the patent-encumbered JBIG algoritm, so we probably won't have them in Debian anytime soon.
• The only difference between Konica Minolta's 2500W and 2530DL seems to that the latter has got Ethernet in addition to USB. Amazingly however they seem to require different drivers even for USB. The former does not work at all under Linux but the latter is supported by foo2lava.
• An excellent source of information, linuxprinting.org, seems to have lost a little bit of its edge. The information there seems to be slightly dated now. Darn.

I started looking at not low end inkjets, too, The first likely candidate Canon PIXMA iP3500 or iP4500 made a quite entertaining research on the linux side. There are no free drivers and Canon's binary ones are strange. Quoting the wnpp bug:

• The packages are actively developed upstream (good!), but not maintained (what?!): each released version supports the then-current printer models only, and is not maintained after release.
• For a start, there is no canonical english web site for these packages, the URL in the package description is japanese only.

I have busied myself with Debian a somehow this weekend. A little bit of snowshoe walking in fresh fallen snows really helped me to gather some energy.

• There is a new upload of findutils in experimental which has locate split-off in a separate package. See this thread on debian-devel for the rationale.
• GnuTLS has got a new stable release pending (with a soname bump and API changes). 2.1.6, the first release candidate, is available in experimental. I did a testbuild of all direct build-dependencies last week, since only two packages failed to build I intend to keep the libgnutls-dev name for the development package.
• BTW new xpdf (3.02-1.3) and polipo (1.0.3-1) backports on backports.org.

To stop myself from endlessly wasting time browsing computer hardware I have actually bought a new computer.

• Giga-byte GA-G33M-S2H mainboard. (G33 chipset with DVI on board. Thanks Björn for the pointer!)
• Core 2 Duo E6750 (2.66 GHz, 4MB 2nd level cache)
• 2 GB RAM
• some other stuff, including an absurdely large heatsink. (Thermaltake SI-128 SE). Unless you are using a huge case you must mount this heatsink on the mainboard before installing it in the case. Otherwise you'll need to rip out all cables and start from scratch as I did. ;-)

Contrary to my initial plans I have gone for G33 instead of G965, since according to all benchmarks I could find the latter is not a lot faster (just a little bit, both are dog slow compared to almost any separate card). This allowed me to use a board with DVI on board and use one of the cheaper E6x50 CPUs (FSB 1333). I am still running mainly etch, just with a newer kernel from backports.org and Holger's Xorg backport. Everything sees to work.

Hardware is painful. I am currently pondering a hardware upgrade, hugin and qtpfsgui are painful on my machine. However it looks like my wishes are too exotic. ;-) I would like a rather quiet machine with Intel onboard graphics and digital video-out (DVI-D or DVI-I connector), 2GB RAM and some 2.something GHz Core2. Only nobody seems to build this kind of computers. "Business computers" like HP's DC5700 machines or Dell's Optiplex are not equipped with DVI out, or if they are the manufacturers try very hard to hide this fact. I then thought about assembling a system myself, but oh my - mainboard producers play the same game. No G965 board wih DVI-out. (I thought if I went through the pain of assembling I should at least be able to get GMA X3000 graphics.) I do know that I could add DVI out by means of a ADD2 card but they are hard to get and absurdely priced (EUR 25). This is just strange, everybody is buying LC displays nowadays and almost all of them have got DVI-in, even the cheapest EUR 30 graphic card features DVI-out.

I have been wondering why xpdf rendered some pdf files really really bad. (Check out this example screenshot for comparison of xpdf, evince and Adobe reader on etch.) The error message Error: PDF version 1.6 -- xpdf supports version 1.5 (continuing anyway) should have given me a hint. But I just did not think that going from 3.01 to 3.02 would make a big difference. However it does; the sid versions of evince and xpdf render as well as Adobe reader. [Updated 2007-08-25 19:05 CEST] I have uploaded a backport to http://www.backports.org/ and less than two hours later it has already gone through bpo's NEW processing and is available for download. Thanks, ftp-master.

I wonder why the standard CDS reference obligation as specified by Markit RED (Bloomberg function REDL) is often a rather obscure, totally illiquid, tiny bond. e.g why would you use a EUR 10 Mio. private placement for Societe Generale instead of one of their benchmark issues?

Found at the exhibition Play with Technology in Technisches Museum Wien:
a small working x-ray machine toy. See description.

I wish hurd was purged from unstable on ftp.*.debian.org since it would stop hurd's build status from influencing the bts versioning code. Currently the bts default view http://bugs.debian.org/findutils redirects to the listing for unstable which strangely lists bug#385896 as open although the fix was uploaded three months ago and is even in testing. The reason is simple: The hurd-i386 autobuilder still has not built the new version of findutils. For the time being I will follow Don's advice and will simply add &arch=i386 to the url.

I doubt I am the first one to wonder how a huge internet based company like amazon.(de at) can manage to be unable to send un-broken e-mails, however I just cannot resist. The regular plain text e-mails about orders (Bestellung erhalten/Bestellung versandt) are broken in a simple way, they are declared as Content-Type: text/plain; charset=ascii while actually containing Windows-1252 characters (Umlauts, Euro-character). I usually fix these in mutt with <Ctrl>-e. This week I returned an incorrectly labeled product back to Amazon and in return received a surprisingly broken mail, header and footer as UTF-8 while the core of the mail is Windows-1252 again. Looks like this in mutt

Guten Tag!
Wir grüßen Sie ganz herzlich und informieren  ine Erstattung zu der
Betreff genannten Bestellung.
Unsere Zahlungsabteilung hat Ihnen EUR 12,29 \374berwiesen.

etc...