So far, it has been a fairly happy hacking day. I wanted to see how far I could go to make a Linux desktop accessible for the blind. Platform of choice is Breezy because:

• it uses gnome, which is the most accessible stuff at the moment;
• it's the distribution with the least amount of raw edges at the moment;
• its gnome accessibility stuff is more up to date than the stuff in sid (I got in touch with Mario Lang and I'll see if I can help)

Uwe, you are not the first to request that I add a comment feature to my blog. Its consideration is on my to-do list, but I am somewhat against it already. Why? Because I do not want to deal with yet another point of attack for spam, and I am really not fond of web interfaces; there are probably nice ways to tell pyBlosxom to integrate comments well with the filesystem, allowing me to e.g. reply via SSH rather than a browser (or better yet: email), but as it stands, all I know are comment forms on the blogs' web interfaces, which require a browser and cookies enabled to be able to add one's opinions. As blog owner, I have to continuously check my blog for new comments with a browser, and leave answers to comments immediately as I am checking them, but tracking forums with browsers is a royal PITA. Thus, an RSS feed would be the right way to do it. However, RSS only alerts me of new comments; I then still have to fire up a browser (or a shell) to leave an answer. Anyway, I am already drowning in emails, I don't need another medium requiring attention and feedback. There's also another reason against comments on my blog: it's my blog. Apart from legal obligations in the EU, I have not fully grasped the pros and cons of giving other people a space to express themselves. If an entry of mine solicits feedback, I include an email address, and I summarise what I learn from the numerous responses I receive in a second blog entry. I will look into ways to integrate my blog with my email traffic, so that comments come in as emails, and replying to them publishes my response on the blog. Anything else I would not want, I think.

Switzerland is bloody expensive, you surely don't want to be shopping here. For instance, I wanted to get myself the Lonely Planet Guide for Laos and Myanmar, both of which have had new editions released in 2005. Their price is a little more than 20 each, including the German VAT of 7%. If you buy the book in Switzerland, it's between 40 and 50 CHF (incl. 7.6% VAT), which is almost 10 more than the German price. Despite the fact that I cannot find the 2005 editions neither in Swiss stores nor on amazon.de, I'd be looking at a slight problem. Either, I can pay the higher price in Switzerland (in addition to shipping), or I can order with amazon.de (or the like) in Germany, who will subtract the VAT and ship it to Switzerland for free. Sounds great, huh? Well, it isn't really, because in Switzerland, receiving packages might mean you have to pay VAT, and not only that, they also make it unnecessarily complicated: If the goods you receive are valued less than 75 CHF, you do not have to pay VAT. However, if you do have to pay VAT, the VAT and a nominal service charge of 10 CHF count towards those 75 CHF, which means you effectively have to stay below (75 - 10 - VAT) CHF, which is 60 CHF or thereabout. Translate that to Euros, we're talking about 38 . Thus, if you want free shipping and no VAT, you have to place orders between 20 and 38 -- 20 is the minimum price to get free delivery if your order is not entirely made of books, in which case delivery would be free even below 20 . You do not want to place orders higher than 38 ever, because VAT in Switzerland is higher than the VAT for books in Germany, and the additional 10 CHF charge is not going to make up for it. So the only way for me to obtain the two books is either through the overpriced Swiss bookstores, or via two separate orders to amazon.de so as to not exceed the 38 limit. The maths gets more fun when you get away from books. For normal goods, VAT in Switzerland is still 7.6%, but 16% (and maybe soon 19%) in Germany. Thus, you will want to stay below 38 , unless the total value of the goods exceeds an amount where the cheaper 7.6% VAT + 10CHF service fee is lower than 16% VAT. This drives the limit up to 121 CHF, or about 77 . Now, if you think this is complicated, stay put. I am looking to buy myself the DrayTek Vigor 2900VG router, for which is around 265 in Germany including the 16% VAT, and 421 CHF including the 7.6% VAT in Switzerland. These prices square, but they're not including shipping, and if I would order the German product for delivery to Switzerland (which they are not willing to do, but anyway), I'd have to add the 10 CHF service charge and the 7.6% VAT. I also would not be able to claim the 16% VAT paid in Germany with EU customs because I am not a full resident of Switzerland. Man, no wonder taxes have had a long history of violence. Now I also received a 35% rebate offer from the Swiss distributor, driving the price down to 274 CHF, which is clearly a hell of a lot cheaper than the German product. So what am I waiting for? Well, Germany being an EU state and all, the German shop has to give you a 30 day return warranty and a 2 year product warranty. In Switzerland, you get no return warranty and only 1 year of product warranty. My dilemma thus becomes: are 30 days return warranty and an addition year of product warranty worth the difference of roughly 150 CHF? Seems like the answer must be "no". But I know that if I order it in Switzerland, it will break after 366 days, by the laws of Murphy. I have now asked the German distributor to match the 35% rebate. If they will, I'll order the device to my parents house in Germany. If they don't, I'll take it up with Mr. Murphy. PS: If you think the DrayTek Vigor 2900VG is not a good choice, speak up now!

It's about time I encrypt my home directory. So far, I've found two tools for the job: dmcrypt and encfs. dmcrypt links in between the VFS layer and the data store. For an encrypted partition or image, it creates a device under /dev/mapper, which you then mount. Think cryptoloop. Encryption covers the whole scale of symmetric encryption, but I didn't find anything related to public key crypto. I am also not entirely clear yet on why a damaged block on the storage medium doesn't screw the encryption block (partially) stored on it, as well as all following ones, given a chaining mode like CBC for instance. However, as this thread shows, I am wrong and technology is right (and works), and I still just don't get it. It feels like another Monty Hall problem to me (which I also never really understood to the point of being able to convince someone else thinking 50/50 is the answer of the alleged 2/3). So then there's encfs, a FUSE-filesystem that seems (I have not tried it) to create a virtual directory for an existing one, transparently encrypting filenames and contents as they are written to the existing directory (via the virtual one), and decrypting them when requested through the virtual one. So it's a 1-to-1 mapping, which has its pro's and con's. encfs seems equally strong on the symmetric crypto side, but seems to lack public key support. For each, there's a proper PAM module: libpam-mount can mount a cryptoloop filesystem upon login (though the README says it cannot do so for OpenSSH logins. If that's right, well, one candidate less...). On the other hand, there's libpam-encfs, which creates the virtual directory on login (and removes it on logout). I can well imagine how any of the PAM modules would do their job just fine on a machine where xdm is the primary/only login method. However, what about SSH? What happens if I log in through xdm, then SSH, then quit X? How do you solve the problem? I am curious and will blog about the things I hear... Update: After rereading my thread, I am understanding, I think. We're talking encrypted block devices (or loop devices). It's not difficult at this stages to tie IV chains in with storage blocks. I am still curious why my 8M file was still all zeroes after changing the cryptotext. Update2: Sesse points out that CBC uses the ciphertext of the previous block as an IV, not the plaintext. So if some block gets corrupted, that messes up the next one's plaintext, but not its ciphertext, so the one after that is OK. Thanks.

I've always wanted to tag my image collection, but it's big and a pain to do by hand. Today I found out that digikam has got to a point in which it can be used as a nice interface to categorise images, so I now have something cool for the tagging work. Now, I'd like to play with the tags using my debtags toolchain. Digikam stores data in a SQLite3 database, so it's easy to convert it into the text format used by tagcoll. Here's the script to do it:

#!/usr/bin/ruby
# gettags - Extract a tagged collection out of the digikam database
#
# Copyright (C) 2006  Enrico Zini <enrico@debian.org>
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
# Usage: gettags [path/to/digikam3.db]
# If the database file is not given, use the one in the current directory
require 'sqlite3'
db = SQLite3::Database.new( ARGV[0]   'digikam3.db' )
# Build a table of id => complete tag name
tagByID =  
db.execute( "select id, pid, name from Tags" ) do  row 
        if row[1] != 0 and tagByID[row[1]] then
                tagByID[row[0]] = tagByID[row[1]]+'::'+row[2]
        else
                tagByID[row[0]] = row[2];
        end
end
# Why does it work with instant queries but not with precompiled queries?
#db.execute("select tagid from ImageTags where imageid = ?", 36887)    r  print r[0], "\n"  
#gettags = db.prepare("select tagid from ImageTags where imageid = ?")
#gettags.execute(36887)    r  print r[0], "\n"  
lastname = nil
lastdir = nil
ids = []
db.execute( "select i.name, t.tagid, a.url from Images i, ImageTags t, Albums a where i.id = t.imageid and i.dirid = a.id" ) do  row 
        if row[0] != lastname then
                if lastname != nil then
                        print lastdir, "/", lastname, ": ",
                              ids.collect    id  tagByID[id]  .join(', '), "\n"
                end
                lastname = row[0]
                lastdir = row[2]
                ids = []
        end
        ids <<= row[1]
end
if ! ids.empty? then
        print lastdir, "/", lastname, ": ", ids.collect    id  tagByID[id]  .join(', '), "\n"
end
exit 0

I didn't understand why if I perform a precompiled query I get a row object that cannot be indexed, while if I perform an instant query then everything is fine (see the comment in the script). I'm still on the learning side of Ruby, so I welcome people tellig me of better ways to write this script, and I'll be glad to update this entries with what I receive. Done the little script, now I can plug my images collection into tagcoll and have a bit of fun:

./gettags   tagcoll related /2005/12-15-03-Taiwan-Newcamera/dsci0051.jpg
./gettags   tagcoll hierarchy
./gettags   tagcoll implications
./gettags   tagcoll findspecials

Cute! And this is another little step to somehow connect pieces of the debtags toolchain to the web, which is something I'd like to explore both for the web interface of the central database, my blog and my picture gallery.