Ending the Experiment

The last post already hinted that I was unsure on how to proceed with this blog. Well, I made my decision now.

Since I don’t have enough time and not enough to say to continue this in any meaningful manner, I’ll stop posting to the blog. It will share the fate of countless other blogs that were started eagerly and then discarded.

I made this post not so much for a hypothetical reader, but more for myself – to place a period at the end of the sentence that is Drangon’s Digital Den. Maybe I learned something good during the experience, maybe not. Maybe my English improved, maybe not. But at least I came away with some new knowledge and a little bit of experience.

Farewell.

Transition to WordPress.com

I grew increasingly unhappy with Blogger, so I decided to move the Digital Den over to wordpress.com.

This has been a mixed blessing so far, because I had to hand-edit most of my previous posts to fit here. At least the HTML is now cleaner than what was previously produced by the Blogger WYSIWYG editor. Apropos editor – why the heck is the WordPress one missing justified alignment?! I find justified text just so much more readable and better looking – well, unless the spacing is screwed up…

There are a lot of features available Blogger lacks: Categories, the neat post preview function, included visitor stats, and a Blogroll. On the downside, I can’t freely edit the template. :-/

About that Blogroll-Thingy: I don’t actually know what a Blogroll is supposed to be. I know that you link to other blogs and stuff, but then again, it defaults to wordpress.com and wordpress.org. I guess I’ll just stick in a few of the blogs I read. And maybe next time actually write a post that has some interesting content…

I guess the problem is, that I just can’t decide whether to keep this thing just for fun or actually devote time to it, so it’s not so half-assed anymore.

Tags: WordPress, Blogroll, Dumb Content

Picasa coming to Linux?

During my daily websurfing I came across an article on DesktopLinux.com, where rumours are spread that Google software is soon to run on Linux. Apparently Google is in contact with CodeWeavers to make this happen.

CodeWeavers is known for its modified Wine-version called CrossOver Office, which – you guessed it – runs Microsoft Office under Linux. Apparently Google would like CW to do the same with its own software, starting with Picasa. Neither of the two companies did comment on the issue though; the article invokes “Sources close to the project”.

Apart from this uncertainity, the thing that bothered me most was the part about other Google applications following suit, if Picasa was successful.

Sure, Picasa is nice, but there are already good alternatives for the Linux Desktop (e. g. iOta, digiKam). Google Earth on the other side, is very tricky to set up using Wine only.

I can understand that they want to try this with a comparatively simple program like Picasa first, but honestly, Picasa is not exactly the Google program the Linux Desktop needs – and hence, it might not exactly be successfull. I hope that’s not going to be the point that denies us other Google Applications…

Tags: Google, Picasa, Linux, Wine

Cool Text

After previously posting about an ASCII Generator, I thought I’d also share the link to Cool Text, a website that lets you render a custom Text Logo.

The FAQ says, that the image can be used without restriction, so I set out to create a Logo for this blog.The image was not very optimized size-wise, so I recompressed it with OptiPNG. That reduced the filesize by more than a third – Makes people on slow connections happy. ;-)

Time to put it up.

Aww, damn. Call me stupid, but I can’t insert the image into the template; the CSS keeps messing up everything… :-(

Well, I can at least post it below, so that you get an idea of what Cool Text-Logos look like:

Dubbing

Through the wonder of DVDs I’ve been watching more and more movies in their original English language. In Germany almost all video content is dubbed (i.e. dialogue is replaced with a German translation); it actually is kinda hard to get original language content unless you buy the DVD, since untranslated movies are only shown about once a week, late at night. And only in cinemas, not on TV.

It’s great to have a movie in your native language, but now that I started paying attention, I noticed that the result is often imperfect – sentences not finished properly, awkward language, timing issues (German words are longer on average than English ones), or ocasionally outright translation errors. I was really startled when I started paying attention, because all but the most gross issues kinda are corrected subconsciously.

The thing that surprised me most was that accents generally are ignored. In retrospect it seems logical, since there is no German equivalent to, say, Cockney. Generally, the only accents kept – or rather replaced with German equivalents – are the clichéd ones: French, Russian, Chinese, Texan, etc. Some others do keep their vocabulary (say, the scottish ‘aye’), but that’s mostly it.
It’s confusing when the dialogue directly touches the accent – sometimes it’s smoothed over (e.g. someone – falsely – claiming that the character in question falls into mumbling sometimes, instead of pointing out that he’s Canadian), but other times no explanation is given.

Another topic is humour. That’s hard to get right, and it is the area where most of the mistakes are made. That’s because humour often depends on word ambiguity, alliterations/rhymes or things every native speaker would know. It is remarkable how successfull dubbing is in this respect nonetheless. But when it fails, you notice.

Still, there are some movies where IMHO the localized version improved upon the original: Pirates of the Carribean does seem to be a lot funnier to me in German. Mainly because of added emphasis here or there – the characters are more over-the-top without it seeming silly. On the other hand, maybe on average there isn’t as much emphasis in English and I just didn’t notice yet…

Tags: Dubbing, Movies, Television

How does Google collect and rank results?

I just found the article “How does Google collect and rank results?” linked on digg, and I thought the author did an excellent job of explaining the inner workings of a search engine in simple terms.

I had a post explaining the same things in draft state, but decided to scrap it – No need to waste time on doing something, someone else has already done better than I could.

Tags: Google, Search Engines, Indexing

IMAP

Since I’m now studying at the University of Kaiserslautern, I’m using multiple computers: The one in my domicile, the terminals of the Computer Science departement, my parents’ computer when I’m visiting home and in the future maybe a Laptop.

The problem that arose, was how to keep my stuff synced between all these. Sharing simple files is not much of a problem since the university machines are always on, and I can access them remotely via SSH.
Browser Bookmarks were a little trickier. I thought about uploading them to the web, but doing that manually would be tedious. After looking around for a bit, I found a nice Firefox Extension, Bookmarks Synchronizer, which does it for me automatically.

What remains is my email. Sure, I could use the Web Interface, but… I don’t like it.
The first thing I thought of was to carry my email with me on a USB-Stick, but that proved to be even more annoying. Then I considered paying my mailprovider GMX for IMAP access. IMAP allows you to maintain all email on the server of the provider, so I would not have to manually sync. I applied for the 30-day trial.

Frankly, it sucks. They wrote the server themselves, and it doesn’t allow for nested folders – no more subfolders for my mailinglists. Dang. They also advertised their Organizer and Adressbook features for the Pro Mail option, but that’s web based as well. Web based only. You can’t automatically sync it to, say, Kontact. Well, if IMAP worked properly, KMail could store Organizer and Addressbook data in the mailbox… but no. Conclusion: Not really worth the money.

Time to see who else offers IMAP. Aha! Web.de. Free of charge, even. Signed up. Signed out. 12 MB storage space. HELLO?! That kind of defeats the purpose, doesn’t it? Why have IMAP at all if the user cannot store all mails on the server? Maybe that’s their way to prevent people from using too many server ressources. Maybe I oughtta pay them for more storage…

Guess I’ll just suffer through not always having my mails with me. It’s not like I’m addicted or something. Really. I can stop anytime I want. Believe me. Please…?

Tags: Email, IMAP, sync, rant

Google Analytics #2

Considering all the fun I had with figuring out what Google Analytics is, I decided to just go and take a look at it.
After signing up, I had to wait for twenty-four hours until Analytics recognized its code, and twelve hours more until the first data came in. Apparently Google was ill-prepared for the initial onslaught of signups since everything was going slooow. A day later, I could not log in at all. Well, after the initial load, everything runs smoothly now – probably because new signups are currently disabled.

The Interface is nice, although a little less concise than what I’m used to from Google. It’s made with AJAX, and thus very responsive and fast – it also provides the user with nice charts and all kinds of eye candy. What made Google Analytics most appealing to me was that it allegedly allowed monitoring a website (or – as in this case – a blog) as if one did host it oneself and had access to all the logs.

Well, that’s a rumour. It does provide a lot of data, but nothing that other tracking services like Tracksy couldn’t provide (Well, apart from the AdSense stuff…). In fact, there’s one thing I’m really missing: you can’t see referer URLs. You can only see referer domains.
For example, the last post generated a trackback on Writer’s Edge. Analytics says that people came here via that domain, however it does not show the location of the link. Maybe that’s because the service is a descendant of Urchin – which is designed for corporate customers with thousands of visitors a day. Guess they didn’t think anyone would bother with single links… Oh well.

I’ll experiment with it some more, but I guess I’ll eventually go back to Tracksy.

Tags: Google, Google Analytics, Web Stats

Google Analytics

I just edited the previous post, and while submitting, the browser was showing “Waiting for google-analytics.com“. Huh? Never heard of that.
Is Google analyzing the post or something? Some Data Mining project, maybe?

I decided to enter the URL directly into Firefox, but all I got was a Google-flavoured 502-Error page (Internal Server Error). Sticking www in front of the URL gives a 404 – File not found.

I tried doing a WHOIS-lookup, but I’m not particularly experienced there. The query did not yield any results. Maybe that’s my fault, maybe the domain is too new or something – it isn’t listed in searchenginewatch.com’s list of google domains either, which was last updated in October.

Fine. Maybe someone else has already uncovered something?
A quick search revealed that I’m not the only one having noticed this – a few other people already spotted it and posted, but noone seems to know much more than I already figured out by now – Georganna Hancock at the Writer’s Edge discovered that the domain was registered in July though – so much for my too new theory…

There are some french posts, however I don’t speak a word of french. They mention a different url: https://ssl.google-analyze.com/urchin.js. A JavaScript file? It seems that this is part of Urchin, a web-analysis company Google bought. The .js-file looks at the Referrer to learn where the user came from. An array of several search engines is listed. Then a lot of complicated-looking code follows. There are two links to a .gif-image (one SSL link, one normal one), but the image is transparent – a web bug?

Well, now I’m outta clues. I still wonder what exactly they are doing with the Blogger posts – the Referrer isn’t very valuable here, since it points to blogger.com unless I’ve missed something substantial…

Edit: There actually was one clue left; I examined the HTTP-Headers while posting this post. The browser requested the mentioned image, however it had a trail of options you would normally expect from a php page… among them my screen resolution, color depth, the current URL and a lot of other data I can’t make sense of.

Edit (2005-11-14): Google Analytics is apparently available to the public now.

Tags: Google, Google Analytics, Web Stats

Data safety

The advice is omnipresent: “Thou shalt backup your files!”

Well, I didn’t bother until recently, because I never really had a nasty disk failure or similar cataclysm destroying my data. When it actually happened, I started to think about a decent backup strategy.

The Files

What actually needed to be backed up?

1. Documents, Sourcecode, Savegames, etc.
2. Mails, Calendar, etc.
3. Application Configuration (e. g. Firefox preferences)
4. System Configuration (e. g. Timezone, Keyboard layout)

Because I’m running Linux, most of that already resides in my home-directory (1, 2, 3), so I don’t have to go hunting for it all over the harddisk. The remainder luckily also is grouped together – all system configuration resides in a directory called etc. I wonder where that name came from…

Okay. All my important files are accounted for, but there are also some files among them which are not very desirable in a backup:

• Caches: They only waste space, since they are automatically rebuilt when they are not available. The best example is probably the Browser Cache, which takes up 25 MiB on my system.
• The Trash: Well, I deleted the stuff, why would I want to back it up? As for accidental deletion, that is unlikely enough to coincide with a system breakdown that I don’t bother – most of my important files arrive by mail anyway, and are stored on the server for easy re-retrieval.
• Spam: Of course my mail is backed up, but there is no need to keep the snake-oil sales pitches.
• Temporary files: Many programs litter the harddisk with temporary files. And because they are temporary, I don’t want them in a permanent Backup. Easy as that. :-)

The Program

Now that I had established what I wanted, and what I didn’t want, I had to look for a program that could make a nice package from my files. The first one I considered was good ole GNU tar – it did most of what I wanted, but all in all, it was a little cumbersome for backup purposes (despite that being exactly its intended usage… after all tar is short for Tape ARchiver).

Enter dar, the Disk ARchiver. I found it on Freshmeat while looking for something more versatile than tar. In addition to full and differential backups, it offers a host of other features: slicing the backup into fixed size chunks, compression (Yeah, yeah. Usually no good for backups, but there are some safeguards, and IMO the risks are bearable), PARchive support, direct access to files (no need to untar the whole archive), strong encryption support (Blowfish) and others.

My Setup

I had a spare partition of 500 or so MebiBytes on a different disk than the one my home- and etc-directories reside on, so I set that up – in case of disk failure it would be safe, but in case of a meteorite impact, it would not. :P

I made a full backup there, totalling about 150MB. I also burnt that to a CD-R, together with DAR and the DAR documentation. Then I modified the scripts from the DAR differential backup mini-howto to suit my needs and put them into the crontab. From now on, DAR would make a differential backup on my spare partition every day. Since the backup is differential, it would weigh in at at most a few MB. Every week or so I could then burn the differential backups to the CD-R the full backup resides on. Once the CD is filled, I will make another full backup for a different CD-R and again fill it up with differential backups.

Another approach would be to let DAR split everything into 10MB chunks (or slightly less, accounting for Base64-encoding-growth) and directly mail the files to, say, MyAccount+Backup@gmail.com.

Edit: I did not take into account that every session on a multisession CD takes 13 MiB by itself, so it is pretty inefficient to burn the differential backups directly. On the one hand, CD-Rs are cheap, but on the other hand I hate inefficiencies such as this one, and so decided to cache the diffs on my USB Stick for now, and only burn them every month.

Tags: dar, Backup, Data safety