betabug... Sascha Welter

home english | home deutsch | Site Map | Sascha | Kontakt | Pro | Weblog | Wiki

Entries : Category [ digital ]
Things having to do with digital world items go in here.
[digital]  [language]  [life]  [security]  [media]  [zope]  [tourism]  [limnos]  [mac]  [athens]  [travel]  [montage]  [food]  [fire]  [zwiki]  [schnipsel]  [music]  [culture]  [shellfun]  [photography]  [hiking]  [pyramid]  [politics]  [bicycle]  [naxos]  [swim] 

17 October 2006

IP Telephony, Here I Come!

Skype, Gizmo, ...and Vivodi

Yesterday I signed up with Internet telephony services Gizmo and Skype. I had been looking into the options for a while, but then just came the time. The thing is, that I need cheaper phone calls to some European countries right now, and yesterday was the day to finally get it rolling...

My first choice was Gizmo, but that didn't work out. Why would I have preferred Gizmo? Because Skype uses a closed source, proprietary protocol. The company also belongs to ebay now, which is ultimately not sympatico. Gizmo follows the SIP standard, which means that can interact with other programs using the same, open standard. But things turned out different.

Signing up for Gizmo was easy enough. Download the program, go through the usual dance to create an account. There you go. Do you? Actually you may have to set up your firewall too. There are some ports that have to be forwarded, no problem for a bithead like me. They even link to where you can select your router and the application you want to "pass through" and you get nice step-by-step instructions. But it still didn't work. Which was obvious to me after a short while of thinking.

The thing is, we are a very progressive household here. We have our telephone service not from the government monopoly OTE service, but instead use the private phone company Vivodi for both our ADSL and telephone service. Vivodi in turn uses VoIP to give us two phone lines, and these phone lines obviously work with the SIP standard. That in turn means that the same ports that I tried to use for Gizmo are in use for our standard phone lines. Unless Gizmo (or SIP) offers an "alternate port" scheme, I don't see me using that (or another SIP service) anytime soon, which is sad.

Skype in turn uses their own, proprietary system to get through all kind of firewall and NAT routing setups. They use different ports from SIP too (since they aren't doing SIP anyway). So their service worked.

Since I want to call "normal" phone lines and not Internet users, I don't care what other people are using. Instead I need a service called "call out". Usually one makes an online payment (for example 10 Euro) and can then make outgoing phone calls to normal numbers for that amount of money. Prices are much cheaper than through the normal telephone service. Calling one hour to Germany cost me about 8.50 Euro through Vivodi, while Skype deducted about 1 Euro. The "prepaid" scheme is desirable with an Internet service like this, since you have a boundary to your spending. The voice quality was fine too, so I'll be using the service more often.

Posted by betabug at 19:56 | Comments (1) | Trackbacks (0)
18 October 2006

Μου μιλάς ακόμα;

Πως να τσεκάρουμε αν είναι ζωντανός ένας εξυπηρετητής

Ένας γνωστός server τον όποιον διαχειρίζομαι με άλλους έχει πρόβλημα και σταματάει τη λειτουργία του κάθε τόσο. Μάλλον πρόβλημα hardware, αλλά είναι δύσκολο να ξέρουμε. Αυτό που θέλω να κάνω, όμως, είναι να μαθαίνω νωρίτερα αν υπάρχει πρόβλημα, δηλαδή αν δεν απαντάει...

Ευτυχώς έχω και δικό μου "εξυπηρετητή" και μπορεί να με εξυπηρετεί για αυτό. Θα μπορούσα να ψάξω στον Ιστό για ένα πρόγραμμα που κάνει αυτή την δουλειά, μα δεν χρειάζεται. Το Unix του server μου έχει όλα τα εργαλεία που χρειάζομαι για αυτή την δουλειά, αρκεί να τα βάλω στην σειρά. Στόχος είναι να κατεβάζουμε κάθε τόσο μια σελίδα από τον web server, μόνο και μόνο για να δούμε αν παίζει, το τι περιέχει η σελίδα λογικά δεν μας αφορά. Για να μην σπαταλάμε "bandwidth" και για να μην έχει πολλή δουλειά ο serveras, ετοιμάζουμε ένα αρχείο με μόνο μια λέξη μέσα "yes". Ας το ονομάσουμε test.txt, και ας το βάλουμε να ζει στο URL

Ξεκινάμε από τα εργαλεία μας με το lynx που είναι ένας web browser στην command line. Δεν θέλουμε πολλή φασαρία, γι'αυτό του λέμε να κατεβάζει το αρχείο και να μας "δείχνει" μόνο το source:

/usr/bin/lynx -source
στην έξοδο μας βγάζει το source της "σελίδας" μας, δηλαδή την λέξη "yes". Οπότε ξέρουμε ότι δουλεύει η εντολή μας, το output, όμως, μετά δεν το θέλουμε. Ας το πετάξουμε λοιπόν:
/usr/bin/lynx -source > /dev/null

Αυτό μεταφέρει την έξοδο στο αρχείο /dev/null, το όποιο όμως δεν είναι απλό αρχείο, είναι ένα "device", συγκεκριμένα "τα σκουπίδια", δηλαδή ό,τι του ρίχνουμε το εξαφανίζει. Ό,τι θέλουμε. Μα όταν δεν παίζει, δεν θα θέλαμε να δούμε το "λάθος"; Ναι, και θα το δούμε! Αυτό γιατί το Unix έχει και ειδική έξοδο για λάθη, το STDERR. Αν κάτι δεν παίζει, θα δούμε για παράδειγμα:

Looking up                                                                                                                  
Making HTTP connection to                                         
Alert!: Unable to connect to remote host.                                       
lynx: Can't access startfile                    

Και με αυτό είμαστε έτοιμοι, μόνο που δεν θέλουμε να το τρέξουμε μόνο μια φορά. Θέλουμε να τρέχει κάθε μισή ή κάθε μια ώρα, ανάλογα με την εμπιστοσύνη που έχουμε στο μηχάνημα μας :-) Γι'αυτό χρησιμοποιούμε το cron service του Unix. Με crontab -e μπορούμε να γράφουμε στο αρχείο με τα δικά μας "crontabs" μια καινούρια εντολή με χρονική ρύθμιση. Ας την βάλουμε έτσι:

27     *     *       *       *   
/usr/bin/lynx -source > /dev/null
(κανονικά αυτό πρέπει να είναι μια μεγάλη γραμμή, την έκοψα για να μην χαλάσει την σελίδα εδώ).

Το 27 και τα αστεράκια λένε στο cron πότε να τρέχει την εντολή, "όταν τα λεπτά δείχνουν 27", "κάθε ώρα", "κάθε μέρα στον μήνα", "κάθε μήνα", "κάθε μέρα της εβδομάδας". Επειδή "πετάμε" το αποτέλεσμα της εντολής μας, συνήθως δεν συμβαίνει τίποτα. Όταν όμως υπάρχει πρόβλημα, τότε έχουμε "έξοδα λάθους"... και που το βάζει; Το cron μας στέλνει mail! Αυτό το κάνει στον server που είμαστε οπότε υποθέτουμε ότι θα το δούμε και σε περίπτωση που ο άλλος server έχει πρόβλημα. Και αυτό ήτανε, με μια γραμμή σε ένα αρχείο γίνεται όλη η δουλειά.

Posted by betabug at 10:31 | Comments (0) | Trackbacks (0)
19 October 2006

Power Day

Blow my fuse, baby!

The last 24 hours were deeply rooted in the power of electricity. First of all on Wednesday evening at home our super quality electricity system blew a fuse, which left me sitting in the dark, right as I was talking on the phone. Then I came into work on Friday morning, logged into only to find that there had been a power failure in the housing, which hadn't worked out so well with the G3 (no, no, I'm lucky there has no permanent problem). And to round things out my new bluetooth headset acted up in the afternoon, which isn't really in the electricity category by itself, but I needed the charger to get it to work again, and the charger was of course at home...

The fuse at home left my room and the kitchen in the dark. This unusual combination of rooms to be on the same circuit is already a sign of how strange our power setup is. All power lines in the apartment have only two cables, no grounding cable. Obviously the setup was for proper three-pronged cabling, but my guess is that the contractor made some extra cash by cheating them to do the job with 2/3 of cables. The fuse board is the old fashioned kind with screw in fuses. That's fine and dandy, but it usually means that either a.) you have no fuses at home when you need them or b.) the spare fuses at home are the old ones from the last failure and you forgot to buy some. Extra note to self: It's nice to own a good flashlight [pats self on back] for moments like these. But it would be even better to leave said flashlight in a well defined place, so I won't have to search for it 10 minutes in the dark.

I then went out at 23:00 to try and buy a fuse. That's not as hopeless as it sounds, as in Greece we have kiosks (or "newsstands") that are open all night, and they tend to have all kind of stuff. The first one I went to didn't have anything electrical, the second had just ran out of the size of fuses I needed. Back home we then discovered that our fuse-board has one unused fuse, so we were able to switch around to get things going.

As for the G3, one problem seems to be that the internal battery is dead. That isn't so unusual for a machine a couple years old, I'll just replace it when I go to Germany. After the power returned, apparently the machine didn't power up by itself. I assume that is because of the dead battery, since normally it's setup to autostart in after power outages. Peter from the "Internet Ulm" club went in and booted up all the servers that didn't do that on their own (thank you Peter!), which were quite a lot. With the G3 he had the extra trouble that the power switch on the outside seems to be stuck or broken, he had to open up the box to get it going.

Then there was the bluetooth headset: I wanted to do a call from the office and had to muck around with the pairing to get the headset to talk to the computer. Despite what people may tell you, bluetooth is still a hack. The only times the G4 at work has ever crashed was with bluetooth problems. This time I went through the little dance of shutting down bluetooth on the G4, starting up the headset, then bluetooth on the G4, then checking through the "Sound" system prefs that input works. It didn't do it. Deleted and recreated the pairing, again the same dance. At one point I thought I had to do it again, switched off the headset and noticed at that very moment that I had input, so I started the headset again, when suddenly everything got stuck. The Mac user interface got unresponsive (mouse still moving, but no clicks registering), though I didn't check through the network if the machine was still alive underneath. A reboot fixed that.

What was worse was that the headset had crashed too. No little blue light flashing any more. No beeps and twingles as feedback for various long and short button presses. No reaction whatsoever. Now on 17 grams of plastic and electronics they don't give you a reset button any more. There is just a button for on/off/multifunction, and volume up/down. My line of thought was that when I plug in the power supply it usually shuts down its function, so maybe it will reset then. But of course the power supply was at home. No bluetoothing and Internet phoning for me then. When I came home my theory proved right, I plugged in the good old power and the blue light came right back too.

When I still was looking for a reviving secret key press for the headset, I went through several web searches. Apparently Motorola's support consists of a page telling you that "when you can't connect the headset to a device, you have to pair them first". Interesting. Then I found out that abc news has the Motorola H500 in the list of "worst product of the year 2006/Q3". Nice to know, even though their reasoning seems based mostly on the headset not having fit their testers ear.

Posted by betabug at 23:58 | Comments (0) | Trackbacks (0)
07 November 2006

Yesterday I installed IE7

It's new, it's uh... nothing changed, CSS needs "fixing" (again)

Yesterday I had set myself the task to check our Zope application with the new Internet Explorer 7, out now relatively fresh. The IE7 update is being "pushed" onto consumers by Micro$oft it seems, but our customers come more from the corporate world, so in going through a few days of logs, I didn't spot any visits from the new Internet Exploder. Installing IE7 on the VirtualPC instance was an interesting experience though. Oh, about the CSS? It's almost fine, but it left me angry for a simple reason...

The good thing with VirtualPC is that I made a copy of the "PC Image" before starting, so I will have IE6 around for testing too, and if anything went wrong, I would have a starting point. The bad thing is of course that it's a bit slow. The "PC Image" didn't have Service Pack 2 (hey, it gets only used to look at web pages in development, why bother?), so I had to start a new personal and intimate relationship with "Windows update". I didn't even have an idea where that was. Now I know and I know it's slow (even slower on VirtualPC of course). Getting IE7 to work took me about 4 to 5 hours, doing multiple dances with "go-to-update/choose/download/restart/repeat". I was still able to do other stuff while watching the progress bars, since after all I use a Mac - and Windows is just this funny thing inside one window.

Come on, IE7 team!

After the update I went through the customer area of our site. There are a few things that need correction. So instead of having to check our CSS in Safari/Firefox/IE, we now have to check in Safari/Firefox/IE6/IE7. No, we don't bother with IE5 or IE5.5, and we do check with iCab from time to time, because of it's support syntax checker for html and css.

The problems were all in areas where we had to work around stupid Internet Explorer bugs. So how do correct that now? An internet search revealed, that the IE7 blog asks people to remove the usual IE hacks from CSS files. Great. Instead they are offering conditional comments in the html headers to use a specific stylesheet for specific versions of IE. Wonderful.

Could I ask those IE blog guys some simple questions? Where have you been 2 years ago? Or the last few years for that matter. Why didn't you speak out when people started looking for hacks to solve these problems, when they started to publish ugly stuff based on combinations of bugs in your products? I remember days and days of searching the web for yet another problem that had to be worked around, and finding yet another weird hack to copy-paste into the style sheet.

Why didn't you speak out about the hacks then?

Why didn't you comment on those blogs publishing those hacks? Why didn't you write the authors, telling them: `This is the proper way to do it, and it will work for all future'? Don't you think they would have listened? I believe they would have picked up the proper way and would have published it for everyone to use, just like they do now.

For us, the solution is simple. We work on just this one web application right now. It's a web application, so I can update exactly one html page (to be exact: one Zope PageTemplate) and include another stylesheet file or two. But I don't want to know how much time and money will be wasted on other sites to run after this mess.

Sometimes with things like these, I ask myself: I have to strive hard do my job right, why do some other people seem like they don't have to do theirs? In my job I have my own standard (not that it's too high, I won't claim that). But appart from that, customers and workmates will run after me if things are not done properly. I simply have to do my job right. I sometimes don't have the impression that everybody operates with that kind of scale. And right now I don't exactly have the impression that the IE team was in there on that level for the last few years.

Posted by betabug at 23:27 | Comments (0) | Trackbacks (0)
10 November 2006


OpenBSD auf meinem "neuen", alten G4

Letzten Sonntag bekam ich von Stefanos von der HelMUG einen G4/466 geschenkt. Nicht wirklich das heisseste und neueste Eisen, aber die G4-Architektur ist so solide und sauber gebaut, dass diese Maschinen noch länger Spass machen werden. Auf die Kiste habe ich dann erstmal OpenBSD installiert. Worauf sich jetzt der Grossteil meiner Leser (naja, zwei von den insgesamt dreien) fragen werden: "Was ist das?" Alle anderen werden sich fragen: "Wieso das, wenn der PowerMac ja auch mit einem schönen, normalen Mac OS X laufen tut?" Mal der Reihe nach: OpenBSD ist ein Betriebssystem, dass aus dem freundlichen Mac eine furchteinflössende Unix-Maschine macht. Die Bedienung ändert sich dabei von freundlich und nett hin zu verschroben und kryptisch. Warum also?

Ich beschäftige mich mit Computern mehr oder weniger seit ich 13 oder so bin. In der Zeit habe ich viel gesehen und immer ausprobiert was immer ich konnte. Basteln, probieren, dazulernen. Inzwischen verdiene ich mit Computern meine Brötchen und ich behaupte mal, dass das nur funktioniert, wenn ich immer weiter dazulerne.

OpenBSD kenne ich zwar auch schon einige Jahre [1], aber mit dem G4 versuche ich mal als Experiment einen Computer mit OpenBSD als meinen Desktop-Computer für den täglichen Einsatz einzurichten. Dabei möchte ich nicht irgendwelche andere Betriebssysteme imitieren, sondern versuche die Unix-Philosophie möglichst weit zu führen und herauszufinden wo ich dabei lande.

Eins ist schon mal klar: Selbst mit Vorkenntnissen geht es nicht so einfach wie bei Mac OS X. Nur schon dass Hardware und Software nicht mehr von derselben Firma sind verursacht Mehrarbeit und Lernaufwand. Für mich ist dabei interessant, dass jeder geschaffte Schritt voran auch wieder ein Lernerfolg ist. So kenne ich mich inzwischen schon etwas besser mit xorg.conf aus, auch wenn der grösste Teil der Menschheit lieber nicht wissen will, was das ist.

Die von mir gewählte Benutzeroberfläche (jaja, die kann man sich aussuchen und beliebig verändern) ist altmodisch bis archaisch, mit einer guten Portion "spartanisch". Ich benutze kein "Desktop Environment" (wie KDE oder Gnome), sondern einen einfachen Window Manager. Momentan ist das ion3, etwas vom spartanischsten was es gibt. Viel weiter vom Mac OS kann man sich nicht entfernen solange man irgendwas mit "Fenstern" benutzt. Die Benutzung der Maus wird auf das absolut nötige Minimum reduziert, praktisch alles ist per Tastatur machbar. Dinge wie das Verschieben von Dateien und Öffnen von Programmen werden sowieso per Terminal erledigt. Ion teilt einfach nur den Bildschirm auf und ermöglicht so dem Benutzer die benutzten Programme zu sehen. Kein "eye candy", keine hübschen Bilder, keine Gadgets und kein Schnickschnack. Maschine pur eben.

1: Aktuell ist OpenBSD 4.0, angefangen habe ich mit 2.7 oder so. Da pünktlich alle halbe Jahre eine neue Version rauskommt und die Versionsnummer um 0.1 erhöht wird, bin ich wohl um die 6 Jahre dabei.

Posted by betabug at 10:56 | Comments (2) | Trackbacks (0)
18 November 2006

Server Downtime, EU Security Measure Casualty

Or "Crash Bum Bang" if you prefer

On Friday I was preparing some things for the upgrade of to OpenBSD 4.0. Due to pilot error, I managed to crash the machine and bring it into a state where accessing it wasn't possible any more. I'm sorry if the downtime has caused any problems, especially to the witch users (the witch being the Zope/apache rewrite rule generator, which is about the only really useful service I provide). There will be another (hopefully brief) downtime on Tuesday evening, when the actual server upgrade should happen.

The problem with that upgrade is of course that I am in Greece, while the server is in Germany. So on Saturday I took the plane to Munich (well, not only for the swerver upgrade). I carried a HD with OpenBSD 4.0 installed on it with me. Given the new EU security craze^H^H^H^H^Hmeasures, I wasn't too sure if that would be trouble. We called the airline and asked. They had a bit of a problem finding someone who knew what a Hard Disk was, but in the end I was able to transport the HD safely in my hand luggage.

The EU Casualty

The EU security guidelines caused a different casualty though: Since one can not bring liquids into the cabin any more, the bottles of greek Ouzo I was bringing for presents had to go into my bag. When the luggage came out on the baggage claim, I noticed the outside of my bag being wet. I had a faint hope that it was due to the rain in Munich. Not really. One of the bottles hadn't survived. Most of my clothes had a distinct anis and alcohol smell and there were shards of glass all over the contents of the bag. I managed to get my winter jacket out of there without cutting myself. The Ouzo didn't penetrate the security barrier of the second compartment of the bag, so things weren't that bad. A run of the washing machine cured most of the other problem cases.

Posted by betabug at 13:14 | Comments (0) | Trackbacks (0)
28 November 2006

Server Failure

Main board hosed

The main board of the G3 that was running and had a total failure. The server is currently being restored on a G4. Not all data is back, not all services are running right now. Things will hopefully get better soon.

Posted by betabug at 17:03 | Comments (0) | Trackbacks (0)
29 November 2006

Site Up, Mail Not Yet works, doesn't yet

As you may have noticed, the site was up this afternoon, down again in the evening, now it came back up (or you wouldn't be reading this). I have restored both the static site and the dynamic content from backups that are a few days out-of-date. Talking about here, for the papaki site there are some more complications due to DNS issues. I could get the real, up-to-date data, but it's on a SCSI disk that -- for the moment -- doesn't work in my replacement machine. Looks like I need a SCSI terminator (and/or a clue about scsi), but the days when I had either of those are gone. The mail server configuration is on that SCSI disk too, so I will have to recreate the setup.

Posted by betabug at 02:02 | Comments (0) | Trackbacks (0)
01 December 2006

Slowly Coming Back to a Normal Server

mail is up, papaki will be next

The server is slowly getting back into shape. Mail is working again, The website is mostly running (with a few missing images in the blog). Next will be bringing the website back up. Need to set up a new VirtualHost for this. But can't be long any more. In the meantime the machine has moved to my employers office, where it awaits completion and some way of shipment to its normal home.

Posted by betabug at 17:57 | Comments (0) | Trackbacks (0)
04 December 2006

Ελληνικό iChat μέσω Terminal και IRC/bitlbee

Ποιος λέει ότι το Terminal δεν κάνει ελληνικά;

Ναι, αυτή η διαδικασία είναι παράξενη και μόνο για όσους αγαπάνε τα Unix και γενικά τα CLI. Χρησιμοποιώ IRC για chat (π.χ. το κανάλι #zope στο για τεχνική υποστήριξη στο Zope). Δεν έχω και όρεξη να έχω 12 άλλα και διαφορετικά προγράμματα ανοικτά για AIM/iChat, ICQ, MSN, Jabber, IRC, ... τα θέλω όλα σε ένα και κατά προτίμηση στο Terminal. Πιο σοκολάτα γάλακτος μπορεί να μην γίνεται, αλλά πιο πολλά chat σε ένα Terminal γίνεται...

Η συνταγή είναι το πρόγραμμα bitlbee, το όποιο κάνει την σύνδεση με τα chat servers και τα εμφανίζει σαν ένα προσωπικό IRC server. Αυτό σημαίνει ότι μπορώ να συνδέω μαζικά με ένα πρόγραμμα IRC λες και όλοι οι φίλοι μου ξέρανε από IRC.

Το bitlbee τρέχει στον server μου, που μέχρι πριν λίγο είχε OpenBSD 3.4. Έχω πάντα ανοικτό το IRC client μέσα σε ένα screen. Με αυτό τα ελληνικά είχαν πρόβλημα. Τώρα το αναβάθμισα στο OpenBSD 4.0 και τα ελληνικά παίζουν άψογα. Από ρυθμίσεις χρειάστηκε το σετάρισμα του "Character Set Encoding" στο "UTF-8", και αν θέλουμε και οι ρυθμίσεις που προτείνονται γενικά για το shell και τα ελληνικά (Βλέπετε ειδικά τα comments από κάτω με το λινκ στο

Κάτι παραπάνω ήθελε το screen. Για να παίζει το screen με utf8 ή πρέπει να το ξεκινάμε με screen -U ή μπορούμε να γυρίζουμε το μπροστινό παράθυρο σε UTF-8 με την εντολή :utf8 on on. Και τα δυο πρέπει να γίνουν κάθε φορά που ανοίγουμε το screen session μας. Το αποτέλεσμα όμως είναι "τέρμα για τα greeklish" στο iChat με τους φίλους από το HelMUG.

Posted by betabug at 12:18 | Comments (0) | Trackbacks (0)
12 December 2006

Τα Ελληνικά μετά το Update

Unicode με το καινούριο OpenBSD 4.0

Η αναβάθμιση του server μου μπορεί να κατάστρεψε το hardware (όπως διαβάζατε έκαψα την μητρική του G3), αλλά από το σύστημα και ειδικά την υποστήριξη τις ελληνικής γλώσσας μέσον Unicode UTF-8 πάμε πολύ καλύτερα. Όπως είπαμε, παίζουν καλύτερα στο Terminal. Αλλά και το Zope είχε προβληματάκια με Unicode στο παλαιό OpenBSD, τα όποια τώρα λύθηκαν. Τα RSS feeds του COREBlog τώρα παίζουν κανονικά. Αυτό σημαίνει και ότι τα δυο RSS feeds μου είναι τώρα "validating".

Ένα προβληματάκι βρήκα το πρωί με το Zwiki του papakiteliatziar: Μου έριξε exceptions στην πρόβλεψη και στην αποθήκευση τον κείμενων. Μετά από σετάρισμα του για το python, όλα πάνε καλά πάλι.

Posted by betabug at 11:00 | Comments (0) | Trackbacks (0)
24 December 2006

Rat Rated!

So who does the rating anyway?

Looking through my browser stats, I noticed several referrers from "", so "who is that?" I thought. Even with extended searches on the web I could only find it on published webstats pages. Deducing a bit, I concluded that these must be accesses from people who rate pages for a "content filtering service", the one from sonicwall. They've come to 3 times in the last 10 days or so. Following that link I came to a fancy little tool that allows to lookup any page on how it's rated by their system, which led to all kind of fun...

Apparently (this weblog) is categorized as "Web Communications" (as is Good enough, given that I talk about a wide range of topics, some of them even having to do with "the web". It could well be that all blogs are thrown more or less into that category. Because as everybody who tried to categorize a large set of data knows: Categorization systems always suck [1]. Invariably you end up at the point where you have to throw all kind of stuff into some category that doesn't really fit, just because it doesn't really fit anywhere else.

Much more interesting is how they rated, my greek wiki site. The "papaki" ("little duckling" as it's often lovingly referred to) is a mixed site about culture, food, technology, and greek stuff. So how did they rate it? "Games". Seriously, who does the rating and how serious are they about it? If you go look at you will notice a nice little drawing of a teddy bear [2]. So this is my guess to what happened: Bored and tired "rater" clicked on yet another site to review. Yawn. Uhmmm... it's kind of foreign, with all those funny characters. Better not try to read it, since "my eyes already hurt from trying to read all of the web at once". Hey, there's a pic of a teddy bear, must be one of those kid gaming sites. Click. Rated.

The result? Some poor corporate droid, who wanted to look up the original recipe for greek "saganaki" for the company barbecue likely gets a "blocked, this is a game site" page.

1: At this point the word "suck" isn't a swearword and/or sexist. It's meant as "sucks all kind of things into a category, like a vacuum cleaner sucks all kind of dust into a bag".
2: Hey, it's the vi bear, that's why it is on a technology/culture page!

Posted by betabug at 12:01 | Comments (0) | Trackbacks (0)
06 January 2007

Vivodi Upgraded our ADSL to 2048/256

But the line stutters...

This noon after a short line outage I logged into the router's web interface to notice that we have a faster ADSL line now. Vivodi (our ISP here) upgraded our ADSL line to 2048/256. For almost all their customers this was done automatically, I've had to order it, since I previously had the smallest package they offered, a meager 256/128 line. The higher speed is nice and welcome.

The downside are currently a lot of line drops. The line keeps up for somewhere between 4 and 30 minutes. Then it starts to loose packets and in the end it completely disconnects, the router reports "DSL connection down". Currently the "uptime" gets longer each time, so I hope it "fixes itself". If it doesn't, I'll have to call the support number, which unfortunately is the worst experience with this provider. "Please wait..." Let's wait and see if I'll have to.

Posted by betabug at 23:39 | Comments (1) | Trackbacks (0)
09 January 2007

ADSL Trouble

We're up... we're not... we're up... we're not!

After last weekends upgrade of our ADSL line to 2Mb, we had a lot of trouble. The line goes down every 10 minutes to 1 hour. SSH connections are stuck. Phone calls get either cut or disturbed - our phone runs on VoIP over the ADSL line. Apparently the VoIP box can recover some phone calls, but it's very ugly. When this happens, the router/modem's web interface informs us that "Your DSL connectioin is down" - who would have guessed? This evening I called Vivodi (our ISP).

After waiting for a long while on hold, I talked to a kid who said they'd temporarely put the line down to 1Mb, so it wouldn't disconnect, then they would go investigate and fix it. Good. So far they already lowered the line speed to 1Mb, like he said. But the drops of DSL connectivity keep on. For the moment it seems to be even worse. Will have to see if their investigation turns up something, or else I'll have to call them again. It's the first time ever that I've had trouble with our Vivodi line, so I hope it comes out fine.

Posted by betabug at 20:52 | Comments (4) | Trackbacks (0)
15 January 2007

Date Calculations in Shell Scripts

Keeping 3 days of backups...

Let's say we are dumping a database into a file for backup purposes. We want to keep 3 days of backups around. We could of course rotate the filenames (similar to what newsyslog does: basename.gz, basename.0.gz, basename.1.gz). But we want to backup the directory to a remote host using rsync - and renaming the files would mean to transmit all files with each remote backup. A better naming scheme is to use names with dates, like basename_20070115.gz. That way names don't change and rsync doesn't retransmit yesterday's file. The problem then is that we have to delete backup files older than 3 days. We need a way to calculate dates in a shell script...

The secret sauce in the (partly pseudo) code is that we use the "unix epoch" - seconds since 1970-1-1 - for our calculations. Let's go through the code:

cd /backup/directory

budate=`date '+%Y%m%d'`
                        # today, normal date format, 
                        # backticks to get the at the string
bufilename=basename_`date '+%Y%m%d'`.gz
                        # same thing rolled into filename

# now delete the file from 3 days back, 
# so we have always 3 days of backups
let deletedate=`date '+%s'`-259200
                        # todays date in seconds - 3 days in seconds
                        # key points are using `let' for sh calculations
                        # and using `date' with the %s format string
delfile=basename_`date -r $deletedate '+%Y%m%d'`.gz
rm $delfile

dosomebackupscript -o $bufilename
                        # create the new file

This works by calculating the filename for the file 3 days ago, and deleting that file. After that is done, a different backup script creates the latest file with the same naming scheme. The script runs in a cron job, when it doesn't run on one day, then a stale backup copy is left over. We're not attempting to solve that here.

Posted by betabug at 23:26 | Comments (13) | Trackbacks (0)
23 January 2007

Scheduled Downtime

Server will move again

We will have some scheduled downtime at and, because I will attempt to move the server to his proper location again. The server will be shut down on Wednesday 2007-01-24 around 17:00 (GMT+2) and should arrive at his proper home and go online again on Thursday 2007-01-25 around 21:00 (GMT+2). There will be no web or mail service (or any other of the stuff the machine does) during that time. Backups are taken care of. Thank you for your patience, please knock on wood and cross your fingers.

Posted by betabug at 18:11 | Comments (0) | Trackbacks (0)
Prev  1   2   3   4   [5]   6   7   8   9   10   11   12   13   14   15   Next