The URL shown on the results page is very different from where I ended up (after a redirect)

Microsoft Security Warning? You idiot this is a mac. I don’t have a disk named C and I don’t have a disk named D. But Windows users are likely to fall for these things and if they do, that’s exactly what they deserve.

There was just one very minor issue advanced permalink migration plugin stopped working momentarily. However whent the perlamlinks structures were recreated it started working again. I suspect that has something to do with the db schema for the new version being  different from the old and that effecting the way the plugins’ data is stored.

I like the new dashboard but I have started to hate the new version of the visual editor already. It seems to be apallingly sluggish (2.6.3 certainly didn’t have that problem) let’s see what can be done about it. Amonghts other new features, I like the fact that threaded comments have finally been made into a feature. I looked at more about half a dozen threaded comment plugins before deciding that all of them were too much trouble. The database had what’s needed for threaded comments for quite a while, glad to see the code catch up with it.

In other matters, I have signed up with openx – an adserver + ad marketplace. Well the market place really hasn’t got off the ground yet but I am not allowed to comment more than that for the moment. I signed up for this because like most geek webmasters, I am disapointed with adsense. What’s worse there is a large descrepency between the number of page views reported by analog and webalizer and the number of impressions reported by adsense.  I have tried ot adbrite as well but it seems to be full of scammers. No not adbrite itself. Most of the products and services advertised on adbrite smells like a fishing camp.

Database error

A database query syntax error has occurred. This could be because of an illegal search query (see Searching Raditha.com), or it may indicate a bug in the software. The last attempted database query was:

INSERT INTO archive (ar_namespace,ar_title,ar_text,ar_comment,ar_user,ar_user_text,ar_timestamp,ar_minor_edit,ar_flags) SELECT cur_namespace,cur_title,cur_text,cur_comment,cur_user,cur_user_text,cur_timestamp,cur_minor_edit,0 FROM cur WHERE cur_namespace=3 AND cur_title='Alis'

from within function “Article::doDeleteArticle“. MySQL returned error “1114: The table 'archive' is full“.

There has been so much spam over the last few years that the MySQL table is full. I was tempted to clean up the database, update the mediawiki software used to power the wiki and start all over but then realized that it’s a lot of hard work for little return.

During the early days, there were a few usefull contributions to the wiki but later on they were inundated by spam. Nearly every legit contribution was accidentally deleted while cleaning up spam. Some pages would be hit even 40-50 times a minute (no wonder the database filled up). I tried combating it by using the mediawiki block list and IPtables. Entire subnets were filtered out to no avail. The spam kept increased and never showed signs of decreasing.

Trouble might have been averted if I had kept mediawiki upto date but that didn’t happen. Unlike most CMS style web applications it’s not possible to ’skip versions with mediawiki’. Well it might be possible now but it wasn’t always so. That means if you can’t keep up with all their new releases, it becomes harder and hard to do the update. I gave up years ago.

In the end I just relied on wget and a small python script that I wrote to convert all the wiki entries into PHP. There are only a few (too few really, yeah pathetically few if you would call a spade a spade) pages which means all the time and effort spent on it is just a waste.

What does this mean? has google become a social network. Will spammers now be able to vote legitimate search results out of existence? Will companies with a large number of employees be able to get the number 1 rank simply by asking all their employees to vote for their pages?

Only time will tell.

Here are the latest additions to the exclude list (in analog.cfg):

HOSTEXCLUDE 218.103.176.25
HOSTEXCLUDE 74.50.117.78
HOSTEXCLUDE 74.50.117.96
HOSTEXCLUDE 69.46.23.155
HOSTEXCLUDE 194.165.42.137
HOSTEXCLUDE 60.255.140.28
HOSTEXCLUDE 198.24.6.168

The first four spambots in this list accounted for more than 50,000 page views per week. iptables has also been updated to drop any requests from these IPS (why waste bandwidth?). Despite filtering out these requests the large differece between the number of ads served by adsense and the number of pages reported by analog remains high.

By the way, you can find out if an IP you suspect as belonging to a spammer really does belong to a one by entering it in the IP lookup table at Project Honey Pot.. Another place where you get a similar sort of service is Stop Forum Spam

Adsense targetting leaves much to be desired and click through rates are abisimal. On top of that there is a huge difference between the number of pages served (as reported by analog and other tools) and the number of banner impressions.

I joined adbrite to find that it’s a poor second best in all these aspects. To make matters worse, some of the ads definitely are sleazy. Not the kind of wares that I wanted promoted on my site. Even so I let things be for quire a while primarily out of laziness but in the end decided enough is enough. Almost all the adbrite ads have been pulled up. There is still a few left in the wiki section of this site. The wiki itself needs serious attention because it’s been spammed to death.

During the time that I was on adbrite things haven’t improved all that much with adsense.

The diagnostic tools says the daily budget has been exceeded while in fact only $0.00 has been spent! When you try to change the budget ad settings say the budget is fine. This is not the first time I have seen such contradictions at google.

The upgrade appeared to be straight forward until I tried to publish my first post with the new version. ironically it was about the upgrade process and in the end it had to be published through <a href=”http://www.wordpress.org”>wordpress</a>. Movable type seemed to choke on my tamplets.

An error occurred publishing entry ‘Upgrading Movable Type’: Publish error in template ‘Individual Entry Archive’: Error in <
mtCalendar> tag: You used an <MTCalendar month=”this”> tag without a date context set up.

WTF is a date context? the manual doesn’t say. Others too seem to have run into similiar sort of problems but none seemed to have come up with a solution – those who have found a solution were keeping mom.

In the end, I thought, Movable Type is good but not that good. I did notice that this new version seems to run a lot slower than the last and that makes it slower than wordpress too. The other good thing about wordpress is that it’s mostly PHP – none of that templating mubmo jumbo so closing the capter on  MT became a no brainer.

At this site, I have almost moved several times but never really got around to it. So for the last four and a half years, I have been using Movable Type. I have come close to moving away several times but never really got around to it. the version though is pretty old. It’s only 3.33. I have been very lazy with the upgrades. So it’s about time that I did an upgrade and I have – to version 4.21

What a big difference there is between 3.3  and 4.21 at least in appearance – unfortunately there is no substance to it. I ended up publishing this post using Wordpress!!to it.

I was wrong. dig and host both return responses very quickly. Even a completely new lookup – one that is unlikely to exists even in the ISPs cache returns in less than two seconds. Anything that’s cached is returned in less than a second and anything that’s cached locally well, that’s just 30 milliseconds – but these records still take 2-3 seconds with firefox. Could it be a problem in firefox? no – mozilla and wget also seem to have the same problem. All in all any networking applications seems to spent most of it’s time doing DNS lookups. I searched around and one suggestion was to try changing the following settings in /etc/sysctl.conf:

net.ipv4.tcp_moderate_rcvbuf=0
net.ipv4.tcp_default_win_scale=0

you can get the same effect (quicker even) by using the following format instead:
echo ‘0′ > /proc/sys/net/ipv4/tcp_window_scaling
echo ‘0′ > /proc/sys/net/ipv4/tcp_moderate_rcvbuf

That unfortunately didn’t have any effect. The same post suggested editing modprobe.conf to switch off ipv6

alias net-pf-10 off
alias ipv6 off

I reckoned I could avoid a restart (AFAIK editing the sysctl file or the modprobe.conf file requires a restart for the settings to take effect) by using rmmod ipv6. It says the module is in use. Next I try stopping ip6table (why has it started in the first place?) two of the ipv6 dependent modules disappear but I am still not allowed to remove the ipv6 module. Next i try stopping networking still no like. So I guess, I am going to have to edit modprobe.conf after all and do a reboot yikes!

Guess what: after the reboot -it’s still slow. Looks like I might have to install a caching name server locally.

Then I read that you can fix it by setting browser.offline-apps.notify to false – that didn’t work either.

The same article goes onto state that removing NetworkManager should do the trick – I didn’t do anything so drastic but simply stopped it. That seems to have done the trick.

This is most annoying but I haven’t worried too much about it because my computer hardly ever shuts down (or firefox closed). But that changed today thanks to three power failures. And on each restart firefox was in offline mode.

Some folks have been so frustrated by it that they were driven to write firefox extensions to disable offline mode. all together. Interestingly Fedora is not the only distribution with this problem, according to the forums it’s present in many others including Ubuntu but didn’t notice many references to Debian.

Iceweasel 3.0 hasn’t made it’s way into the debian stable repositories that wouldn’t really matter if it was possible to install the precompiled binary from mozilla.com unfortunately things don’t quite work out like that because I got an error with the installer (unfortunately forgot to note it down)

So if you are keen to use firefox 3 the only option is to switch to the unstable repository. The documentation advices against mixing the two but sometimes there is no choice.

Adding the unstable repostory to synaptic (which is really adding it to apt) is easier said than done. When you reload after adding the new repo, it errors out.

E:Dynamic MMap ran out of room
E:Dynamic MMap ran out of room
E:Error occured while processing scim-modules-socket (NewVersion1)
E:Problem with MergeList /var/lib/apt/lists/

http://us.debian.org_debian_debian_dist_sid_main_binary-1386-Packages

E:The package lists or status file could not be parsed or opened.

The web is full of pundits and some of them have commented that this error is because there are two many repositories. How rediculous is that? fortunately there are some really knowledgable people out there too. One of them suggested that this is because the default space allocation for the cache is too low.

I reckoned it might be something like this when I first so the error and initially thought of changing the ulimit (this is what needed to be done with k3b on Fedora), but it turns out that on debian ulimit by default is unlimited.

The solution is to add something like the following to the apt configuration files:

The number of course being the new limit. You can choose something appropriate.

All projects seem to be effected by it.

If you try to access those links you should see a 400 bad request error. But that is not happening and this is not the right time to fix it. If I try to fix it now, even the few pages that remain in the google index will be dropped off. The solution is to use a rewrite rule to persuade the google bot to stop adding pages with :443 in the URL. The rule is simple

RewriteCond %{SERVER_PORT} !^80$
RewriteRule ^(.*)         http://www.raditha.com/$1 [L,R]

Now i just have to wait patiently for the :443 entries to disappear from the index and the correct port 80 entries to reappear in it. That’s when it is time to fix the SSL issue.

The first thing to do in a situation like this is to take a bakup. Then I typed in ‘about:config’ into the browser and changed the signon.SignOnFileName to signon2.txt still no dice.

The next thing to do is to create new profile and copy the files from your old profile into the new folder. In case you have never had to create a new profile before; you need to start firefox from the command line using the -ProfileManager parameter. So did it work? not exactly but we are on the right track.

Immidiately after starting with the new profile, this error was generated:

I wouldn’t have seen that error if I had not been working in the command line. So that’s a lession for GUI addicts. Such a property does show up in ‘about:config’. I reset the property and restarted the browser. Hey presto, the passwords do show up. Then I closed the dialog and opened it again only to find the passwords have once again disappeared. I don’t recall seeing the name of Hudini on the developers list. Never mind.

Along with the entries in the passwords dialog the signon.txt file (I renamed it back from signon2.txt) has disappeared too. Good thing I bothered to take a back up. I repeated the process only to see the same result. Then I thought to try to export the passwords and try to work with that. Fortunately there is a Firefox extension just for that.

Installing and extension involes a restart so the password file had to be restored again. But yes, the passwords can be exported using this extension. And they can be imported right back into the default profile. Normal service has been restored.

Interestingly enough, when I first applied for the logmine name on sourceforge it was taken by a squatter. That was about two years ago. Upto this date they have not released anything or even made a single commit.

In recent weeks I have started getting a few cents a day from adbite. That’s right a few cents a day. It’s nothing to write home about, but you would write anyway if your adsense earnings were even lower. And the trend with adbrite seem to be upwards where as adsense was stuck in the same rut for years. Even though the amount of traffic to this site continues to grow (albeit slowly) the revenue from adsense never grew.

Some of the shortcomings of adbrite (the lousy directory for example) remain but they have attended to one of my gripes. The referral links didn’t seem to be working properly. That is I was not getting the dues from an advertiser that I referred. But after about three weeks and two emails (which were not answered) they have put that right. So the adbrite button goes back up.

Well not a lot has changed. Just the liquid layout has been done away with in favour of a fixed width design. I came to this decision not because fixed with designs are part of the hyped up web 2.0 scheme of things but because most notebooks don’t have the 4:3 aspect ratio and are wide screen.

With a liquid design a site would be harder to read on a notebook because it’s so wide but at the same time you can argue that a lot of screen space is being wasted too.

They don’t appear to have any ‘run of the network ads’. That is any ads that are displayed across the network or sites that belong into a certain category. So adbrite at the moment are showing ads from adsense. How does that happen? you can specify an alternative ad source which will be used when the ad company doesn’t have any ads of their own. In this case they don’t seem to have any ads at all.

The reason that no ads are displayed could well be because my site is not listed in the directory. If the directory listing is not automated, they should have a manual activation phase (If I remember correctly adsense had such a system).

A couple of days ago, I was browsing through the amazon web services site when I came across what they billed as a success story; Adbrite. I didn’t pay too much attention to it at the time but today, when I visited Alexa, I came across the same thing again and signed up even with out bothering to read up all the nitty gritty details.

Whether Adbrite turns out to be better or worse than Adsense only time can tell but hope springs eternal in the heart of geeks. One thing for sure, their referral program is a heaps better than the adsense program. You will get better profit share. The other advantage is that everyone knows of Adsense already so no one will follow your referal link, but adbrite is not as widely known yet, so there is a good chance that someone might. So my friend, if you are thinking of signing up for adbrite. Just click on the button on the right right

This is a screen shot of an entry on my new photoblog. One of the ads is for a product by our licensee Thin File, that is probably because I have a link to Rad Upload at the bottom of the page. What about the other ads?

Yahoo on the other hand hates Safari. It’s more accurate to say yahoo hates OS X all together. Yahoo chat for example is not available for OS X, Hey wait a second gtalk isn’t available either but ichat supports jabber, so that means you can use it with GTalk.

The new Yahoo mail sounds very exicting, it probably is exciting too but not if you are a Safari (or even a Camino) fan. So what’s the solution? set your default search engine in OS X to be firefox.

(Click to enlarge)

This screenshot is from http://www.linux-mag.com/content/view/2830/ at the time I checked it out, There was an MS ad running on it (as you can see). Would an M$ site or even a pro M$ site ever carry an add for an open source project? I don’t think so.

I clicked on the add just out of curiosity. It was blocked by firefox. Unlike IE, firefox knows what’s good for you! There is however an option to override firefox’s popup blocking, even then the ad didn’t do anything more than take me to another page that had nothing more than the ad. This time it seemed slightly more animated.