Recently, Wordpress 2.8 was released. This, & other events, prompted me to take a loot at vodex.net, as it's been largely abandoned for some time... imagine my delight to discover the Wordpress install wasn't even working anymore for some reason; the Livejournal post import was broken, I couldn't log in.. it was a mess. The old version, kept behind to enable custom modifications to integrate with Livejournal / the main site / various PHP applications, didn't help. (Security hole? I'd had the forethought to use a hidden path for Wordpress itself)

I use a lot more social media now, & have largely abandoned Livejournal, along with a lot of other people. (They could have been facebook, you know; shame they don't expose enough data to Google to let me find that very LJ post by someone on that topic). Anyway, I'm working on converting the entire site to Wordpress's considerably improved feature set, including various plugins for FriendFeed, Google Reader, etc. (I especially like the All-New Livejournal Importer for Wordpress, which is a much better job than WP's original LJ import script, or my own LJ import script ... except I don't want to duplicate posts on the site, I want to link to them. Might use it for a private LJ backup tool, though).

So, watch this space. Or rather, that space...

I've added a few things in MailMaker, the spoof Daily Mail generator:

• option to create a Daily Express. It's not a perfect match; when I have some more time, I'll refactor so you can use the Express font/layout. And do The Sun as well.



• made meaningful 'alt text' to the images



• created an RSS feed of latest covers
  



• and - bit warily - added some extra images, including, of course, Madeleine McCann. Not sure it might turn out a good idea (for hopefully obvious reasons of taste, hence the delay), but like Diana, what Mail/Express cover is complete without some?
  

Yes, I seem to have fully caught the twitter bug.

I've made use of Paul Stamatiou's RSS to Twitter script to automatically pimp for technology's sake announce new Livejournal posts on twitter. It'll have to run as a cron job for now, unless I can find a way for Livejournal to ping whenever a post is made.

Woohoo! I've just gone to Technorati, found out that old addresses for blogs and stuff were out of date, and claimed my canonical LJ address - using OpenID! It took longer to type this than to do it. And first real use of OpenID, too.

heh, it said it'd been last updated 573 days ago. That's very not right.

Sorry, spod post.

One of my favourite cartoons is Ruben Bolling's Tom the Dancing Bug.
Now and then, he runs a 'Super-Fun-Pak Comix',
an entire page of parodies of cartoon strips (and a masterclass in succint humour).

I've made a page on vodex.net that shows individual Fun-Pak-Comix at random, as well as including a title banner.

It's at
http://vodex.net/superfunpakcomix.

I've written a script to remove duplicate iCal events after my Google Calendar got messed up - http://vodex.net/ical_prune/ .

I'm migrating my website from PHP to Ruby on Rails. Last night, moved the PHP source code form my old hosts to RailsPlayground. Whilst good value (they do Rails, and have about 10 times the bandwidth/storage/features as my old host for less money) their site/support forum seems to have very dodgy availability. So was largely on my own for figuring out how to host PHP as I'm not that experienced in remote barebones hosting.

Turns out Rails routing applies across the whole site, not just in the railsapp directory, so PHP has to end in .php, and it has to go into the public folder, which in hindsight makes sense (most PHP apps live in the public area of a site).

Here's the roadmap:

• I have until the nameservers update to arrange for pretty URLs like vodex.net/goalpost to point to dirty ones like vodex.net/goalpost/index.php - don't want to mess with visitors/search engines. This is a temporary kludge, so I'll probably just alter Apache conf. a bit. I only have one Rails app live at present (Universe on Rails) and it's low profile so it's not important as I'll...
• Install a Ruby CMS/blog like Typo, and use it instead of the custom PHP setup I used to have.
• Develop something like my current Wordpress Livejournal importer for Typo, using the LiveJournal Ruby module, thus standing on shoulders of giants (LJ's XML-RPC API seems a little counter-intuitive).
• Move on from there.

In related news, work upon a mockup document management system at the University proceeds apace, partially due to productivity gains from said Ruby on Rails. It's true - Rails encapsulates best practice out of the box, leading to much better overall progress despite the learning curve. It's all about the framework - Stuff that would have taken me months in PHP (even with PEAR) has taken just days in Rails. It's also giving me experience in tools I didn't make use of when developing GoalPost - Subversion version control, test-driven development, and similar agile methodologies.

At present I'm changing the view from an initial makeshift scaffold layout involving fieldsets to a much solider layout. Changing layout in rails is a snap, much easier than, say, in PHP, where you ended up with HTML-producing code inside logic classes *cough*.  Good workflow here – determine a view/layout (separate from the concerns of the logic creating it), then produce the current design in Dreamweaver or Nvu, say, and then componentize it into layouts. Much cleaner.

The Daily Mail Maker got a traffic spike of 180 visits/700 pageviews on Thursday 16th March, getting on for five times my usual site traffic. All from a single unreplied-to link from b3ta.com (here), it seems. I guess they have a lot of office worker traffic :)

For future reference to self, to find individual referrers in Google Analytics, find Referring Source under Marketing Optimization, and then, (to get individual referrers, which up to now I couldn't find) Cross-Segment Performance with Content. A little annoying I couldn't find that, as I knew it must capture the data...

Clare has seen the last of the nurses - her leg is fresh scar tissue almost to the knee and is still nauseous with trouble standing, moving etc. but it's officially the "months whilst it heals" stage now. As a benchmark of her progress, apparently she will be able to do a job if she has her leg up all day and doesn't stress herself. In a few weeks.

I am clearing out the bedroom a bit each night as we are having the ceiling repaired on Friday after the storms a couple of weeks back. I am very scared of re-infecting Clare by her catching her leg on the edge of a dusty cardboard box etc.

There is talk of my contract being extended/re-negotiated beyond the end of my contract her at the University to do a (vastly improved) unit database. I'd really like to do it. But the wheels of academia can move slowly, and I'm taking no chances, hence I've updated my CV and will be informing agencies/local companies. This is something I should have done over a month ago, but of course The Cellulitis Monster happened instead.

And lastly, I stupidly had a salesman come round from Mosaic ERP, as we need work done on the front of the house. He stayed to 10pm despite Clare being in some pain, and went through his sales patter despite us making it clear he would be wasting his time if he quoted more than a grand or two. He then quoted £10,000 for chiselling out some crumbling bricks and putting Tex-Cote gel over it, and despite us saying we couldn't afford it, refused to leave (despite us asking to) until we signed his timesheet. (A week from now I'm finally starting to accept he didn't make us sign a contract whilst we were sleeping standing up.) I thought he was actually going to assault me by the end, and I'm not joking. Surprisingly, our own enquiries turned up that most of his claims weren't accurate. Despite assurances about no further phone calls we've had three so far, about how they would knock the price down by putting adverts up and the workmen will only be standing around otherwise - the way he immediately hung up when I said "no" was interesting. And I wax lyrical about it here because web searching for Mosiac ERP shows up little results, and, well, how million people use the web for research now?

Update on Clare. She now has a Viscopaste PB7 Zinc bandage, which has helped with healing incredibly. Had two nurses round to check blood pressure/circulation around her leg, with a nifty ultrasound gadget that put her heartbeat on speakerphone. Leg now resembles corned beef, but instead of congealed lard around it, we have dead skin to be slowly washed off. Leg's hardly leaking anymore, and is now under a compression bandage.

I've also given the cellulitis pictures a permanent home at vodex.net.

Already after a couple of days of relative rest, I feel considerably better - although I still went and bought these exotic-looking fruits in Asda. Next to the kumquats and lychees; the size of plums; yet dusty-orange and with hard skin. The tantalising label read simply 'Stone Fruit'. What would they be like??? Got home and - "you idiot, those are apricots!"

Discovered that vodex.net's web counters hadn't been working since December - bbclone doesn't work if you put the include in a function, whch running a CMS which I of course do, and it errors silently by default... Fixed now, but it now records every 404. Ho hum.

Ahh, little things, little things...
The domain vodex.net is now active is now my website's home, replacing the .me.uk. If only vodex.com was for sale...

Clare is slowly making progress. She now has a dressing made of some PB7 zinc wet bandage that doesn't hurt as much as she expected. Still in a lot of pain and has trouble standing or walking, continually groggy and light-headed from all the medication she's on, and won't be fit for much for the next month or so. I'm still having to do al the cooking, shopping, cleaning, etc. She still doesn't have a job, of course, though it's probably good that she didn't get one before this happened or she'd probably have been sacked

She gets home visits from nurses, which is variable - the dressing on Wednesday was so painful I had to take it of and re-make one myself, which took about 40 minutes with lots of tears & a bit of blood, as her skin is like purple-black raw steak. The nurses that came round yesterday are going to put in a compliant, as the emergency phoneline I called to see if I could take the dressing off should have come round and replaced it - but instead they told me to take it off and wrap it in a bath towel. Which would have re-infected it again, of course. I'm trying to do whatever the medicos tell me to do but I instinctively ignored that one, which was nice.

I have a week off work from this afternoon, and am not going anywhere apart from Nodland & looking after Clare at home. I don't think I've ever been so tired, even in some of my past jobs.

In unrelated news, my new domain vodex.net is under development, and it's a good opportunity to re-organise its folder structure (stop yawning!) to a flat system, which in turn means I get to re-write my rather crufty menu-generating code. I've had a look around and nothing seems to fit my requirements - putting arbitrary URLs into an arbitrary level structure, with the menu 'folder' being expanded if you're on the current page. (This happened last time, which is why I wrote my menu before). PEAR may do it, but not all web hosts support PEAR. In fact, it's surprising just how poor many of the scripts existing actually are. Again, the promise of code re-use starts to fall a bit flat the moment you go off the traditional path.

BTW, train of thought this morning on way to work - did I ever say that I got in the paper for seeing a UFO?
( In Here )

Huzzah! In one night (and after quite a day), I have produced my first ever Wordpress plugin.

http://www.vodex.me.uk/projects/lj_feed_import/

It's still under development (needs to be added to the admin option panel rather than just running every time), but it automatically imports Livejournal posts into Wordpress, and not the other way around as most plugins want (post to Wordpress, auto-post to Livejournal). This is what I want (LJ is my primary publishing tool).

It needs more work, but not at this time of night...!

Aha! My little experiment in practical search engine content discovery is accomplished. Namely, I put a page up about something very obscure on the 20th June, and waited until search engines found it.
Google, for example, has only just stumbled upon it - apparently when I updated some other pages yesterday, it rummaged around my site (good to know it's in the 'check for changes daily' category!) and discovered it then. (Note how it doesn't have the 'cached' date for the page like it does for the Misc page that links to it.) Perhaps it's due to how it sees the 'freshness' of my site?



(Also showing up on MSN and Clusty.)

I was asking about how to do this some time back, but clean forgot to mention that I'd implmented it.

http://vodex.net/wikipedifier/

Easily converts text with [[Wikipedia markup]] to full-blown links (original plan to automatically search plain text for existing articles was too computationally/networking intensive and error-prone as to what makes a phrase).

It doesn't let you convert links like [[Eagle (comic)|]] to just 'Eagle' yet, however.

I've uploaded the menu navigation script that I wrote for the site. I expected to be able to just re-use a public script, but none of them either do what I want (just refer to the page and create the menu without interaction) or look nedlessly complex in a Slashdot-user kind of way.

Anyway, it's at http://vodex.net/php/nav_menu/

I also had to write a trivial contact mail script. Funny how hard it is to find a simple script to do what you want to do, that *just works*. The promise of open source software, eh?