Well, tonight I spent the night loading Wordpress 2.5 (which is pretty slick). The main reason for this is that I discovered earlier tonight that some lowlife was exploiting a vulnerability in the previous blogging script to create redirection pages for spam websites.
So…hopefully it is all fixed now, and I begin the long road to clear my domain name of being associated with spammers.
Please do forgive me great googlebot! I promise I wont use old hackable scripts again…
Thats a whole-lotta-caught-comment-spam!
After a little tinkering here and there, and a few hours where everything was broken and I was to be found furiously grimacing at the laptop screen, the tag system is now finally fixed!
The tag system got broken when I first added the morehistory.js or EverScroll as I like to call it. The break was caused by some rather inflexible code which I had used to get wordpress to pickup the right posts to be displayed.
After an extended stretching session, it is now limber enough to be able to call only posts with particular tags (such as imagery, tech, video etc). The same inflexible code had to be fixed on the AJAX portion which makes the content appear as you scroll, this was a little more complex, but is now also fixed.
I also added a tweak to remove the rather nasty category pages and replace them with the nifty everscrolling tag pages, so you can now press the category links at the top of the post to be spirited away to a listing of posts in that category, displayed nicely…with everscroll 
Then let me put the entries on this site directly in your inbox.
Since my rss subscription list is growing on a daily basis (except the days when it shrinks), I thought I would also add the ability for the rss-challenged to subscribe to updates via email instead.
Thanks go to feedburner for making this feature so simple to add.
Version 3 of the blog is now online, many things are probably broken…I will work through them.
The design is influenced by numerous other sites which I have a great deal of respect for, most notably Khoi Vinh’s beautiful Subtraction and Michael Heilemann’s technologically amazing Binary Bonzai.
Perhaps the biggest thing about this update is the wordpress hacking that went on before hand. This all began a number of months ago when The Humanized Reader first came to my attention. This is the future, I thought…page-less websites, no more next buttons brilliant! I then got thinking about how I could implement a similar thing to the blog, and one weekend when all other entertainment had run out, I decided to dissect the code and work out how it was working and whether this was even possible.
I will be the first to admit that JavaScript is not my strongest language, not by a long way. But after a little tinkering, I managed to bodge a working version of the humanized code into the existing wordpress code that runs the blog. So now when you reach the bottom of the page, the little ajax request object goes speeding off and grabs the next five posts, and lo, one ever-scrolling edwilde.com (well, until you reach the first post - I tried, and gave up after 10 minutes of scrolling).
Unfortunately, in the process I have somehow managed to break the tagging system, I am sure it isn’t critical though and it will be fixed in time. I also dread to think what this does to the server load, I will be implementing a caching system pretty soon to elevate this. In the meantime, I am sure it will survive.
Over the past few weeks I have been heavily involved with creating, tweaking and all too often just reading rss feeds. It all started with my desire to put an rss feed on the sculpture website for the newly designed news section. This meant that I had to code the rss myself, and validate, test etc…
This was actually quite entertaining, as I didn’t really know that much about how rss was structured and what standards need to be adhered to. Although at times it was rather frustrating, as I found that most rss readers really don’t give any leeway for bad characters, or bad structure. Literally one & in a item would stop the whole thing from displaying…with no error message either.
Of course the most difficult part was actually explaining to people what I was doing and why it was worth the time and effort to do it…hopefully it will become evident as time goes on. I kept trying to simplify and simplify, but as soon as I mentioned rss, their eyes glazed and drool began to drip from the corner of their mouth. Ahh well…
Anyway, the main point of this is that once I got a validated, workable rss xml document up and going, I discovered feedburner. Quite a few popular blogs use it, and I had always assumed it didn’t really do much which was useful to me. How wrong I was!
Not only is it a complete doddle to setup, it also has some neat little features. The first and best is the way it presents your feed. No longer are newbie IE users sent to a bewildering xml document when they click the subscribe button, feedburner has a nice gui for your feed, allowing you to automatically add it to almost all the popular readers. Personally I do think the landing page could be a little more newbie-friendly though, perhaps a brief non-tech explanation as to why rss is good as well as some download links to popular readers, or links to web-based services.
It also has a nice statistics engine, as you may have noticed by now I kind of have a statistic addiction…the business model for feedburner is based around the need for more statistics, you pay $5 per month to get all sorts of extra info about your readers…seems quite a good plan.
Some of the other benefits are the feedflares, which add little links to the bottom of each entry (’email this’, ‘add to del.icio.us’, ‘digg it’ etc). and possibly most useful for the sculpture site is the ability to subscribe via email, perfect for the non-rss-enlightened reader. You can also chuck in your del.icio.us bookmarks or diggs for the day into the feed, although for some reason not both yet.
All in, feedburner is a fantastic service, adds value and costs nothing…which is what we like!
The photo archive is now online, thanks to the magic of the FAlbum plugin for wordpress. The photos are taken from my Flickr account.
Now all I need to do is take some more photos…
Ha…no way I would sell my domain name for this! still, good to know though…
Just incase you haven’t guessed by now, this website attempts to put a value on a domain name. It uses things like google/yahoo ranking, as well as the structure of the name itself (how long, how memorable etc). It was also nice to see that the sculpture website was rated very highly…so much so, that there is a warning that the information is only an estimate.
Made some changes to the stylesheet, including a new header image. Still not 100% sure whether it will stay or not, it was design out of necessity as I will (am aiming to) add some more sections to this site including my video work, flickr set and perhaps even an about page. Now the bottom part of the header is dark enough to discern (perhaps not read) any further pages which are added, although there is still work to be done…
Is currently only displaying my diggs as I am still yet to decide on a good method for integrating the other feeds.
The options I can see at the moment are:
Take your bets people…which one will I choose…?
This (potentially) will be a listing of 15 links which will be located at the bottom of this page, next to the previous 15 (get it, get it). The functionality is running on my offline site, well…just about running…
The future 15 will be compiled from a set of rss feeds, generated on other sites by me. The RSSindication plugin gives me the ability to display an RSS feed on the page, this is the simple part. The plugin also offers the functionality to combine one or more feeds into the single display.
What it doesn’t do (yet) is sort that combined list, it just combines them sequentially. This isn’t much use when you only have 15 slots to be used, as they simply end up as just the first feed’s content.
What needs to be done here is that the items in the feed must be sorted by the date they were added. This is one of those things that is real simple in concept, but usually turns out to be a pain in the ass once I look at the script which needs editing…
[5 minutes later]
Even worse, I just had a brief scan of the code and it looks like it is supposed to be doing that already (makes sense, i guess)…except it doesn’t appear to work.
Assuming that gets fixed, I would be adding my diggs and stumbles as well as my del.ici.us inbox (although I *technically* have little control over what is displayed there).
[15 minutes later]
This may make little sense to most, but as I am writing this in a typically unfocussed way I am also looking into the problem. It would seem it isn’t a script problem at all, simply the fact that only digg actually has the right tags in the rss feed to do this properly, stumbleupon has one date for the whole feed while del.icio.us has a different format date. I guess the rss standard isn’t quite as standard as i thought…time to come up with a ‘plan b’
To continue the current trend of beta’s on websites, this (as you see it now) is the first iteration of the design. I added the screenshot just incase I ever get around to either finishing or changing it.
It is an example of automatic design. I didn’t really have much of an idea of what I wanted the design to look like, other than having a single column layout (mainly because I really like the feel of binary bonzai). Simply a case of tweak something, decide whether it looks better or worse, tweak something else etc…
Of course all of this was done through the nifty tft goodness of my laptop. Most of it is somewhat invisible on my crt so I may need to whip round the css and darken the greys a little.
The big ‘e’ was a lucky piece of photoshopping, the design needed something to balance it…so I used the design equivalent of duck-tape to do it.
In terms of functionality, the main task at the moment is to get my diggs and stumbles at the bottom of the page, next to the previous entries. The related entries plugin looks entertaining, although it works using a field which requires manual entry (due to the conversion from my old blog). So I might just cheat and try and edit the plugin instead
As far as content goes, I do aim to add my portfolio to the site. I doubt this will be difficult, simply time-consuming. As well as horribly self-reflective, no doubt…
phew…categorised 429 posts since the last post. Still haven’t really mentally visualised a style for the site yet. So instead I am fiddling with the scripts on my local copy of the site.
One thing I have found, is that wordpress does look very intimidating for a casual programmer, lots of files, classes and functions. But it is nice to see that the very basic header, loop, footer structure, one would expect, is there.
The other nice part is that it seems to be the ‘done’ thing in this massive community to simply re-touch the bits you don’t like in the script, maybe even add something new. Pah! and some say the open source community doesn’t exist.
So much yet to do…no real rush tho, i hope…
Well, the archive is copied over to the wonderful wordpress. Of course it still each post needs to be properly tagged, categorised and named so that the nifty related articles plugin will work properly.
Oh yea, and I need to customise the theme a little, although K2 default is very nice, it isn’t exactly standing out from the crowd.
all fun though…
Back to basics, due to the complete lack of action on the blog redesign I have kicked myself into a simpler way of working. Get the content up, and edit the template…rather than edit the structure, backend, template then attempt to get content into it.
This does have a few downsides, my conversion from the collection of txt files to a neat sql database was not without it’s problems. This means that most of the posts which are now in the database are a little wonky, so expect to see lots of ?’s floating around, coupled with missing images and bad spacing…hopefully I will get chance to try and tidy them soon. Might be a good procrastination before putting a new template together…nothing worse than uncusomised templates!