Previously I could have (and planned to) set this kind of thing up via the post by email feature, but there’s a certain elegance about having a native app for the purpose. I’m sure this will briefly spur my interest in it though and I’ll do some fiddling with the pop3 method for comparisons.

Some initial observations:
* No way of doing anything other than plain text, hence the stars (unless HTML works..)
* No way to control posting settings like tags etc. An overall “moblog” tag would be good as an automatic, which pop3 can do.
* The full version of Wavelog allows posting of media such as images, audio and video files. This could well be the crux of my (future) comparison with email bloging, as it’s a really useful feature I have an idea for.. You’ll have to wait and see..
* No way to set the cut-off point for post previews.
* A couple of minor bugs I’ve noticed and will pass on to the authors. Maybe I’ll even get a free licence out of it.. (Can but hope, eh?)

Overall though it seems to be a good piece of software that does the job, with plenty of room for feature additions in future versions.

Edit note: This post has been edited since posting to tidy up, insert the break point, etc. It turns out HTML is supported, but a T9 phone keypad isn’t really the ideal input mechanism for writing code. A more major flaw, to my anal web-standards self, is that the whole post gets created in a table with the different sections being the cells. There’s really no need for this when the same can be achieved through float and clear, though I guess it would have to be a bit cleverer to cover the many quirks of CSS.

So, opinions. A mixed blessing really. User interface elements are a lot nicer, particularly the new reflective dock (which seems to be a bit controversial round the web). There’s also a neat little hidden trick for putting spacers in the dock, a problem I had to hack around in Tiger by placing “fake” apps in the dock which threw up errors on accidental clicks.

The spacers tip comes courtesy of a post I found on the MacNN forums, which basically involves adding extra entries in the dock’s plist with the command:

1. defaults write com.apple.dock persistent-apps -array-add '{tile-data={}; tile-type="spacer-tile";}'

Stacks are a good idea, but I can’t see myself using them beyond my current “Downloads” stack that’s tidied up my often messy desktop. The one thing I don’t like here is the “stackyness” of the icon given to them. I think the idea is that it looks like a pile of files that you’d make on a desk, except it’s just a bit too neat. No pages sticking out at jaunty angles, folders with their contents falling out, or even any shadows to show height. I’d be happier if I could just replace the icon with a custom one that represented what I used that stack for.

Spaces is the only other major change I’ve used, which is an achievement considering the number of times I’ve tried and given up with virtual desktops in the past. There’s not much to comment on, other than it being implemented really well, and I’ll probably stick with using it for a while.

Oh, and the much touted, improved networking in Finder. Or rather, the lack of it. Finder successfully discovering my Windows shares had always akin to me being able to hit a playing card with a dart while blindfolded,  but thankfully when Steve strolled out last MacWorld to show off Leopard he told us Finder would now have improved networking features. I still hit the same problem of only being able to discover Windows PCs on a full moon, and now to add to the mix I get errors (6602, for anyone interested) when trying to connect to password protected shares on Vista. Clearly a huge improvement.

Luckily I have found a workaround for this problem, but oddly it comes in the form of a simple Applescript:

1. tell application "Finder"
2. mount volume "smb://mediacenter/c$" as user name "frosty" with password "pword"
3. mount volume "smb://mediacenter/d$" as user name "frosty" with password "pword"
4. mount volume "smb://mediacenter/e$" as user name "frosty" with password "pword"
5. end tell

Which begs the question if it’s this easy to get working in Applescript, why does it always fail in Finder? And moreso, why do these two methods not just call the same underlying API?

Another downside is the default icons. I know, it’s easy enough to change them, but Apple pride themselves on their design – why choose these? As soon as I find some folder icons that agree with me these will be gone.

Anyway, enough with the negatives. Overall, Leopard’s quite an improvement over Tiger. There’s the big, noticable changes like the Dock, Spaces, Stacks and Time Machine (which I haven’t tried yet, and probably never will), but I’ve also noticed a lot of minor changes that just make it that little bit nicer to use. Toolbars are now almost universally consistent, rather than the mishmash of plain grey, gradients and brushed metal that were round every corner in Tiger. Terminal’s been improved with the addition of a remote connection window (to be fair this could’ve been there all along and I’ve missed it). The translucent menu bar is far more subtle than I expected, and not a problem to work with on most backgrounds. I’m sure I’ll come across more changes as I carry on using it, and hopefully the networking issue will be fixed in 10.5.2 due out sometime in January.

I won’t hold my breath though.

Check the “The Great Outdoors” category for my photos of Etherow Country Park and Werneth Low.

The wonders of Flickr…

Having hit this problem for a client site I’m working on, I decided to come up with a solution for it. In its current unpolished form, the Dynalinks plugin parses link tags as the page is constructed, so links will always be up to date. As for the unpolished part, it currently requires text to be inserted manually as so:

1. [link:14"]click here[/link]

This format will always provide a current link to the page with ID 14, no matter where it’s moved to in the site.

The plan is to extend this with a TinyMCE button to make inserting links easier without having to look up a page ID. Possibly down the line it may include dependency tracking, so a page can’t be deleted when it’s linked to, but maybe this is a bit overkill.

Update:

The script’s had a bit of an update so it now works with the URL ‘link:[page_id]‘, to fit in more with the editor and ease of use for non-techies. This also means it doesn’t need its own TinyMCE button.

Grab the plugin over at its project page.

I’ve spent far too long on this so far, but it’s been a learning experience, being my first delve into WordPress plugin development.

The planned feature list is: (probably a bit ambitious)

• Images grouped into albums
• Automatic retrieval of EXIF information
• Tagging and comments system separate from main WordPress data
• View tracking
• Bulk uploading with zip files
• Online image editing, including rotation and cropping. Possibly other features, once I look into GD’s capabilities
• Full integration into WordPress frontend, with template tags and content tags
• Added buttons in TinyMCE for inserting tags
• Optional use of mod_rewrite for tidy URLs

How many of these features will make it into the v1.0 release is debatable, but I intend to get them all working over time.

If this works as an exo-skeleton for the inevitable high-impact crashes, sign me the hell up!

This post brought to you by the ridiculous inventions thread on SomethingAwful, and the letter J.

• Upload the sin city art in design

I’m sure there’ll be more, but that’s a daunting enough list to get started on for now!

Semantic Markup

There are numerous advantages to using accessible, semantic markup when creating web documents. As the term suggests, this technique allows content to be more accessible to a wider variety of people, through the increasing use of assistive technologies such as screen readers. When a document is structured semantically, the task of representing this to different audiences becomes much easier.

While semantic markup and the semantic web are separate concepts, the adoption of these technologies by the wider community can be seen as comparable. Both require additional effort on the part of the page author, and advantages can be seen in both cases.

Despite the accessibility benefits of using semantic markup, and this now being a legal requirement for any organisation offering a service to the public, a recent BBC report found that the majority of leading websites still fail to meet basic accessibility requirements.

So why are so few sites accessible? For many, other than accessibility advocates, there is no reason to put in the extra effort involved. Unlike with conventional programming languages, web browsers will consistently display syntactically invalid markup, giving them no reason to learn the ‘proper’ way. In addition to this, WYSIWYG editors often output invalid or inaccessible code, as it is difficult to add semantics to the page without a knowledge of context.

The W3C’s Loss of Control

There is also a perceived lack of faith in the W3C, who originally set out to establish standards for web markup. Official specifications have become stagnant in recent years, with the last major revision being XHTML 1.0, published in 2000. The draft XHTML 1.1 specification was released shortly afterwards, but 6 years later a final version has still not been published. In contrast, unofficial developments and proprietary technologies have experienced prolific growth and adoption.

Javascript (officially ECMAScript), RSS (and also the Podcasting extension), Flash, Pingbacks, and more recently microformats are all unofficial/proprietary developments which have experienced near-universal adoption. Flash has arguably become the new standard for embedded media in web pages, due to its cross platform compatibility and non-reliance on video codecs. In the time since the last revision to the official XHTML specification, Flash has received 5 major updates. While this technology is inherently inaccessible, it represents the speed of web developments outside the W3C.

RSS was originally developed by Netscape (as with Javascript), but now has 3 branches in development: RSS 1.0, RSS 2.0 and ATOM. Despite its name, RSS 2.0 is a separate development branch which ignores the official RDF specification (unlike RSS 1.0), and instead uses simple XML without the concept of DTDs. This, like Flash, has become a universal mechanism without the support of the W3C.

I’ll post the rest of the article here in a few weeks when I won’t get done for plagiarism or something..

[link to digg story] [direct link]

Next Storyyyyy…

etc.