Transparent PNGs in IE, File uploads in SVG

By now everyone has seen Sam Fortiner’s excellent post about implementing PNG in Internet Explorer, which is great news, and fulfils a long-hoped for hole in IE functionality.

More amazing is the factoid found in the comments during a discussion of whether IE should support SVG natively:

Did you know that they have added file uploads to the latest SVG drafts? It’s utterly crazy for an image format to have features like this.

http://www.w3.org/TR/2004/WD-SVG12-20041027/api.html#fileupload

He’s not kidding either – that link really works, and really does link to the part of the SVG 1.2 Working Draft, which is a last call document. I don’t think I can really offer any insightful commentary into this; I’m too stunned.

Game Completeness (or not)

I’ve been slacking quite heavily on my computer games this past few weeks (read: year) which has left me with a load of games which I would have liked to complete, but haven’t:

  • Super Mario Sunshine (57 shines)
  • Prince of Persia: Sands of Time (c. 50% of the game)
  • Legend of Zelda: Ocarina of Time (Forest Temple, c.30% of the game complete)
  • Ico (10%)

Less important, but still outstanding are my GBA games:

  • Legend of Zelda: Oracle of Seasons (end of game boss)
  • Legend of Zelda: The Minish Cap (3rd or 4th dungeon of five, can’t remember)
  • Advance Wars 2 (c. 50% of the game)
  • Golden Sun 2: The Lost Age (c.40% of the game)

Of all of these I’ve listed, The Minish Cap is the one I was enjoying most, and am likely to continue with, despite it being on the GBA (fortunately I have a GBA player so can play it on my Gamecube); Ocarina of Time hasn’t really grabbed me so far, and, because I’d already completed Legend of Zelda: The Wind Waker, the controls feel wrong and the combat more limiting, so I don’t know how much further I’ll pursue it. Super Mario Sunshine has been a good challenge so far, and fairly enjoyable although marred by a bad camera – I’d certainly like to finish it.

In addition, I still have Legend of Zelda: Four Swords Adventure which I got for my birthday in January but haven’t even started! With any luck I’ll find out some of my friends have got GBAs so we can play it multiplayer, the way it should be 🙂

Despite all this though, I’d still really like to pick up Paper Mario: The Thousand Year Door and possibly Tales of Symphonia. I think I’m a glutton for punishment 🙁

Personal web proxy: MindRetrieve

Since Jim stuck it in his del.icio.us bookmarks a few weeks ago I’ve been using MindRetrieve both at home and at work to proxy my web browsing, and I really can’t fault it.

As you browse it caches and indexes local copies of webpages so that you can search through them later, and even view the cached version if you’re offline. It strips the pages of all HTML except for simple paragrpahs, presumably for easier search indexing, which means no CSS o images, so the cached version is as plain as can be, but this is a fair exchange for disk space and proxy speed. It’s written in Python and uses (I think) PyLucene for the indexing, and is, as a result, very quick. Additionally, since it just uses a query parameter on the search page, it’s very easy and quick to set up a Firefox keyword so you can search from the address bar: I just type “find dvd” and up come all the pages I’ve viewed with “DVD” in them, ranked by relevance.

I’ve used a fair few personal proxies in the last few years, ranging from the very simple to very complex, but MindRetrieve really hits the sweet spot for me. It’s easy to install, set up and use. The proxying is so fast that you forget your’re using a proxy at all, and the searches are fast and the results are relevant. It has the added benefit that the admin UI and search results are all editable because they’re plain HTML and CSS, so you can make them look however you want, even to the extend of adding JavaScript to add in linkbacks from del.icio.us and so on if you like.

Simple and does the job, I don’t think praise gets much higher than that 🙂

HOWTO: Relax at the weekend

Whilst looking after my ill girlfriend this weekend (or rather, not; preferring to let her just get on with it), I’ve discovered the trick to have a fulfilling yet relaxing weekend: read the weekend newspapers.

Honestly; it sounds trite and prosaic but who actually does it nowadays? We spend so much of our time plugged-in and hooked-up that to take the time to sit in the lounge, sun streaming through the windows and a newspaper (or one of its many weekend supplements which are oh-so-common) on our laps is almost unheard of. Nevertheless I can inform you, first-hand, that it is a remarkably rewarding experience which I recommend to one and all.

FOAF TV

So, how about this for a FOAF application idea. How to find out what other
people you know are watching, and hence, what you should be watching.

Take a service like tv.bleb.org, GreaseMonkey it to add a button “I want to
watch this” to each programme on the page (maybe with a “repeating
item” checkbox), publish it as an iCalendar file (so you can subscribe
to it in your calendar app and be reminded when it’s on) and rdf-ical,
and link to the rdf-ical in your FOAF.

Then spider the “What I’m watching” files of other people listed in your FOAF file, and see what they‘re watching that you‘re not. You can probably derive a
best recommendation by who has the most in common with you. You can
also spider out further to find out what other people are watching that
you‘re not, to see what you missed.

There was some talk recently I think about Social TV in terms of Tivos (like Tom’s post about Social Software for Set-Top boxes…
and there was something else on PVRBlog recently that I’m too lazy to
go and find), and on-screen display of the other people you know and
what they‘re watching (via the magic of webcams and so on), so you
could watch a TV show collaboratively and IM/voice chat during it
(namely going “I can’t beleive Rafael wasn’t gay!” if you‘re watching
“Playing it straight“).

So there we go then – a nice project for someone: auto-discovery of other people within six degrees of FOAF watching the same show that I am, right now, and being able to chat/instant message with them. Ideally right on your telly via hacking MythTV or whatever else people use, or is hackable (I have no idea), but via the computer is good enough.

Is there a dearth of FOAF apps?

Someone posted to the rdf-dev mailing list the other day, saying FOAF had had its chance to do any good, and had failed. Time to close up shop.

He missed the point. FOAF is a data format. It’s the surrounding application infrastructure which needs a kick up the arse, not the format that needs killing.

As of right now, I can’t give you one good example (off the top of my head) of an application which uses FOAF to make my life easier (or, in Jamie Zawinksky’s terms, get me laid). I can use several FOAF browsers, but frankly, whoop-de-do.

I posted a week or so ago about spidering people you’re subscribed to, and discovering their blogrolls and so on, hence being able to find who they’re subscribed to that you’re not. Where’s the FOAF app that does this? I need to be able to generate FOAF from my reading list! (Aaron Boodman also had something about this on his blog, but I can’t find it right now)

A little while ago I wrote FOAFlicious, an app which generates a FOAF file from your del.icio.us inbox. What this means is that you can then spider the inboxes and find out who everyone you subscribe to is subscribed to, but you don’t – it didn’t occur to me at the time to do this, but I’m reasonable amazed no-one else has done it in the meantime.

I don’t know if I’ll have time to do that anytime soon, so please, someone, do it for me! Write all those cool FOAF apps that will show that FOAF really is worth it!

Jabber transport stability

Peter Saint-Andre has published issue #22 of the Jabber Journal and lists what he thinks need to be priorities over the next 12 months.

I certainly agree with his first two (friendler clients, easy-install/config servers) but the others I could probably take or leave. For me a more important issue would be maintaining stable transports in the wild.

Most people using a Jabber client for the first time probably have at least one account on another IM system (for the sake of argument, let’s say it’s an MSN Messenger account). The first thing they’re going to want to be able to do, having been seduced, encouraged or forced to use this Jabber client and having heard of its abilities, is to connect to their MSN account. Let’s skip over the difficulty of browsing services on different servers (entered manually btw) to find a suitable transport that actually works, and skip over the barrage of authentication messages they’ll get when they sign in for the first time, not knowing what to do with them. They’ve got a transport, it works, and they’re chatting to their friends when one day – no response. No chatting for them that day! Or the next! Or the next! And typically this will persist until someone running the server notices and restarts the service, except of course it will go down again at some unspecified point in the future for an indefinite period of time. But with their new knowledge of browsing for services, the intrepid user has already gone off and found a new working transport, and signed up for that. Until that one goes down. Let’s not even talk about what happens when they’ve signed up for another but haven’t deleted the first transport from their roster and it finally comes back online.

I’ve not really been using Jabber clients that long – a bit more than three years I guess, and in all that time I’ve probably used about ten different MSN transports, switching as each one proves itself to be unreliable. I have two MSN accounts, one on anywise.com and another on admin-networks.org – they’re both pretty good and have very low periods of downtime, but there are other problems, like when I close my Jabber client, I’m not logged out of my MSN accounts. It’s all such a trial.

Hopefully we’ll start to see not only much more reliable transports in the near future, spearheaded by PyMSN-t and the other Python-based transports (which I’d be running myself on jabbernet.org if I could figure out how to get Python 2.3 running on Debian stable, I’m more of a RedHat guy), but also better detection, configuration and a great end-user experience.

Bigger back button extension for Firefox

A little while ago I wrote some CSS that applied Fitt’s Law to make the back button in Firefox bigger.

At the time, someone said I should make it an extension, and I did (based on Chu Yeow’s “making unred tabs obvious” extension), but forgot all about it.

It wasn’t until my Firefox installation got corrupted and I had to do a clean re-install that I realise how utterly vital it is, and I became really glad that I had made it an extension. Jeez – how do people cope with that insanely small back button? It’s crazy!

So, here it is: Install Fitts’ Back Button.