OpenID features

I've been thinking recently of two features which would instantly make OpenID infinitely useful.

Firstly a background application that, on start up, would sign me into my OpenID provider and authenticate me on all my OpenID enabled sites. This removes the need to have to log in each time (well the first time for each session) and also removes the confusing redirections, which could be a vector for a phishing attack, especially to users who are new with the OpenID concept of consumers / providers.

And secondly (and more complicatedly) a way share information between OpenID authenticated sites.

For example, I currently have a Twitter account and a Facebook account. If I wanted my updates on Twitter to also update my Facebook status, I have to pass my credentials to the Facebook / Twitter application before that could happen.

If OpenID supported the idea of trust between applications (maybe limited by specific APIs) this could be done automatically without giving the Facebook / Twitter application full access to my Twitter account.

Again, this would have to be done in a very smart way so users won't accidentally sign their accounts away.

And yes, due to past mishaps, I did check Ted Dziuba's site before posting this :)

The Mac finder team, Star Wars fans?

You decide

Melbourne's cashless society finally matures

It appears the Safeway on Chapel St is leading the charge on bringing Melbourne's EFTPOS system into the 21st century.

Spotted, the other day, in said supermarket.

[edit: the new thing is PINs for credit cards]

World of Goo

Another amazingly designed physics based game, World of Goo.

http://2dboy.com/games.php

A very indepth review (from a game designer's perspective) below:

http://blog.wolfire.com/2008/11/world-of-goo-design-tour/

Firefox theme for OS X

I had to recently recreate my Firefox profile again, so I've been (one at a time) reinstalling addons as I find I need them.

I realised today that I needed to install the delicious plugin, so a quick search of the word "delicious" and up pops up a search result of a theme called "grApple Delicious".

On closer inspection it looked like a Safari style theme for Firefox. I've always been a bit niggley about the default Firefox theme on OS X, but it's never irked me enough to decide to try another theme. With this theme staring me in the face I decided "why not".

Here is a screen shot of what it looks like:



Ok not 100% perfect. Everything at least fits the OS X look and feel (which I never though the default Firefox theme ever did) and even the close button on the tabs are on the correct side of the tab (the LEFT hand side, dammit!). The tab itself is not joining the top half of the frame though, is the wrong shade of grey, and the spinner / favicon is on the left hand side of the tab which looks awkward.

A bit of googling reveals the original site which created the theme, located at takebacktheweb.org. A quick look and I notice there's another grApple theme they've created called Yummy. The screenshot looks like it addressed the niggles that I had, so I download the theme. Screenshot of it running on my box is below:



They are minor improvements but I think it just adds that little extra bit of polish. The tabs are the right colour, the spinner is on the right hand side of the tab, and they are no favicons on the tab (favicon is only displayed on the location bar).

Perfect. My new default Firefox theme on my Mac :)

e-texteditor

Being an old Linux user from way back, my default choice of text editor has always been vi. When editing shell scripts, vi; when coding, vi; taking notes vi.

When I started my first development job and moved over to Windows I brought vi with me.

Sure there was Eclipse and TextPad but every now and again I'd encounter something where I need to do a column editing function, or a search and replace via regular expression, or just a very large file that I needed to load up which didn't cause the entire OS to hang.

Being a developer involves working in a team, and one time or another, they'll be a developer at my computer debugging a problem who would encounter vi for the first time. Unable to comprehend the reason they need to go into insert mode to type something, they would give up exasperated, stating that I should use a real editor, like EditPlus.

I've given nearly every single Windows editor a try and haven't found anything that would easily replace the functionality that vi provided.

Simple things like "find as you type", and highlighted search results would be basic criteria for me (note, both these features have made it into browsers like Firefox and Chrome, which only goes to show where all the innovation currently is).

Enter e-texteditor. Recommended by an old co-worker of mine, and branded as a TextMate clone for Windows the editor it is very sparse. I did my usual barrage of tests, first thing being "find as you type" and to my surprise what do I see:

Not only was there "find as you type" but as if almost like I was in vi again, highlighted search results!

And to replace the highlighted results, no need to open up another search window but instead, place the cursor in the "replace" text box and hit the "all" button. Everything highlighted is now changed, and is an easy way to see (just by sliding around the document and looking for highlighted text) what is going to get changed with a find/replace before actually performing a find/replace.

A bit more digging around and I find it has a built in indent guide, which draws vertical lines at where the indents are.

The indent is a bit smarter than usual too. Not only can you specify how many spaces a tab is, but you can specify soft tabs (i.e. an indent is equivalent to 4 spaces). This means every time you hit tab 4 spaces will be used but also (when using the cursor keys) if you move back over those 4 spaces it doesn't traverse one space at a time but 4 spaces at time.

Another nice touch is that to display non printing characters, i.e. spaces, tabs, carriage returns, instead of sifting through a menu option, all you have to do is highlight the area and all the non printing characters are displayed

My second test is usually column editing. Most text editors usually need to be put in a column edit mode, but e-texteditor all that is required is that instead of using shift and the cursor keys to highlight the text, you use alt instead and can highlight a column straight away. Then ctrl-c to copy, or just straight away type over the selected columns.

This hooks in with another nice feature they have where if you wanted to modify a couple of words in a document, you just highlight each word individually and just start typing. When select these terms, you have essentially put the cursor in those highlighted positions and when you type you literally see yourself type in half a dozen places at a time.

This fits in nicely with the column editing paradigm. As when you select a column, you are essentially just selecting that word. When you copy, push delete, or just type, you are typing over that word.

There is a video of this on the home page of the e-texteditor homepage which explains this better than I can explain it.

Finally regex, which works a charm. It's a little finicky to get to (I can't seem to find a keyboard shortcut for it) and you have to click the little magnifying glass icon next to the search to enable it, but the regexes themselves I haven't found anything fundamentally wrong with it so far. But that could be the fact that I've only been doing rather primitive regexes so far.

One more feature I'd like to highlight is that it is compatible with TextMate bundles (syntax highlighting definitions) so nearly any language you can think of is supported.

There is also a keyboard binding to switch between different bundles in the document, which comes in handy when editing a marked up file which contains a soup of HTML, CSS, JavaScript, scriptlets, etc and you want to highlight the different parts of the document as it isn't available in a single bundle.

This was about 2 months ago and since then I've gotten a lot more use out of this. Strangely also reducing my reliance on vi, and have even gone as far as purchasing a license!

I haven't really got any niggles with this except it crashes a lot more than I would like it to (not that often but a text editor shouldn't crash at all) and it also has trouble loading large files (i.e. files larger than 100megs) which vi just eats for breakfast.

I think vi will still be around a bit but I think I've managed to find an editor which I can use on a day to day basis where people won't freak out using when jumping on my machine.  

Gmail Themes

Who says google doesn't know design.

One fatal flaw with Open ID

If your Open ID provider is down, you'll no longer be able to log into any any sites which you've registered with that provider.

Ok is all ... 

Knowing when to break the rules

One thing which has irked me a lot recently is a piece of UI "correctness" that I've noticed when using Chrome.

Note the screen shot below:



If you can manoeuvre your mouse into the area that I've circled (in between the tab and the "plus" button) and click and drag the mouse around, you'll notice that the entire Chrome window starts moving around. The reason for this is that when you click this area, you are (essentially) clicking on Chrome's title bar and (as a feature with most OSes) when you click and drag a window's title bar, you are able to alter the position of the window.

What's wrong with this behaviour? Nothing really. It is extremely hard to trigger, and when triggered (and after having a good look at the UI) relatively easy to figure out why your Chrome window has repositioned itself about 5 pixels to the right, instead of opening a new tab.

Sometimes though it's ok to break the rules.

There's a rule in computer science called Fitz Law, which talks about how easy it is for a person to click on something on the screen with the mouse.

Long story short, the easiest places for a person to click are (in order of easiest to hardest)

1. Where the mouse is currently
2. At the edges of the four corners of the screen (all four corners are second equal)
3. And all third equal, anything hard up on any sides of the screen, i.e. hard top, hard left, hard right, and hard bottom of the screen
4. And for the purposes of this example: everything else
   

The reason is that to hit these positions, you can just "flick" the mouse right to each of the four corners of the screen (or to the hard left, hard right, etc) and don't have to slow down to find the exact spot, or go back again if you've overshot.

If you're using a Mac, to open the Apple menu you don't have to click right on the "Apple" icon (even though the screen shot below gives that impression). Just click on the top left hand corner, and the menu opens. Same with the spotlight window (on the top right hand corner).



If you're on a Windows machine, same with the Start menu and also the Task Bar. Notice how though, on the Task Bar (below) each task box doesn't actually extend to the bottom of the screen. If you click hard bottom of the screen, underneath the task, the mouse will actually "jump" up 3 pixels and selects the task (go on, give it a try if you don't believe me).


Which brings me back to Chrome's title bar "bug" which should (IMHO) open a new tab if the surrounding area around the "plus" button is clicked.

Competition in the spreadsheet space

I got a reply from some smart arses regarding my last post on more competition in the spreadsheet space and thought I'd post the following screenshot as a sign of what I overlooked:

Attachments with mail.app

I've always been confused with the way that attachments are handled in mail.app (the Mac's default email application).

Traditionally, if you'd wanted to attach a file to an email you'd just drag it into the composition window and (usually in a separate location in that window somewhere) there'd be a list of attachments that the email had.

Once sent, on the receiving end, you'd receive the text of the email, and a list of attachments (which you could then and download or preview individually).

On the other hand when dragging an attachment into the compose window of mail.app, it doesn't drop the attachment into a separate section of the email, but inserts the attachment inline.

This causes a few problems with some mail applications as:

1. There usually isn't an easy way to download images which are inline
2. You can't select all inline images and download them all in one hit
3. And more an annoyance than a bug, but when downloading the email, all the images will be downloaded there and then, instead of just downloading the text
   

The worse thing was that it was inconstant. Sometimes it would send the attachments inline and other times it wouldn't. But this was only on machines which weren't using mail.app. When using mail.app though, all images would always be displayed inline and there would also be an option to download all inline images in one hit. (Yes things are beautiful within Apple's wall garden).

I needed to find a way to send emails from mail.app so that people receiving my emails had a way to download the images I sent.

So I did a bit of googling and found an option (which many people claimed on message boards) called "Always send Windows-Friendly Attachments" which would fix the problem



A bit of experimenting with this and no. With respect to this problem this option doesn't do anything at all.

After spending a bit more time searching on the Internets I decided to check out the problem myself and maybe it was the way that mail.app was sending the emails and differences in the MIME encoding.

Viewing the source I realised that on emails where attachments were inline, the email contained MIME types of both plain text and HTML, and emails where attachments weren't inline the email only contained plain text. I did a quick test email to myself (note how I've got the word HTML underlined - more on this later):



And receiving it on the gmail end:



This confirmed one half of my suspicion. (Note how in the gmail screen shot there is no option to download the image which is attached)

I then decided to send this email again but forcing plain text this time (keyboard shortcut apple-shift-t):



Loading up in gmail shows up as expected, with no inline images, and an option to download the image.



So the reason for the inconstancy? Even though I've sent emails which have had the attachments downloadable (like in the second screen shot) and I've never forced plain text before (until this day), is that mail.app (even though I've set it to always send rich text emails) only sends emails in HTML if your email contains HTML formatting (the reason why I underlined the word HTML in my first email).

So looking back in my history, it seems that most cases that mail.app has inlined images is when I've sent an email with a link (HTML formatting) and an image from that link.



So in future, if I am sending a set of photos, I'll be forcing plain text emails. Otherwise it's the lolcat image with the relevant hyperlinks.

(Or I could just disable HTML emails altogher)



Note: In the raw email sent, mail.app doesn't just compose the email with the plain text at the top and the attachment at the end. It actually sends a plain text MIME component, then the attachment (the inline image) then another plain text MIME component at the end.

Most email readers (gmail included) seem to have trouble displaying this image inline when surrounded by plain text components, and instead puts all plain text elements together and provides an option to download the image.

Chrome Omnibar - Search contents of a particular site

I didn't notice this before, but I was using Chrome's Omnibar to load up Facebook, and before I finished typing in facebook.com on the RHS a little widget popped up saying "Press Tab to search facebook.com".

So I hit tab and the Omnibar brings up another widget on the LHS that seems to switch the Omnibar to search only the contents of Facebook.

Very useful, as the majority of the time when I'm searching I don't want to search the whole of the Internet, but just the contents of a particular site, for example Wikipedia

Strangely, the Facebook search results doesn't bring up the search dialogue like in the Wikipedia search result, but just a link to search

I did a bit more poking around and discovered that which sites are searchable are stored in the configuration. Just load up options, basics tab, hit manage:

And there's a list of sites which Chrome can search the contents of

Excel modal windows FAIL

I would have thought certain conventions in UI would have already been firmly established in this day in age. For example the the final result of the modal windows vs tabs debate ending in not one or the other, but both

i.e. as in the web browser space:

And similarly, Firefox, Chrome, IE all following suit.

Why is it then, that Microsoft Excel (the reason why the PC became popular in homes) still doesn't support modal windows which aren't limited by the application frame itself?

At least they got their priorities straight. Spending time on the Ribbon. And the ability to load up spreadsheets with more than 65,000 rows. Oh and fixing that bug which blew away your undo history the instant you saved an Excel document.

Maybe it's time for another competitor to enter the Spreadsheet space?

Joe the Plumber

This Joe the Plumber thing is getting blown out of proportion.

Here's a graph to solidify just how fair McCain's tax cuts are to 60% of tax payers (last three bars on the graph).

Actually, Joe the Plumber isn't going to be paying more tax with this $250,000/year income. He'll be paying the same as he did last year.

Who thought physics could be entertaining?

Remember Crayon Physics Deluxe?

I've recently stumbled upon a similar physics type game called Fantastic Contraption which is like Crayon Physics Deluxe, but not vapourware :)



I didn't realise how these games drew quite a few parallels with my profession until I saw other people's solutions and how I realised that (like coding) there is no right solution to solving the problem, but they are degrees of elegance in solving the problem itself.

I also found myself using common design patterns, such as all terrain vehicles, triangles, cranks and catapults.

Seeing other people use these same systems you could kind of figured out those who knew what they were doing and those which were just adding wheels here and there, making their stick "just that bit longer" so they could hack together a solution which would solve the puzzle.

These would make an interesting job interview question :)

My must have Firefox addons

• Australian English dictionary: Hey, I'm in Australia right now, why would I have a New Zealand English dictionary?
• Firebug: For all your web application debugging needs
• FireGestures: Going back and forwards is just a left and right swipe away, and flicking between tabs is as simple as a quick "left mouse click, right mouse click", and vice versa to flick the other way
• GreaseMonkey: When the formatting of a site is nearly "just right" and needs that slight tweaking, or when auto-complete for login fields just doesn't work
  
• Personal Menu: Gives me the option of hiding the menu bar (and quickly making it visible with the alt key)
  
• Session Manager: The ability to restore accidentally closed tabs / windows, and also remember entire browsing sessions (for those times when I need to restart FireFox but don't want to lose all my windows)
• Tabs Open Relative: When opening a link in a new window, opening the tab to the right of the existing tab instead of right at the very end of the tab list
  

What problem are you solving?

I've been following the progress of Stack Overflow since it's announcement on Jeff and Joel's blog, and one of the things that has irked me is that they've decided to provide authentication to Stack Overflow only via Open ID.

So for those who don't know what Open ID is, it's easiest to explain by providing an example. So lets say you want to register for Stack Overflow with Open ID.

The first step would be to pick a provider, i.e. Claim ID, where you register a name, password, email address, etc and at the end of your registeration get an endpoint (i.e. http://claimid.com/fooUser).

You'd then log into Stack Overflow and instead of creating an account you provide your Open ID endpoint (http://claimid.com/fooUser) and Stack Overflow will redirects you to Claim ID where you would log in with your Claim ID name and password and select the "trust Stack Overflow with my details".

Claim ID will then redirected you back to Stack Overflow and that's it, you've created an account on Stack Overflow.

So the advantage is that once you have an Open ID, creating accounts is really easy as all your details are stored on your Open ID provider and all you need to provide is your Open ID endpoint.

It also means that if you should want to log into another site, you'd just navigate to that site, put in your Open ID endpoint and you don't have to type in your user name and password as you're already authenticated with your Open ID provider.

What that essentially means though is that now you have a single name and password to log into all your sites. GREAT!

Wait a minute? Isn't having the exact same name and password for all your sites considered bad practice? And not only that, isn't there already a way to log into sites using the exact same name and password, which is BY CREATING ACCOUNTS USING THE EXACT SAME AND PASSWORD?

The pilgrimage to Mecca

What's a pirate's favourite letter?

"P" of course.  For pirate.  

What else would it be?  

"R"?  

That's so immature.  

Asthetics as a product feature

Found posted by DHH on the 37signals blog

http://www.37signals.com/svn/posts/1247-theres-no-shame-in-looking-good

Aesthetics is a feature in itself. One that I — and most the rest of the human race — is perfectly willing to let trump other functionality.

I think you’ve fundamentally misunderstood why people buy beautiful products, if you think it’s all about projection. While there’s certainly something to that (and I see absolutely no shame in that either!), it’s at the core about people feeling good about that which is pretty. That doesn’t make us shallow, that just makes us human.

While I completely agree with his sentiments, I'd have to say I think it's prominently western cultures which even has the notion that form can be placed on the same level as functionality.

http://online.wsj.com/article/SB122143317323034023.html

"The iPhone is a difficult phone to use for the Japanese market because there are so many features it doesn't have,"

Being someone of asian descent brought up in a westernised country, I'm sure I don't have to point out the dilemma that the iPhone presents to me :)

Varying degrees of evil

I had to find the following page for a friend today, which is a HOWTO for switching Azureus from the new "Vuze" layout to the old "Classic" layout

http://www.azureuswiki.com/index.php/The_Azureus_Experience

Classic Azureus:

Vuze:

There's nothing necessarily bad about the Vuze layout. It provides easily searchable content and a way to subscribe and publish to channels which you would ordinarily have to do manually (a boon for noobies).

But for some reason I can't seem to shake the feelings of evil from it. Namely the fact that Azureus no longer does what it was originally designed to do (download torrents) but seems to be providing a whole platform for delivering first rate value added media content to end users.

The old bait and switch (well, not really switch as there is still a way to get back).

I also can't shake the feeling that this "platform" will soon start to charge money for its use, and that those who have been suckered into either publishing or subscribing to the channels (i.e. have put a lot of effort in) will soon have no choice but to pay the cost.

This model sounds kinda familiar

And familiar again

The beginning of a new meme?

http://isbarackobamamuslin.com/

New Facebook

Contrary to most people I love the new Facebook. They're only 3 things I'm ever concerned with when looking at someone's profile which they've incorporated into the new design:

1. Their wall
2. Their info
3. Their pictures

I also think the links are better labelled, and the layout of the nested tab system is very easy to understand.

Sadly though, it appears the Facebook team isn't as confident in their new redesign as they should be and instead have littered the entirety of their site with tool tips which ordinarily wouldn't be too bad were it not for the way they designed them.

Firstly the tool tips aren't naturally part of the design and they stand out when you first load a page. This detracts from the main content and instead of taking in the new layout, you're busying trying to read all the orange boxes everywhere.

Secondly there is an option to remove the tool tips once you read them, and once you remove them there's no distinguishable way to get them back. This makes the user less likely to remove them and leave the layout in that jumbled mess.

1. If your layout is easy to navigate you shouldn't have to have help scattered everywhere
2. If you need help it should be easily discoverable
   

Product or service?

There's an old joke which goes, you can tell what you're getting (a product or a service) by when you get charged.

If you're getting charged before, you're getting a product (i.e. shoes, iPod, etc). If you're getting charged after, you're getting a service (taxi, plumber, etc).

When do you get charged when you go to McDonald's?

With this in mind, when do you get charged when you use Twitter?

What about Google Maps?

Statistics

The New Zealand Reserve Bank reduced it's OCR the other day by .5% (the 2nd time in rougly the last 2 months).

I wasn't aware that they had dropped the OCR in New Zealand prior to that, but I was aware that about a week ago Australia's Reserve Bank did the same thing, but only by .25%.  

I wonder what impact this has on house prices.  I have no idea how I could sensibly measure this, but being curious I loaded up the charts for both currencies vs the US dollar.  First, New Zealand:

Now Australia:

Afraid the LHC will destroy the earth?

http://www.hasthelhcdestroyedtheearth.com/

Also has an RSS feed, and some interesting comments in the source.

Innovations in EFTPOS

The regular way of paying by EFTPOS at the supermarket:

1. Wait for cashier to scan all your goods
2. Swipe your card
3. Cashier asks if you want cash (answer is always no)
4. Cashier asks "is that on savings?" (I don't know why they ask this in Melbourne, I'm completely capable of pushing the button to choose my account)
5. Type in PIN
6. Wait for transaction to process
   
7. Done

Saw this at the Safeway off Chapel Street the other day:

1. While cashier is scanning your goods, swipe your card on the machine
2. Pick either "full amount" or "full amount with cash out"
3. Pick account
4. Wait for cashier to finish scanning goods
5. Amount then pops up on EFTPOS machine
6. Type in your PIN
7. Done

EFTPOS has always been the slower paying option and ideas to speed it up have always been technology related.

It's good to see a solution which involves just a bit of process tweaking, which works just as well as getting faster hardware.

Swipe!

Spotted on 37signals

http://www.37signals.com/svn/posts/1240-swype-keyboard-for-touchscreens

A new input method by one of the creators of the T9 Predictive Text input system.

Brief usage rundown:

• Display graphical keyboard on screen
  
• Instead of punching out individual letters, just put your pen down, drag (roughly) over the letters you want to type
• Lift your pen up
  
• The word automagically appears

Incredibly simple. Not the obvious solution. In my mind, the "logical" progression from T9 on a full keyboard would be the iPhone solution, where you type out the text roughly and then it tries to guess what you type.

I would imagine an input system like this would increase the iPhone's usability infinitely, especially in portrait mode where the keys are incredibly skinny.

Ok I'm back

Lets try this again.