But later, if you rename your Project (emergent design: I’ve refined my ideas so I want to change the names of things), you can also rename the Group too to match. But the directory stays with the original name. You can supposedly rename the directory – but look at that shitty procedure: edit the project file in a text editor and perform a search-and-replace. I’ve tried this about three times on my project and it ends up in some shitty place of hell with linking errors complaining that the file is not in the old directory name. I’ve done a find . -type f -exec grep -H 'oldname' {} ; on the bloody project directory and there’s no reference to the old name left that I can locate. Revert revert revert! Try again. Fail again! Revert again! Rinse and repeat.

Get your goddamned act together Apple. This stuff is a hilarious joke! When you rename the group the fsking IDE ought to rename the corresponding directory or at least give you another option to do this! It’s like some nightmare out of an early-90s Borland C editor scenario. Wait … scratch that, those things usually handled this way better than Xcode does today.

And renaming Test classes should not be that hard either. Developers need to rename things after they’ve created them! Christ on a biscuit. Code organisation 101.

[face palm:[head desk]];

Also, this is a semi-description of the installation ‘TELEMAT’ we had at TISEA in 1992 that Amanda mentions.

3:712/634   Scot Art               Woolloomooloo NSW     Sys-Ex BBS

Note slides 4 & 6 particularly. Note that “tablets” are expected to outsell “laptops” this year. See how quickly the “Wintel” model of operating system and hardware is falling off? This is a rapid change in fortunes for the Microsoft behemoth.

In 10 years, maybe less, Microsoft will be a bit player. Few people will use an MS operating system; certainly their sales of new operating systems will be insignificant. Mostly people will buy just business productivity apps off them (Office/Exchange).

Of course, they may improve their Surface experience and perhaps catch up to iOS and Android, but I think at this point, they will still be a bit-player in that market. Why do I think this? Because of their insistence on the Windows 8 “you only need one operating system” mindset. They seem heavily invested in. They may be able to break this decades-long allegiance to the great god Baal^H^H^H^H Windows, yes, possibly. But I don’t think they have realised their lunch is being eaten quite so quickly.

Just for the record, I also think that OSX sales will also be tiny in 10 years time. Only “people like us” (developers) and a small minority of “power users” will buy “computers” that aren’t tablets. There’ll still be servers, of course, but that market’s dominated and will still be dominated by flavours of Linux and VMWare running on IBM/HP style hardware platforms (with a small share to their proprietary OSes for existing high-end customers). Of course, Windows may still play in this space, for people with dot-net services. And it may be that’s how MS saves it’s bacon. But, that stuff just doesn’t have the revenues that Windows-to-consumers has; the alternatives are basically free (thus limiting price) and the volumes are relatively low.

And so, as a *dominant market force*, it’s already living on borrowed time, unless sometime in the next 12 or 18 months it gives up on the “one size fits all” approach and creates a SurfaceOS that really knocks people’s socks off. Do you think that’s in the Microsoft DNA? Not bloody likely. But my bet is that they won’t try, and even if they do, it will still be a horrible frankenstein’s monster (Ballmer’s monster)?

While I think that Android had some time to iterate over their touch phone and tablet interface until they got it right, I don’t think that Microsoft has that luxury. Android tablet makers have been learning as they go along, as has Apple. But they’ve been busy learning these lessons while their market share was relatively tiny. Microsoft has to catch up and overtake them — soon to be the market-leading portable computing platforms — more or less in two or three leaps; they’ve already expended one.

I really ought to make a more hard-core effort to learn Objective-C and the Apple APIs.

You can customise which engine you use by default; I use xelatex. However once I did that I found it still compiled the file I had open with pdflatex. After clicking about a bit I found that one of the icons in the information section (the little cog) allowed me to change which engine I was using for the current file, which is apparently remembers.

The build started to complain about a file I was \input’ing which could not be converted to UTF-16, However this is not a problem I have encountered with the same files under TeXShop. For the sake of getting things to work I commented out the \input line and continued with my demo.

Like most of these programs if you’re used to a really powerful editor (e.g. BBedit, Textmate, EMACS, a programmer’s IDE) you’ll find the editor lacking in many advanced features. It’s not even up to TexShop’s ability yet. That’s why I’v always found it necessary to keep versions of those editors around and available to edit file if I need to do complex search-and-replace based on regex, for example.

It has some syntax highlighting which is customisable to some degree. The default syntax highlighting is a little too subtle, everything being various shades of blue, except maths macros, which are red (not being a maths/science person, I always feel a little left out). I tried out the maths macro just to see and it puts only the dollar signs ‘$’ in red, the content between them is left in the default colour.

Some more work could be done on the syntax colouring. All arguments in braces {}, including the braces themselves, are coloured the one colour. So if I write a footnote, it looks like this following screen cap:

(I’ve change the default colour here to green). As you can see, what is hidden from immediate display, is my use of embedded formatting commands inside the \footnote{}, viz, the ‘\emph{lectisternium}’; it’s all just a big block of green. I think a number of solutions ought to be applied to improve this:

• embedded, recursive, arguments should be coloured differently (perhaps just getting darker with additional recursion, or cycling through a preset palette selection).
• commands inside commands ought to be highlighted.
• certain common commands like \footnote \emph \textit \textbf \section etc should have separate colour schemes.
• the brace colour should be subtly different to the colour of the argument content (eg. slightly darker).

By default the editor is set to “hard wrap” which meant my soft-wrapped tex file line ran off the right of the page. There is a global option to change this, which I had to discover, only after looking for, and failing to find, a “view” style option to apply to the current window.

The “live update” is an interesting idea but its execution needs improving. Of course while you’re typing the poor thing is just about continuously compiling or trying to compile to PDF. But what happened is that the PDF preview did not jump to the place where I was typing. It did seem to jump around (to the last page in my case) but not exactly to where I had made changes. Which sort of defeats the purpose a little.

Well, that’s about as much time as I should spend on this, took me longer to write than to demo.

1. Update pronunciation of map components. Multiple issues here.

2. Map data need to contain the “Display” name (shown on the map) and a “pronunciation” name (for the text to speech). How can the voice mis-pronounce a place like Greenslopes? That’s two common English words, Green / Slopes, in a compound word.

3. Things like “STATE ROUTE 22″ or “NATIONAL ROUTE 3″ do not need to be SPELT OUT which is what it does currently (i.e. “Ess Tee Ay Tee Eee Arrr Ohh You Tee Eee Twenty-Two”). I’m looking at you Tom-Tom. The first time I encountered this in the device I nearly had an accident as I was at a complex and unknown intersection and suddenly the device is spelling out words which I am trying to piece together what it was telling me while navigating a high speed intersection / freeway off ramp.

4. Have an easy way to program “I don’t need this part of the route” into the device. For instance, to get from my Home to the major freeways/tollways nearby. I know how to do that, and I have preferred routes (paying attention to things like time of day, difficulty of turns across oncoming traffic, etc) for each freeway nearby which are different from the ones the device likes. Its really annoying to drive 100 metres from your house and have the device furiously be recalcing and telling me to turn around or make weird turns to get back on the route it wants me to take when I KNOW WHAT I AM DOING.

5. Sometimes what I need are directions to get onto the freeway in the right direction from where I am. Once I’m on the freeway going in the right direction (to/from city), I know where I’m going. The navigational problem is that I don’t know the current local area and the best way to get to the right freeway on-ramp.

6. Stick to the Main Roads on the way there. I’ve driven to places where it’s told me do make some STUPID SHORTCUT down a series of local streets because that’s 1500 metres shorter than sticking to the main road. NO. Stop it. This is OK for locals, but those people don’t want me transiting through their local streets.

7. Long-distance travelling modes. I know that Euros don’t usually have this problem so they never really think about it, but it’s certainly true in the USA and Australia. Sometimes I know where I’m going and how to get there. What I want to use the device for is to warn me about speed cameras, tell me my current speed (more accurate I think than the speedo and important for the speed cameras) and give me on the screen the distance to the destination and time of arrival. The last two won’t work without a destination programmed into the device. So they need a sort of mode to shut the navigation instructions up – this ay do also for points 4 and 5 … a “shut up until …” mode or even a simple always-visible on-screen button or one on the hardware that mutes only the navigation instruction voice when you don’t care about but leaves the safety camera warnings ding.