Interesting Monkey

More Police

To follow up on the last post, I think the Police were a much better live band than a studio band. The earlier records were under-rehearsed and the best parts weren’t worked out, and the later records lost all spontaneity. Most of the songs, including the three in the link above, got much better live.

There are some exceptions though. I always really liked the studio versions of Message in a Bottle, Roxanne, Spirits in the Material World and Secret Journey. But probably the song where I most think the live versions are inferior to the studio version was Every Little Thing She Does is Magic. The high hat in the verses and the bell during the choruses are so spontaneous, then well produced. I don’t think it’s possible to duplicate that performance. Sheer magic.

The Police

One of my favourite aspects of the Police live, was their use of digital effects in combination with their accoustic playing. Stewart would trigger these drum effects live with delays. He’d use them on all the drums and symbols and the high hat. Essentially, he’d use them to continue rhythms that he started playing. This version of Walking on the Moon is one of my favourite examples. It’s most obvious on the high hat here.

And the song that follows, Bring on the Night, is my favourite version of the song I’ve ever heard from any show, and is definitely better than the studio version which was a few years older. It took me forever to find the 1 in the rhythm, and I love Andy’s guitar part. Try counting as it switches from the verses to the choruses, and you realize the whole chorus sounds transposed by half a beat.

Google Chrome

As a test, I’ve used Google Chrome (on a mac) for the last week. The mac version now seems to be as feature rich as the Windows version, so I thought I’d give it a try to see how it compares to my browser of choice, Safari. I’ve always had a thing for browsers, and I have a general distaste for them overall. I think tabs are horrible, back/forward buttons are insane, the URL bar is ridiculous, and so on. I even wrote my own browser once (well, I started with an existing browser and hacked my own, which I think was not the best idea now, but anyways..).

After the week, I think I’ll go back to Safari, but Chrome is my second favourite browser. I wouldn’t suffer much to switch. Sure, there are some things that bug me now, but they would go away as I adjusted to doing things in different ways. There are some things I prefer in Chrome over Safari. And so, it might be only the familiarity that is bringing me back to Safari. Regardless, I’ll keep my eye on Chrome and might consider switching back depending on how it evolves.

To start, the tab interface is beautiful. Safari 4 beta switched to ‘tabs on top’, but it’s implementation was not great, and for the official release, it switched back to tabs on the bottom. Chrome’s implementation is more novel, practical, and elegant. The animation when one opens and closes tabs is fantastic.

Another novel aspect of the tab bar is the ability to ‘pin’ tabs. What this does is make the tab very small, and removes the close button.

This is slick for web apps that one wants to have open all the time. If one prefers using web apps to desktop apps, this might be reason enough to prefer Chrome to Safari. The concept makes sense because it separates the “browsing” tabs (the non-pinned ones) from the web apps. It might make sense to push this further though and allow to do things like hide the URL bar and the toolbar for pinned apps, even by default. If one is going to keep those tabs open, then buttons just clutter the interface. And as an example, if you want to keep gmail or facebook open in a pinned tab all the time, one probably doesn’t want to navigate off of it or ever use the back or forward buttons. Webapps often don’t particularly jive with those buttons anyways. It would be nice if the browser kept a list of ‘pinned apps’ somewhere, and whenever you browse to one of those sites, it automatically loaded it in a new pinned tab without the UI clutter. Sites could opt-in to the list as well. I think this will go a long way to allowing full-featured webapps, and I can see why it is natural for Chrome OS. Personally, I tend to use desktop apps wherever possible, so this isn’t too big a deal for me. For Safari users, one can sort of “emulate” this pinned tab behaviour by using Fluid.app which is a way of turning webapps into their own desktop or menubar app, so that you can make your own gmail app for example. This enables you to only use the main browser, for browsing rather than using apps.

The notification of site loading progress is way better in Safari. All that Chrome uses is a turning wheel. Safari uses the rightmost inch of the URL bar turning blue from left to right to indicate notification. Safari 3 was even better than Safari 4 as it would use the entire URL bar as a progress indicator, whereas Safari 4 uses only part of it, but at least there is some sense of progress. Constantly in Chrome, I find myself confused as to whether a page has reloaded, or whether it has started reloading. I look up, and all you see is a spinning wheel, and that doesn’t much help. There is no sense of progress. This small difference has a big effect, and is a major reason for me switching back. (Incidentally Safari on the iPad is going back to the old Safari 3 based full URL bar notification as shown in this video). Also, as pointed out by Gruber, Chrome puts the close buttons on the right side of each tab. This totally messed with me and I always shot for the left side. Granted, this is a habit that could be changed, every tab and window on OS X has the close button on the left, so if you differ from that you better have a good reason. I’m not sure why they don’t switch the favicon and the close button.

A more minor gripe in Chrome is that clicking on a bookmark folder opens it, and clicking it again, opens it again instead of going away. This is different than everything else in OS X, where any button that pulls down a menu, or the menubar itself, closes the menu when you click it a second time. This is the standard Cocoa behaviour. It’s a small thing, but I definitely kept messing up over and over.

Also, Safari has a lightweight rss reader built-in. Although it is not as feature rich as Google Reader, it supports encrypted feeds. This is probably not a big deal for most people, but for me, Safari is the only rss reader I use that supports them.

Safari’s implementation of “Top Sites” (similar to the new tab page in Chrome) is quite a bit better. The biggest problem with Chrome’s implementation is that it is not possible to choose and fix the position of a site at a position. So you can’t really customize it as a website launcher, and you lose spatial memory. Also you can only fit 8 pages on there versus a maximum of 24 on Safari, and so you can’t really use it as a fully featured launcher.

Overall, I think Chrome has potential, and I would use it ahead of Safari 2 say. And depending on how one uses their computer, one might prefer it. If I used Windows, I’d probably use it as my primary browser.

personal uses for iPad

There has been tons of discussion on the positives and negatives regarding the iPad. I personally love the idea of the device and think it has great potential. I won’t dwell here on any philosophical reasons why this is or is not “revolutionary”, or whether it will take off in the future. I personally enjoyed these pieces, and will just refer to them.

Instead, I want to present some potential uses of the iPad for myself, which I find promising. Of course, the real uses for the device come from the App Store. So in truth, I have no idea what this will ultimately be. Some of my ideas may not exist in app form until the iPad has been out for a year. But, below is a list of things I believe will be possible eventually, and if implemented well, may be preferable over other approaches. The list completely involves my own interests and leaves out other niche interests (such as art or music, which are there as well).

• travel laptop - It will be the only laptop I carry when I travel on vacations or conferences. I will likely bring the physical keyboard dock to use with it. There are some things that I need a Desktop/Laptop for, but I’ll be able to VNC (screen share) into my office computer if I really need those.

• work - I am going to likely carry it back and forth to work every day. My macbook pro is 6 lbs and I never bring it to work. But at 1.5 lbs, the iPad is much lighter and smaller, and I see tangible benefits to having it at work over my desktop, as discussed below.

• meetings - I’ve never brought laptops to meetings, but this is the perfect computer to jot down notes. Currently, my system is to grab a piece of scrap paper when I attend a meeting, jot down important notes, put it on my desk, then ignore it until it is no longer relevant. I figure digital is an improvement. I actually am pretty good already at prioritizing digital stuff, so I think I can do it (I have successfully used a job-specific variant of inbox zero for the last eight months and kept my work inbox under 10 messages the whole time).

• presenting - When I teach, I do everything “powerpoint style”. Every class I teach is in a different room, and “your results may vary” as to how the rooms and equipment are set up. I don’t use Keynote or Powerpoint, which are both relative torture if you use lots of math or algorithms. My slides are written in LaTeX (Beamer), which creates pdf files, and I present in Adobe Reader since that is what’s available. Since the iPad has its own version of Keynote, and there is going to be an “iPad to VGA” adapter, I’ll just use this combo instead of using different computers. It’s a portable presenter. But since I don’t use Keynote, I’ll pipe my LaTeX script into this handy PDF to Keynote converter and I’ll get a Keynote file as output (each pdf page becomes an image on each successive Keynote slide, so the text is non-editable. But that’s ok for me since I only want to present). Beyond being able to use my own computer for teaching, other advantages for presenters can be seen in Engadget’s hands-on with the iPad. When they are using Keynote, you can see that they stumble on to a simulated touch laser pointer, and even cooler, a pen to write on the slides. That makes the iPad effectively similar to a Smartboard. I can leave blanks on my slides and fill them in by hand as we go.

• sketch notes - It would be useful to jot handwritten notes quickly using a stylus. A company called Pogo sells stylii which work with the iPhone, and they have already confirmed they will work with the iPad. I definitely prefer doing research by handwriting on paper over typing. Likely, the “devil is in the details” as to how effective this will be on the iPad, but there are many advantages to writing on an electronic display over paper. So the question that remains is whether or not developers can minimize the disadvantages of electronic over paper. There are already a couple of apps for jotting handwritten notes for the iPhone, Finger Memo and iJot, but apps will really need to be better optimized for a larger screen. Here’s a little video of using iJot, where one can see the compromises that need to be made on the tiny screen. Make this letter paper sized and it’s a gigantic improvement.
  
  There are some aspects that would be an improvement on an iPad over regular pen on paper. For example, how many completed sketch notes do you have in your pocket right now? Probably none or maybe one. The iPad has effectively infinite storage back in time to allow you to look at previous notes. All your research notes are available whenever you need them (and of course an infinite supply of blank pages as well). Then, organizing and retrieval becomes important. So, it would be nice to be able to keep a database of previous notes, which can be tagged and catalogued. If you need to refer to previous notes while writing, you can display two notes side-by-side in landscape mode. Typically, a handwritten sheet shrunk down to this size it is still completely legible. In general, handwriting is far larger than typeset print. Or, it may be be possible to quickly print a previous page for reference. Although handwriting recognition could be built in, personally, I don’t care about this since I’d have the handwriting be a rough draft, and then type newer drafts. For me, this is just one of many stages of editing. Plus, it would be useful for math, since it is difficult to type first drafts of math research in LaTeX directly into a computer.

• reading papers - I read a draft of a thesis every week or two right now. There are some people that mark everything up in Adobe Pro or Preview by attaching typed notes to the screen. I personally don’t like staring at a monitor in a fixed position for 8 hours straight while reading. At least I can move around an iPad and change positions. What I’m hoping for is to be able to use a stylus with an app to annotate pdf’s with handwriting, and then save out the annotated version. An iPhone app that already does this is Aji Annotate pdfs.

• academic papers - I would like to be able to organize academic papers and keep them with me. The Papers app on the iPhone already exists for this purpose and syncs with the mac. I’d like to always have all my papers with me, be able to annotate them, and bookmark them in various ways. Indeed, being able to search for any text in any paper you’ve ever looked at is very powerful.

• reading books - There has already been a slew of textbook publishers sign on to put their books on the iPad. Inevitably, in a decade, we will only be buying books digitally and reading off a screen. But in the short term, the “devil is in the details” on this one as well. There are certainly big advantages of having a digital copy of books. Search is a huge one (it’s cool to search all your books all at once for one that contains some specific term), bookmarks, annotations, etc… But, I’m worried about the implementation. For example, with paper books, I can lend them out to students. And I do that. A lot. I have a feeling that there will be DRM on the books restricting what you can do with the files (likely we’ll be going through the same struggle as with music only to reach the same conclusion). If I can’t lend books, then the digital copy would have to be far, far cheaper than paper versions. If it’s DRM-free, then I’d pay a lot more to be able to lend. So, I’ll probably do this eventually, but I’m not optimistic in the short-term.

• reading news - I already love reading news on my iPhone. I prefer my rss reader on my phone over my laptop since the reading workflow is better. Once you enlarge the screen, this will be a fantastic news reader. I’m also interested in the new ways in which newspapers and magazines are going to present their content.

• watching TV and movies - One can rent iTunes movies directly on the iPad, which will be great on the road, in a hotel, etc. Of course, there is much hullabaloo about the device not supporting flash. But youtube and vimeo already play on the iPad, and I’m willing to bet that within a year, hulu will play on it as well. I think it will be great for TV and movies.

• VNC/remotes - I suspect that there will be lots of innovation on VNC (screen sharing) clients in the app store. There are already lots of clients for the iPhone, but on a larger screen, they become much more useable. If done properly, this can be the touch screen mac that many nerds wanted in the first place. Indeed, a VNC client doesn’t need to present the screen exactly as it appears on the server. There can be touch-optimized controls rendered on the client side to act as a hybrid remote. There can be additional gestures supported and more direct manipulation. In fact, by default, perhaps you can only control the server from local controls, but switch to “regular VNC” mode if there are not optimized local controls.
  
  For example, if I wanted to control iTunes, perhaps I could see my desktop, but only as a background image. I might have touch controls that is only rendered on the client which gives optimized controls and drop downs with all artists and albums. There are lots of VNC apps on the iPhone app store, but Remote Tap partially implements this idea by overlaying local controls overtop of being able to see the remote VNC Desktop (although I don’t think you can control the remote machine without the local controls).

I’m interested in seeing what developers cook up in the app store and hope that many of the ideas above pan out.

Can't resist one more Zeppelin song

Not many rock songs trick me into thinking I’ve got a butt.

toasters and computers

There has been speculation all over the internet about the soon to be released, Apple Tablet. I would like to toss my opinion about the tablet in the ring for discussion. I might just shave my hairy finger in order to use one (recall that I’m a monkey).

The two articles that have “nailed it” the most for me are this piece by Diaz at Gizmodo, and Gruber’s piece. Both articles are definitely worth a read.

Both discuss the need to move past traditional desktop OS paradigms. There are lots of benefits gained by not exposing the file system. Hierarchy is often conceptually difficult unless one is trained in dealing with it. Although in principal, “regular users” shouldn’t need to have to deal with files and folders on Windows, Linux and Mac OS X, unfortunately files are still the standard way of moving things around and organizing. Heck, you need to manually move files to install new applications on OS X. That’s a mess for newbies.

In ten years, I believe it will essentially be standard for computers to not have the file system exposed. Of course, there will be a pocket full of nerds who will still prefer access to the filesystem, just as there are still computer scientists who live at the command line, but they will be few and far between. To be honest, I think most nerds can survive quite well in sandboxed environments, even doing programming language related tasks. Stick your compiler in the sandbox and you can survive. And for the non-nerd tasks that remain for them, the simplicity that will be necessary for the non-nerds will improve their experience as well. Computation in a sandboxed environment is still universal computation. It’s just only universal in the sandbox.

The article by Diaz discusses so-called “appliance-based computing” conceptualized by Jef Raskin. The idea is that a computer could allow to use disjoint appliances, where the interfaces for each are completely adapted to that particular appliance. And indeed, a touch screen is the perfect mechanism to lay out an unlimited number of different appliances, which themselves only do something conceptually very simple. And one can switch appliances and have the interface adapt fluidly to the new appliance. And on the iPhone, the app store allows to have a massive number of different possible appliances. Desktop operating systems really don’t fulfil this objective with the physical keyboard and mouse, and with the exposed filesystem. In order to successfully apply appliance-based computing, it helps to have a touch screen, where the controls adapt to the appliance.

The consensus seems to be that Apple’s tablet will have a 10” touch screen. But what will the OS look like? What of the interface? On a phone, you mostly are “using” one app at a time, partially as a result of the small screen real estate. But when we scale the screen up, in principal, we could have multiple windows or apps that one could use at the same time. Is there a screen size where appliance based computation should switch from one full screen app, to multiple apps side by side simultaneously? I suspect that indeed, there is some size where it becomes ridiculous to run everything full size. Certainly, on a 23” screen, it is easy to fit two web browsing windows side by side, and I suspect, forcing everything to go full screen becomes harmful. So, likely there is some size where this should change. At 10”, I believe they will have found that it still has more pros than cons to run every app at full screen. This helps cement the “appliance-based” computation paradigm, and reduces complexity to the user. When you split the screen in half, there is a lot less that you can do, and user input becomes significantly more complicated as well.

One of the other benefits of appliance based computation on the iPhone is the reduction in complexity between “switching apps” and “quitting apps”. There is essentially no such difference in user action. You hit the home button, and start a new app. This is what you would want to happen if the OS was really just collecting together disjoint appliances. I suppose if the world was full of happy fairies, it would make sense that every appliance have the option of whether or not to continue to run all the time. But this simply does not scale from the perspective of engineering, even with more beefy hardware. And indeed, many users have 100 apps installed (on average, iPhone/iPod touch users have downloaded over 30 different third party applications). Having this number of apps is better for everyone, but having all these apps running all the time is absurd. Moreover, managing the distinction between quitting apps and hiding apps creates additional interface and conceptual complexity (we’re talking about grandma here).

Right now, the standard behaviour on the iPhone is that hitting the home button “almost always” quits the app. Well, that’s kind of messy, isn’t it? All third party apps, and most Apple installed apps quit the application when you hit the home button. There are a few background processes that do keep running though. For example, if you leave the iPod application, it does quit the iPod app (it really quits), but there is a small process that stays open in the background that doesn’t have the iPod interface loaded, but just plays the music, and displays a small menubar icon while other apps are run. There is also a little background app that stays open receiving messages from some Apple server, so that when apps quit, their servers can pass messages to Apple’s server and then on to your phone. This allows only one, very small background app to run instead of every app staying open to receive messages. So there are two little processes that stay open. There are a couple of easy ways to extend this concept without having every app be open.

• OS wide audio player - Right now, if a music playing app other than Apple’s iPod app quits, the music stops. I suspect that in the Tablet, and the next iPhone OS, third party apps will be able to tap into the background app that plays iPod music, and continue to play music in the background. Then, the apps themselves can be unloaded from memory, but it will appear as though it is still running (the iTunes store app seems to do exactly this on the iPhone, as you can play streaming podcasts over the net in the background with the same menubar icon as the iPod app).
• Location aware apps - For some apps, it makes sense to have events triggered based on your location. But it is also not nice to run your battery down and have the GPS running in the background. However, Apple knows where you are already, roughly anyways, with cell phone tower triangulation, and also with your nearby wifi routers. The cell phone triangulation is “always on”, and the router location is on, if it looks for wifi signals. So, it seems likely, that apps will be able to tap into both of these, on both the iPhone and the tablet. In the background, it will be able to tell within a large radius, where it is all the time, and every 15 minutes it will be able to tell within a smaller radius with wifi (I don’t think it will have the battery to sustain a constant wifi connection, but if it can, then that’s even better).
• There may be some other limited background processing that is allowed. For example, right now, if you get a phone call, your current app quits, and the phone app starts, and then when your phone call ends, it re-opens your previous app. The app itself usually tries to recover as gracefully as it can, after having it’s computation cut-off. In the future, if there is enough RAM, I think that apps will be able to “suspend” while tackling urgent processes, and then resume, without removing everything from RAM. This helps too.
• Time triggered apps - The clock app appears to be running in the background as well on the iPhone, but I suspect again, that the Clock app itself does quit, and there is something running in the background that triggers based on time (this is perhaps just launchd like on the Mac). I suspect that third party apps will be able to tap into this as well, and trigger events based on time. No extra third party apps need to be running then.

Are there any other classes of apps that would benefit from background processing? Maybe. But how much will they be benefiting? I don’t think much, as long as apps launch and quit very very quickly and sensibly (which gets much better with every hardware improvement). It will be hard to tell that they are essentially not running all the time. There is some nerdy things that I wouldn’t mind to do in the background, like heavy duty computation, but I think that will be reserved for much bigger screens with more horsepower. We don’t need to introduce that complexity for a screen 10” or smaller. Perhaps they will allow apps to suspend, and periodically wake up to do something, but that might be reserved for later.

Other than that, I believe the tablet will look largely like the iPhone OS. There will likely be many more supported gestures, which become easier to perform on a larger screen. There will obviously be the app store, which is crucial with appliance based computation. The input method is the big open question. I suspect it will be a large on-screen keyboard where you can touch type. I’m very doubtful there will be anything stylus-like that ships with it. It’s just too slow. There will be a third party market for them which will be important for artists, and specialists, but other than that, it will be all keyboard. Will there be a physical keyboard that one can use with it? I don’t know about this. It depends on how much Apple wants to encourage people to get used to typing on the screen. Sometimes forcing users to do something is better than giving them what they want.

Thus, in terms of physical predictions, I think it will look like a big iPhone. I think it will have a wifi only version like the iPod Touch, and a 3G+wifi version. I think the cell companies will sell tiered plans such as a “very lite” plan of 100MB/month for $15/month. This would be perfect for older folk who want a home computer with internet and could successfully serve as their only computer, and the only internet they’d need.

All in all, I think the tablet announcement will be very important as the first occurrence of large-screen appliance based computation has been brought to the market. As I suspect that this will become the standard in the years to come, it is a pretty big deal.

self-control

Just reading this is exhausting. I’m getting some iced cream.

9 things

Livescience has compiled a good list of the 9 things we learned about humans in 2009.

Of interest to me were that 44 different bacterial species live on your forearm. Take that Howard Hughes.

Also, late risers are better able to focus than early risers. I had only heard negative things about late risers before, so I’m happy to see this.

Zeppelin

“In My Time of Dying” - Bonzo is so relaxed when he plays. He never rushes anything, and actually he usually plays deliberately a little behind the beat to give the music a laid back feel. He’s the anti-“Stewart Copeland” of drummers. I go through a Zeppelin stage at least once a year.

Tony Williams

My favourite Miles Davis drummer would have to be Tony Williams. The guy plays a million miles per minute, is completely identifiable, and grooves too.

Check out his drum solo. I’m kind of immune to drum solos, but this one almost made me cry.