The Ubuntu Shopping Lens Debacle

It seems that Canonical are attempting to raise some money by inserting affiliate links into your Unity Dash when you type in a search term. Currently the main target for these links is Amazon, but I’m sure others will be added over time.

I’m not a fan of the Dash at all. I find it slow and cumbersome to operate with a mouse, and frustratingly ineffectual when I try to type a search term. Of course it doesn’t help that my 1980s vintage clacky keyboard lacks a Windows key, so all the “just hit the META key and type…” posts are useless to me. My dislike of the Dash aside, however, I think adding affiliate links to the desktop is a great idea. Free software often struggles to become financially self-sufficient, so more experimentation in this area is a good thing.

Except that the implementation in Ubuntu was an obvious PR disaster from the outset. Did they really think that people would be happy that the search terms they use when looking for files and applications on the local disk are also being sent to Amazon and other retailers as a matter of course? Mark Shuttleworth was quick to point out they’re not “putting ads” into Ubuntu and Jono Bacon posted that no user-identifying data is sent, but that just ignores the fact that there could be user-identifying data in the search terms themselves.

Mark Shuttleworth’s answer, early in the comments on his post, seems a little short-sighted to me:

…the Home lens of the Dash is “search everything”. If you want to search locally only, use the hotkey to specify the specific scope you want, like Super-A for apps, or Super-F for files.

So if I don’t want to send my search terms to Amazon, I have to search each separate scope individually (remembering a few more keyboard shortcuts along the way). That sucks.

It looks like there may be an option to disable these searches coming along as a result of the backlash. But even that solution is overkill – it becomes all-or-nothing, you either get affiliate links or you don’t. Wouldn’t it be better to just put a button at the bottom of the local search results to “Search our online shopping partners”? Perhaps beside a more general “Search online” button to perform a search using the user’s preferred search engine. No information would be sent out until one of the buttons was pressed, allowing the user to keep their local searches local, but making it trivially easy to perform affiliate and other searches when they want to share their search terms with the world.

I hope that Canonical, and other Free software vendors, can find sustainable ways to make money. But treating your users’ data – and that includes their local search terms – as anything less than confidential by default is not the way to go about it.

End of life for Ubuntu 10.10 (Maverick Meerkat)

Today I turned on my computers at home and at work to be presented with an alert telling me that the operating system on them, Ubuntu 10.10 “Maverick Meerkat”, has reached its end of life. This event has reminded me just how frustrated I am with some of the decisions Canonical has made in the transition to their Unity interface…

I’ve been a Ubuntu user since its first release, choosing to live life on the bleeding edge of the regular six-monthly releases, rather than relax with the LTS releases. Generally this approach worked well for me, until the problem trilogy of 10.04, 10.10 and 11.04.

  • 10.04, “Lucid Lynx”, was an LTS (long term support) release which used the old Gnome 2 interface. It’s still supported for another year, until April 2013.
  • 10.10, “Maverick Meerkat” was a normal six-monthly release which used the Gnome 2 interface. Support has just expired for this version.
  • 11.04, “Natty Narwhal” was a normal six-monthly release which used the Unity interface and is supported until October 2012.

Back in 2010 I went through my usual two upgrades, the first putting me on an LTS release, and the second taking me back off that track to the six-monthly cycle. I had expected to upgrade to 11.04, but then Unity happened.

I’ll admit that I’m one of the people who doesn’t like Unity. In its current form I find it far less efficient than the workflow I’d established with Gnome 2. It doesn’t work with my preferred choice of focus-follows-mouse. Most of all, it’s fundamentally broken with my multi-monitor setup (something that the developers are actively addressing, though, so I might be in luck with 12.04). That last point meant that upgrading to Unity just wasn’t an option, regardless of my opinions of it as a user interface.

That’s my real gripe. If Unity had first appeared in 10.10 then I would have found that it didn’t work, stuck with 10.04, and still have until 2013 for Unity to get fixed, or for me to find an alternative. So please, Canonical, if you decide to make any more significant changes that may stop people upgrading by failing to work with their not-very-exotic hardware, at least do it in the version immediately following an LTS release.

And with that news I shed a tear

As I come to the end of another day wrestling with code in an effort to get it to run in Internet Explorer 8, I find this post about some real world statistics from a large UK healthcare site.

Internet Explorer accounts for 55% of the numbers…

With 47% IE8 is the most used browser, followed by IE7 with 24% and IE6 with 19%. IE9 only accounts for 10%.

So 55% of visitors are using IE, and 90% of those are on an old version. If people really must continue to use Windows XP (and I appreciate that sometimes there isn’t a choice), then please at least switch to a different browser… you know, one from a vendor who hasn’t abandoned support for your platform like Microsoft has. Until you do, your just helping to hold the web back for everyone else.

Joe Consumer has the Same Unity Problems as Me

The recent Ubuntu Developer Summit included a session on Canonical’s user testing of Unity and Ubuntu. Have a read of the results over on OMG! Ubuntu! Here’s my favourite bit from the introduction:

Although a small number of testers are used in the qualitative testing, they are selected very specifically to best demonstrate the target audience of Ubuntu. They are a mix of Windows and Mac users who spend at least 10 hours a week on a computer. They tend to know how to download music, attach peripherals and other fault sedate computing affair [sic].

Perhaps this is one caveat in the data accrued thus far: the focus has been with “Joe Consumer” representatives and not those more advanced or all ready [sic] familiar with Linux.

So the testing isn’t aimed at more advanced users, or those already familiar with Linux. I’d count myself in both those groups. Yet interestingly the main issues they found with with Unity and the Dash were exactly the same as my complaints. Okay, my list also add a few extra techy issues, but for the main part my problems with Ubuntu 11.10 aren’t due to me being an ‘advanced user’ who is ‘already familiar with Linux’ – they’re the sort of problems that affect even Joe Consumer.

The article tries to put a positive spin on the results, but I can’t help feel that in producing a system which elicits similar complaints from both Joe Consumer and advanced users, Canonical may have managed the incredible task of uniting geeks and noobs into one coherent unified group, bound by the issues they face using Ubuntu these days.

Unity indeed!

Why I’ll be skipping Ubuntu 11.10

My normal desktop machine runs Ubuntu 10.10. I decided to skip 11.04 because the UI changes introduced were, quite simply, downright broken with my particular dual monitor setup. A few days ago version 11.10 (Oneiric Ocelot) was released, so I decided to give it a spin to see if it was worth upgrading to.

By “give it a spin”, I mean that I have run it in a virtual machine, and live from CD. I’m using the latter to write this post. I haven’t done a full install to my machine so it’s possible that some of the problems I’ve seen are a side-effect of running it “live”. But most of the problems I’ve faced are fundamental UI issues which would certainly affect a real installation as well.

I started out writing a wordy description of each of the issues I found, and how it affected me, but I realised I was going to be typing for a very long time. So I’ve decided to just list the obvious issues I’ve stumbled across so far as little more than bullet points (albeit wordy bullet points). I might expand on some of these in future posts.


General Issues and Annoyances

  • No Focus Follows Mouse option
  • No option to “shade” the window when double clicking on the title bar
  • No “Show Desktop” button for getting to files and folders on the desktop when you’ve got windows open and covering them
  • The Nautilus toolbar has been neutered to the point of just being a location bar with a search button. Any other options require a keyboard shortcut or a trek up to the menu at the top of the screen.
  • The four default workspaces are arranged as a 2×2 grid. I prefer a 1×4 linear arrangement so that I only have to bind keys for moving left and right, and don’t need to maintain a mental map of the workspaces locations. I couldn’t find a way to change this.


The Application Launcher

  • On my dual screen setup the launcher is always over to the far left of the leftmost monitor – which is a long way to go if I’m working at the right of the rightmost one. At least with 10.10 I was able to add extra launchers or even a whole application menu to the Gnome panel wherever it suited me (I’ve got two application menus, one on each monitor).
  • The conflation of launcher and task manager means that clicking on the launcher icon for an already running app brings that app to the front – it doesn’t launch a new instance. If you want to launch a second Nautilus window or another terminal – both perfectly reasonable requests – you have to middle-click on the launcher icon. But that’s not exactly discoverable; where is the context menu option to launch another instance?
  • I’ve got plenty of screen space; I don’t want the launcher to auto-hide, but it seems that I have no choice.
  • When testing 11.10 in Virtualbox with Ubuntu in a window, the auto-hiding of the launcher can make it very difficult to regain access to it when the mouse moves to the left of the window. It’s better than it was in 11.04, but could still more easily be solved by letting me turn off the auto-hide option. I suspect the same issue would cause problems with remote desktop use.


Problems with the Dash

  • The Dash application launcher appears to have four launcher icons with limited customisation options. Only one of those would I use regularly – why can’t I change the other three to launch whatever I want (e.g. replace “View Photos” with a launcher for Inkscape or Chromium – rather than just a different photo viewer)
  • The keyboard shortcut for opening the Dash isn’t obvious. If you press ALT on the Unity launcher bar, the other launchers get a number associated with them as a keyboard shortcut, but not the Dash. It turns out that it’s bound to the Windows key, but nothing tells you that, you just have to guess. I haven’t found a way to change this, which is a problem given the lack of a Windows key on my keyboard (1984 vintage, with buckling spring keys – far better than the squishy membranes that pass for a modern keyboard)
  • “More Apps” in the Dash gives an uncategorised list that’s essentially useless. You can apply a filter to limit by categories, but that entails an additional click for browsing through each category. At least with the old “Applications” menu, browsing through all the installed applications – by category – was simply a case of moving the mouse.
  • The Dash uses its own UI widgets including “window” buttons which don’t match your chosen theme, and a scrollbar which is only a couple of pixels wide, and sometimes doesn’t respond at all.
  • If you do apply a filter, then close the Filter panel, there’s no indication that the filter’s even applied, which can leave you wondering where all your apps have gone.
  • Do I really need to see “Apps available for download” all the time? With icons the same size as my installed apps? This just distracts me from the visual target I’m actually looking for. If you must include something like this, use much smaller icons aligned to the bottom of the window. Non-installed apps are not as important as my installed ones, don’t give them the same visual priority.
  • The Dash has room for at least three rows of icons. Why do I get one row of installed apps, one row of downloadable apps, and some blank space? Why not show me some more of my installed apps instead?


The Menu Bar

  • My screen is wide enough to easily accommodate both the title of an application and its menu bar – so why does the menu have to be positioned so far to the left that the app name has to fade out beneath it?
  • For that matter, why not show the menu all the time – faded a bit until you mouse over it, if necessary – so that I can see the destination for my mouse moves before I start moving.
  • Of course all this would be a non-issue if I were allowed to turn off the unified single menu and return to the days of in-window menus. Believe me, I’ve got enough screen space for that. But Canonical would prefer to save me a few vertical pixels in each window instead, whether I want it or not (FWIW I have the same issue with the single menu bar on a large-screen Mac; it works well on the smaller screens of old, but is far less efficient when your mouse has to perform a rodent marathon across a large desktop.


There are other issues, but already this is sounding like a whiney list of gripes, so I’ll stop there. The fact is that a lot of these issues don’t need to exist, if it wasn’t for the arrogance of Canonical’s design team and their vision for how a desktop should operate. I don’t want a return to the days of a thousand and one configuration options, but the choices they’ve made increasingly seem aimed at small screens, netbooks, phones and tablets. Are desktop users with large screens really so much of a dying breed that decades of user interface choices can be thrown out so easily?

And I’m aware that some of these issues can be “solved” via hidden configuration options or by downloading additional software. But if you focus on that, you’re missing the point entirely. It’s not that I want every possible configuration option exposed (that’s what drove me away from KDE in the past), it’s that I want at least enough flexibility to be able to set up my launchers and documents where I want them on my computer without having to resort to searching the web for solutions.

I’ve been using Ubuntu since the first release, but increasingly I think 10.10 will be my last. I’m prepared to hang on until 12.04 – the next long term release – to see if things improve, but in the meantime I might just go and grab some other live CDs to give Gnome 3.2, KDE, XFCE and some other options a try…

Will Lucid play havoc with the Dell Mini 10?

There has been a lot written about the change in position of the window controls in the alpha- and beta-releases of Ubuntu 10.04, Lucid Lynx. I haven’t really got much to add that’s not in this excellent post, and it’s already been discussed to death in this bug report. Basically the problem is that the window controls have been moved from their historical position on the right of the title-bar (where Windows also has them), over to the left:

The rationale behind the change – what there is of it – seems to be that the Ubuntu developers have some ideas for things that they want to put in the top-right of the window. I’ve always felt the window title bar to be a bit of a waste of screen space; I don’t want to see it removed entirely, but rather would like to see it gain some more useful functionality. For example, the window title on Apple’s Finder can also be used to navigate to parent directories, or as a drag-and-drop widget for the parent folder.

So I’m definitely not averse to them adding more power to the title bar, but why does that require the widow controls to move now? Why does it require them to move at all? Unless they’re going to magic some extra space from somewhere, I don’t see why the window controls can’t stay at the top-right, and the new widgets go to the top-left, or to the right of the title bar but just left of the window controls. There may be a very good reason why their new widgets should go to the top right, but with no indication as to what those new widgets might be, it’s a little hard to judge.

So what’s this got to do with the Dell Mini 10, specifically? It’s all about the trackpad.

I have a Mini 9, which I think is a great machine. It has a traditional old-fashioned trackpad, with a pair of buttons underneath it for left- and right-clicking. At the weekend I was working on a friend’s Mini 10, which has the buttons integrated into the trackpad itself. The result was that almost every time I tried to click, the mouse pointer would jump down the screen a little, often meaning I missed the target I was aiming for. Although my friend (who has owned the machine for a few months now) fared much better, even he had more mis-clicks than I would consider reasonable.

I’m not averse to integrated trackpad-and-button systems per se. The recent multi-touch trackpads on Mac laptops seem to work quite well in this regard – perhaps because the whole trackpad is a button, so there’s no need to move your finger away from the target to initiate a click. That’s not the case with the Mini 10: you always have to move your finger to the “button” area at the bottom of the trackpad, but doing so is liable to be interpreted as a desire to move the cursor. It also prevents the “move with the finger, click with the thumb” approach to trackpadding that I prefer. This isn’t the only machine I’ve ever used with such a troublesome trackpad, it just happens to be the most recent.

These jumpy mouse-clicks are problematic enough when window controls are at the right, but putting them on the left makes them a prime target for mis-clicks every time the user tries to open the Applications menu. Apart from the claim of mystery widgets in the future, I have yet to see a good reason for moving the window controls, while I’ve seen plenty of good reasons to keep them where they are. The Mini 10’s trackpad is just another one to add to a long list.

DVD Menu Overkill

Previously in My Green Life
‘I own a lot of DVD box sets…’
‘…tediously animated DVD menus…’
‘Worse still are those that drop you straight into a graphical “chapter selection” screen…’

Thank goodness I managed to jump out of the car when it was passing that handy boulder which hid my exit from the camera, just before the car sped off that nearby cliff.

Having escaped from certain death with little more than a cosmetic cut which will be gone by the next scene, it’s time to continue my diatribe against DVD menus. Last time it was the use of spoilerific images that got my back up, but this time I want to talk about the many obstacles that are put between me and the programme I want to watch.

Generally DVDs fall into two categories: features (films, long documentaries and anything else which fills most of the disk with a single programme), and series (several programmes on a disk). In the former case I want to start the feature playing with as little delay and hassle as possible, but I’m happy to accept a few more steps to access the extra features on the disk. In the latter case I usually want to watch a single episode at a time – so I’d like to select and play an individual episode with ease, but don’t mind a few more steps to access the extra features associated with either the episode or the disk as a whole.

In both cases my efforts to get a programme playing quickly are usually thwarted by several interstitial elements which serve little purpose other than to get between me and the content I want to see:

  • Trailers for other programmes. These are usually skippable, but not always. Stick them on as an extra – not something that plays by default. They’re particularly annoying when you forget they’re there, put a DVD in the drive and go out to make a drink expecting to return to the screen showing the DVD’s main menu.
  • The anti-piracy advert. A stock piece of propaganda telling me that piracy is bad and I shouldn’t be a part of it. Often it’s not possible to skip this, even though I’ve seen it dozens of times before. The irony of having to sit through this every time I want to watch a legitimately bought DVD is not lost on me. This is the sort of thing that makes me tempted to download torrents of my DVDs just so that I don’t have to put up with this interfering, accusational, nannying waste of my time.
  • The copyright notice. This is the page of text which tells me that I can’t play this DVD in a hospital or an oil rig. I assume this also prevents me from playing it in a hospital that’s on an oil rig. Do I really need to see this every time I play a DVD? And if you’ve put it on the first DVD in a box-set, do I really need to watch it on every subsequent DVD in the set as well?
  • Das kopyrighten notis. Or something like that. From the general formatting I think this is the same as the previous item, except I get to see it in a different language. Or more likely in several different languages – none of which I understand. Once again there’s often no way to skip these, leaving me to waste another five minutes trying to find the most amusing scandinavian word I can out of the jibberish characters before me.
  • Publisher/Producer/Studio/Tealady Idents. These are the little animations that tell you you’re watching something from Fox, or Dreamworks, or the BBC, or whomever. Then there will be another ident from some other company who had something to do with the DVD, though as you’ve never heard of them before it’s hard to tell what it was that they did. And possibly a third or fourth for good measure.

    I think these are here as advertising – or at least I can’t think of any other reason why they should be put in such a prominent place. If they are advertising, then they don’t work. I can honestly say that I’ve never met anyone who buys their DVDs based on studio awareness. Good reviews or previous familiarity with a film, series, actor or director, maybe – but never “I really enjoyed that last 20th Century Fox movie I saw, so I’ll buy another one from them even if I’ve never heard of it before”.

  • Don’t forget the THX ident that might be on the disk. Great for showing off your surround sound system. But you know what else is great for that? Actually playing the bloody film!
  • There! I’m finally at the DVD menu. Oh, hold on, I’m not really. Yes, it’s there, but some pillock decided it would be a great idea to show me their animation skills for a minute or two before I can even interact with the damned thing.

And so, after somewhere between two and twenty minutes (for the full trailers + anti-piracy + copyright + kopyrighten + idents + animated menu experience) you finally get to a point where you can actually choose what to watch. Hurrah! If you’re very lucky the selected item will already be “Play main feature”, and you can just click the OK button to get started. Of course there are likely to be more idents before the film actually starts, but at least you’re onto the home straight.

For series, however, you’ll often find that the default option is to “Play all episodes”. Really? What genius thought that one up? This episodic programme, which was originally broadcast in weekly segments, should actually be watched in chunks of four episodes at a time? I’m far more likely to watch one episode per night than four episodes back-to-back. Why not start with Episode 1 selected but design the menu so that “Play All” is just an arrow-press away?

At this point there’s often another pointless animation before I land on another page whose sole purpose is to let me get to the extras for that specific episode. Couldn’t they have been relegated to an “Extras” option on the main screen? Or shown as a second item alongside the main menu entry? Most of the time I just want to watch an episode with no director’s commentary and no original storyboard pictures, so make that the default.

So there you have it. In my ideal world all the trailers are sent to the “Extras” screen. The anti-piracy and copyright notices are relegated to a line and button at the bottom of the menu which says something along the lines of: “This product is protected by copyright. Select this here menu item to see more details”. The idents can go away entirely, or be moved to the copyright information page. And any animation on the menus should not get in the way of me being able to actually use the thing as quickly as possible.

The likelihood of any such changes are remote at best. In the meantime trying to watch a single 30 or 40 minute episode of a programme can easily take five minutes longer than needed – over 10% of the episode time, and that’s on one of the less obnoxious DVDs! That torrenting option starts looking better and better if it lets me get straight to the action.


No, it’s not a misspelling of “Nomination”. It’s just me wondering if the English language needs a new word.

It’s common to see the word “decimate” being used in reference to a near-complete destruction. Dictionary pedants are quick to point out that decimation actually refers to a reduction by 10% not to 10%. Clearly there’s a need for a word to describe a 90% reduction, or similar – so I’m proposing “nonimate”.

This is a tongue-in-cheek suggestion. I’m not an etymologist, so it’s probably incorrectly formed and means “by a ninth” rather than “by nine-tenths”, but then “nonodecimation” isn’t likely to catch on. The question is whether you prefer mis-use of an existing word, or misconstruction of a new one.

Anyway, I’m off to steal some apostrophes from greengrocers to stick on road signs.

Museum? “Exhibit” or “Gallery” would be more honest

A few years ago I visited the Design Museum, just down from Tower Bridge in London. It was, as you might expect, a museum dedicated to design. It dealt largely with product design, though there was some content about architectural and graphic design, too.

As you might expect of a museum largely dedicated to product design, there were lots of products which were considered iconic or interesting in their design. I can’t remember the specifics (it was many years ago), but I do recall an original Apple iMac, a Dyson vacuum cleaner, a bit about the Mini, and so on. What I particularly liked was that the chairs that were dotted around for visitors to sit on each also had a descriptive tag tied to it – they were part of the exhibit too.

Fast forward to last week. We went to London for the day, and decided to visit the Design Museum – largely based on those memories of my previous visit. What a let down! It cost £8.50 each to get in (so £17 for the two of us), and the “museum” part that I’d so enjoyed was gone entirely. Instead there was one large exhibit about design in London through the ages, which was interesting but very text-heavy, and a couple of smaller exhibitions: one showing daring architectural models (though only a couple of the designs actually seemed to have been built), and one showing the work of Javier Mariscal.

The latter, however, looked largely like a super-fan’s bedroom. Just walls of merchandise and comic books, and hundreds of drawings hanging from the ceiling of the entrance. There was little explanation or description, making it hard to follow any specific themes or development in his work. One wall, in particular, was just filled with hundreds of products themed around the Cobi mascot he designed for the 1992 Barcelona Olympic Games. That was it: just a wall of products all themed with the same mascot.

We didn’t spend as long as we perhaps should have in the first exhibit, because we thought we would be eating into time for visiting the “real” part of the museum – but we didn’t exactly rush it either. We spent probably about 45 minutes in there – then found the other two exhibits to be so poor that we had finished the whole “museum” in less than an hour. Definitely not worth £17!

Don’t let this post stop you from visiting the Design Museum if you really want to… just make sure you know exactly what it is you’re going to see, so that you can decide if it’s worth the price. Based on my recent visit I think the name is misleading: I wouldn’t really consider it to be a “museum” anymore, at least not by most people’s common definition. Perhaps it’s time to rename it to the “Design Gallery” or “Design Exhibit”.

Today’s Life Lesson

If you find a Wagon Wheel at the back of your office drawer, with a Best Before date of 2003, despite it appearing to be perfectly fresh, under no circumstances should you actually try biting into it.

Not only was it tooth-breakingly hard, but it tasted vile.