Getting it not-quite-right

I always think it's fun when you can catch those nearly-there-but-just-not-quite-right ideas which always look so funny ten years down the line, right when they actually happen.

Motorola's Lapdock is an absolutely textbook example of the genre.

You can kind of see how they got there from here, after all. "Man," they thought, "it's a pain when you're out working, and you get something on your cellphone that you'd like to look at on your laptop, and you pull out your laptop, and all the stuff from your cellphone just isn't there. You have to send yourself an email or re-type the link or whatever. That's stupid!"

They're right. It is. Things are so much more convenient if you have the same information accessible on your phone and your laptop (and, the jump they did not make, everything else too).

Then they thought "hey, what if we just made your phone the canonical repository of all that information, and turned all your other devices into dumb shells that do nothing until you plug the Phone Brain into them?"

Well, congratulations: you just hit on exactly the wrong solution to the problem.

No-one wants the Nerve Centre of all their information to be their cellphone. That's just silly. You don't want a 'laptop' whose power is artificially limited by the battery and heat envelope of an itsy-bitsy cellphone frame. You don't want a 'desktop' you have to plug your phone into for it to be any use. You don't want to have to use a headset to take calls when your phone is powering your laptop shell. You don't want to trust all your vital data to a $2 memory chip inside your phone and have it locked into that phone so things get very awkward when you want to upgrade. None of those things are good things.

It gets outright sad to see Motorola continue to plug away at this loser of a strategy when it's painfully obvious that just about everyone else managed to arrive at the right answer: use a framework which synchronizes all that data across all your devices via this crazy thing called 'packet data transfer'. You know, the Internet. The Cloud. And junk. Whether it's iCloud or Firefox Sync or Google calendar/contacts/Chrome sync or whatever the hell Microsoft calls their thing, everyone else figured it out. Motorola, for some bizarre reason, continues to doggedly push at their approach which just about answers the original need but provides nothing but limitations in comparison to the other way of fixing the problem...

Test Day time again: Eclipse Fedora packager stuff

It's Test Day time again tomorrow (2011-10-13)! This week's Test Day is a bit of a special interest one: it's mostly going to be interesting to Fedora packagers. We'll be testing the Fedora Packager for Eclipse plugin, an extensive plugin for the Eclipse IDE which turns Eclipse into an ideal environment for maintaining Fedora packages: it integrates with the Fedora git repos and the fedpkg tool to make it super-easy to branch, modify, and test build Fedora packages. So if you're a Fedora packager or you're learning about packaging, this could be a great event for you. If you haven't tried this Eclipse method before it's great fun to give it a shot - even if you don't wind up sticking with it you'll probably have a good time, maybe learn some tricks, and help test the environment for those who do use it full time.

As always the event will be happening on Freenode IRC in the #fedora-test-day channel, and you can use WebIRC if you're not a regular IRC user. All the testing instructions are on the Test Day page. Please come along and help the team test and improve this great plugin!

Oh snap

I'm generally happier working on projects which are small and sort of grassroots-y and generally held together with duct tape, which is why if you ask anyone who works with me at RH they'll tell you I recoil instinctively whenever anyone asks me anything about RHEL (which I've still never actually run). But this is a fascinating dispatch from the Really Freaking Huge Code Systems side of the fence, talking about Google and Amazon architecture. Also has the best tech zinger of the decade:

"But I'll argue that Accessibility is actually more important than Security because dialing Accessibility to zero means you have no product at all, whereas dialing Security to zero can still get you a reasonably successful product such as the Playstation Network."

Oh, owch.

Thanks to HP and Canonical Simple Scan team

I'd just like to say a huge thanks to HP and the folks at Canonical who work on Simple Scan - at long last, HP fixed up the ADF support for my 1212nf multifunction printer in hplip, and Simple Scan has completely awesome multiple document scanning / saving capabilities. So I can just load up a stack of pages into the ADF, run Simple Scan, and very easily turn them into a set of properly-sized PNGs on my NAS drive. This is going to help me clean up my filing cabinet no end. You folks rock! Thank you!


If you ever find yourself looking at the four or five spam mails you get daily and wondering if SpamAssassin actually does anything any more, try rebooting your mail server and forgetting to start the SA service. When you wake up the next morning to 56 spam mails, you'll have your answer...

Thinking about Steve Jobs

When I first heard that Steve Jobs was dead, I shrugged and carried on working. I never met him or interacted with him in any way, and he had little impact on my life; it's sad that he's dead, in the general way that it's sad when anyone dies, but I have no context for any deeper feeling. Thousands of people die every day and we pay it little mind.

Since then, though, I've been reading all the responses to the news, and now I find I feel a large sense of frustration at the way his life has been interpreted in many quarters.

I read a particularly galling article just now in one of my least favourite papers, the Globe & Mail, ostensibly directed at the Occupy Wall Street protesters but really directed at the sense of self-satisfaction of its readers, which crystallized everything that to me seems wrong about much of the reaction to Jobs' death.

The G&M's theory is that everything Jobs achieved, he owed to Wall Street; that all his great achievements could never have happened without the modern American capitalist system embodied by Wall Street. No Wall Street, no Jobs, no Apple.

I think it's interesting to look at this claim in the context of the four, to me, most significant phases of Jobs' career Apple Phase I, NExT, Pixar, and Apple Phase II.

I'd say that all of Jobs' most significant achievements came in the first three of these. Apple Phase I was a true early Silicon Valley company: it was not driven by conventional capitalism at all, it was driven by a couple of kids with an innate interest in early personal computer hardware screwing around in a garage. It doesn't seem to have been motivated by any kind of desire to build a world-straddling corporate behemoth; it was motivated by a desire to build cool machines that did interesting stuff. And that's what Apple I built: cool computers with which people did deeply interesting stuff. They learned to program, they improved the efficiency of all kinds of processes in all kinds of worlds - big business, yes, but academia and small business and the home and all the other sectors of human activity. Early Apple computers, like all other early personal computers, were fundamentally rough-and-ready, open, learning environments. You poked around and figured out how stuff worked. You traded ideas with friends. You thought of new ways to do things. A couple of Apple ][s were part of the whole menagerie of computers my family had at home when I was young, on which my Dad, myself and my sisters played - not just in the sense of 'playing games' but in the sense of 'playing around', which is a much more interesting and useful one.

NExT was very similar to Apple Phase I: it was still fundamentally a company made up of the kinds of people who screwed around with Apple ][s when they were kids. It built interesting and hackable computers and software with which people in all the same sectors did equally interesting and productive stuff. As everyone involved had grown up a bit and there was more money sloshing around in the industry it looked more like a conventional corporation than did Apple Phase 1, but it was still at heart a geeks' company, like Apple or the early HP or many other such companies. Geeks will be geeks, whether they're being paid for it or not. Money might help them do things a bit faster, and few geeks say no to large quantities of it, but it's not the point.

Pixar is one of the most old-fashioned businesses in the world: it's a company of actors and writers. They tell stories, just as people have told stories for years and years. Since Disney bought the company the money side has come into play with crass cash-ins like Cars 2, but all the really great Pixar films are fundamentally combinations of great technology with great stories: that's why they're great, that's why we love and remember them. People have been writing, remembering and telling each other great stories since before we had any kind of economy at all, never mind a post-Enlightenment capitalist one. It's hard to believe the great stories that the wonderfully talented storytellers at Pixar have produced would not have been told, loved, and remembered in some form or another no matter when, where or under what kind of economic systems those storytellers had been born.

Which brings us to Apple Phase 2. Finally we have a part of Jobs' career which, as the G&M argued, could likely not have happened without modern capitalism. The current incarnation of Apple is not an engineers' company like old Apple was. It's not really driven by a desire to do cool and interesting things with personal computers. It's driven by a simple desire to make an awful lot of money, to be the world-striding corporate behemoth Apple Phase 1 never was. It's very good at that.

What does Apple Phase 2 produce? It produces, essentially, fripperies. Polished, shiny, well-designed, addictive, yet ultimately superfluous toys. Apple's personal computer side is no longer the heart of the business, and no longer aims so hard at the tinkerers or creative people it used to. (It's still hanging on to a lot of those people because the competition is so poor and Apple's legacy is so strong, but I doubt anyone could argue it's at the heart of what Apple does any more). The heart of Apple's business is now consumer electronics, and despite the vaguely aspirational nature of its advertising, those consumer electronics are fundamentally toys. They are dedicated to no higher goal than to allow us to pass dead time elegantly. The iPhone works as a phone, sure, but then so do any number of other products made since the 1980s. If the iPhone hadn't existed our practical ability to communicate with each other wouldn't be affected one whit. What people really do with their iPhones, iPods and iPads is to play - and not in the sense of playing around, really, but just playing.

There's nothing wrong with unproductive play. All of us do it at some time or another. No-one can be 'on' all day every day; we get fatigued and just want to sit down and do something relaxing for a bit. But the key point here is that the forms of unproductive play just really don't matter all that much. Given the choice between listening to some music on an iPod when you've got ten minutes to wait for a bus and no-one else is around and just sitting there staring at a bush, you'll probably go for the iPod, sure. But people have been sitting around waiting for things to happen for millennia, and we've always come up with some way to distract ourselves; from reading a book to whistling to adding up numbers in our heads. It's just not important. No-one is going to look back on their life and see the ten minutes they spent listening to Lady Gaga while they waited for the number 52 to Darlington as a transcendent moment in their life. It was just dead time that they filled.

I don't think there's anything terribly evil about the products Apple Phase 2 makes, specifically. If there's any evil to them at all it's a symptom not a cause: a symptom of societies that are probably tilted somewhat too far towards making and buying utterly unnecessary fripperies. It's not worth getting all hot under the collar about how closed down and un-hackable modern Apple hardware is, but it is worth noting those attributes: they're simply a consequence of the purpose of modern Apple hardware, which is not for you to learn how electronics work or learn how to write programs or come up with interesting new ways of doing things (or even interesting new things), but simply for you to play around. They're toys. No-one ever got hot under the collar because their Barbie Doll wasn't fully user-serviceable; it's a toy. If you want a toy, by all means, buy one. If you don't want a toy, buy something else. (Again, Apple computers are still quite usable as computers, and I understand those who find Apple computers meet their practical computing needs, but throughout Apple Phase 2, they've only ever gone in a more toy-like direction, never back the other way.)

So really, the only phase of Jobs' career to which modern capitalism was essential was the last and most superficial one. All the truly significant things Jobs did and achieved were not essentially dependent on the capitalist economy at all. They existed within it; certain good features of capitalism may have contributed to their development. But they were all manifestations of things that have happened for millennia, under any conceivable economic system. If the iPhone and iPod and iPad were erased from history tomorrow, maybe we'd all use slightly clunkier cellphones, and use them a bit less. But would human experience be fundamentally altered? No. If Apple Phase 1 and NExT hadn't existed, or rather if the whole early and middle Silicon Valley scenes of which it was a fundamental part hadn't existed, the worlds of computing and even science might be noticeably retarded in development in comparison to where they really are. If Pixar hadn't existed, the beautiful stories it told might have been lost to millions of people; the inspiring and uplifting effects of great stories are significant forces in peoples lives, things you remember to the end. Toys, usually, are not.

I'm not anti-capitalist in particular. I suspect it may be to economic systems what Churchill described democracy as being to political systems - the worst choice apart from all the others. But it does frustrate me to see unthinking or even cynical capitalism-worship, this idea that the greatest thing Steve Jobs ever did was build some shiny toys that made some modern-day Apple shareholders very rich. It was not.

(The other thing that annoys me about much of the response to Jobs' passing is the way it contributes to the largely mythical concept of individual exceptionalism, but I'm sure I'll get to that another time. :>)


I really, really wish I'd known about the 'mapchars' option to mount.cifs before I rsynced my entire music collection to a CIFS share (when I switched from storing it on my HTPC to storing it on a NAS box).

Turns out every track with a question mark in its name went missing during that process, as you can't create files with ? in their name on a CIFS share unless you mount it with 'mapchars'.

Now I get to go through and find and re-rip all the missing tracks. Happily Picard can help a lot with identifying albums with tracks missing, but it's still going to be a pain in the ass...

EDIT: Top tip: if you find yourself in this situation, luckily, it turns out rsync is smart. It doesn't skip the files, it hides them. So "02 - Does He Love You?.flac" becomes ".02 - Does He Love You" . Once you fix your mount options you can just rename the files as appropriate. All together now: thanks, daddy rsync! To conveniently find all files that were so renamed, do:

find -name ".[0-1]*"

(why don't I mount the share with NFS? The DNS-323 seems to have a bug where if you mount a share on it via NFS, then leave that share idle for a while, then try to access it, the box hangs.)

Printing Test Day this Thursday (oh, and some Beta thing came out today)

So we had this little Beta release today, you may have heard about it. Go grab the Beta, have fun - but be warned, it's a fairly Beta-y beta. It's fine for a Beta, but it's not like some other recent Betas which have been near production quality. Do read the common bugs page, and stay tuned to this blog or the Planet or the test list or the forums to keep an eye on future developments. Forewarned is forearmed!

Aside from that, Thursday - 2011-10-06 - is Test Day time again, and since the Beta is done, I've actually got around to letting you know about it a bit in advance this time. Specifically, it's Printing Test Day, which will as always be efficiently and effectively run by the printing team (aka Tim Waugh). As always, if you have Fedora and you have a printer, you qualify for great discounts - wait, sorry, you qualify to come along and work for us for free, isn't that just as good?

Grab your printer, grab a Fedora 16 live image (there's one on the Test Day page), run through the tests listed on the page, and add your results - what could be easier? You don't need a permanent installation of Fedora or a Fedora account. If you'd like to chat with Tim and the rest of the crew during the event, just join #fedora-test-day on Freenode IRC - you can use WebIRC if you're not familiar with IRC. Please come along and help us ensure Fedora 16's printing support is better than ever!

Disappearing add-ons in Firefox

Well, this is messy.

Mozilla managed to release Firefox 7.0 with a large bug which means that any add-on for which an update is pending when you install the Firefox 7 update will apparently disappear after the update.

We managed to release the update to Fedora 15, thanks to positive feedback, even though I tried to alert release engineering to except the update from the stable push after Oliver Henshaw informed me of the bug (thanks, Oliver). Oliver and I added negative feedback to the report, but due to the way Bodhi is currently set up, this does not result in the update submission being cancelled; we've wanted to make this work better for a while, but with the current Bodhi feedback setup it's quite difficult to make it behave 'correctly' in all circumstances as the only variables we can go on are 'proventester, registered user or anonymous feedback', 'time of feedback', and 'positive or negative'. Bodhi 2 will introduce different types of feedback, so we should be able to introduce some kind of 'parachute clause' like 'I have tested this already-approved-and-pending update and it contains a huge bug, un-submit it now!' which could be triggered by one or two proven testers, say.

Anyway, the upshot is this bug is out there and probably affecting Fedora users. It bit me - my Noscript has disappeared. It's actually quite likely to happen, as add-on authors often push updates the day before or the day of a Firefox release.

Now thanks a particularly bone-headed policy on the part of Mozilla, you might find some trouble getting out of this situation - I did. Mozilla published a workaround for the issue. However, you can't find out what it was, because when they released Firefox 7.0.1 (which fixes the problem), they took down the workaround text and replaced it with text which just says 'update to 7.0.1', as you can see on the page (unless they've now changed it back in response to my nagging). It can be very frustrating that Mozilla consistently refuses to acknowledge users who don't install Firefox direct from upstream, but use distribution packages. Folks, just because you've released 7.0.1, does not mean all your downstream distributors have. Removing the workaround text is just plain rude and unhelpful.

Anyway, I found another workaround mentioned in passing in the bug report, which works for me at least. You can find it here. I don't know its provenance or what it does - well, it clearly comes from some guy at Mozilla called Ed Townsend, I guess. But that's about all I got. Try it, or bug Mozilla to put the workaround back.

ABRT Test Day today/tomorrow - 2011-09-26!

Yet another late Test Day announcement: today/tomorrow, 2011-09-26, is ABRT Test Day, particularly focusing on ABRT's new libreport integration. As always, you can join #fedora-test-day on Freenode IRC to chat with developers and QA team members to get help with the testing and any issues you encounter, and if you're not familiar with using IRC, you can join in with WebIRC. All the test instructions are on the Wiki page, and it's very easy to test ABRT. Please come along and help out if you can!