Heading to FUDCon Lawrence

Along with half the rest of the FedoraVerse, I'm heading to FUDCon Lawrence...in about six hours, in fact. Zoiks. I have vague plans to do a barcamp talk on the new Anaconda and a hackfest on i18n stuff, though that may well change. I have much more definite plans to drink, bowl and play poker, very likely in that order. See you all there!

edit: Note - if, at FUDCon, I keep calling you 'man', this is because I have forgotten who you are. If you find yourself in this situation, we can go one of two ways. Feel free to discreetly remind me who you are yourself. Otherwise, I will wait until I hear someone else call you by your name. Either way, once I've been reminded who you are, I will call you by your name loudly and enthusiastically at every possible opportunity for the rest of the weekend. This is both to foster the illusion that I knew who you were all along, and - with any luck - to ensure that I at least don't forget who you are again for the rest of the weekend.

The more you know!

Fedora 18 released (at last)

Yes, as you may have heard out there on the interwebz, Fedora 18 is now available!

Here on Planet Fedora said announcement has been accompanied by a chorus of exhalations and mutual backslapping - it's been a hard release to get done and a hell of a lot of development and release management work went into it. But that's all inside baseball to you, Linux user with a life to live...so this post is me looking at F18 and pretending I didn't know how much hard work it took to get it to where it is. I hope no-one's being overly optimistic about the delays meaning F18 is an incredibly awesome, super-perfect release: the delays were necessary to get the huge amount of development work in F18 finished to our baseline quality standards, not to polish everything to a perfect shine. The delays mean F18 isn't going out horribly, horribly busted, they don't mean it's the MOST SOLID FEDORA RELEASE EVER.


By all means, check it out. F18 in general seems to be a pretty solid release. The delays were almost entirely related to the new installer and upgrader: outside of those two areas, the delays really have made the release more solid. A running F18 system is a pretty nice experience in most ways. One thing that might trip a few people up is the migration of i18n, keyboard layout and hostname configuration - short version, system-config-keyboard doesn't work any more, use 'localectl' (or GNOME control center) to configure locale and keyboard layout, use 'hostnamectl' to set hostname.

The new installer and upgrader are a really decent first cut. But with my Official Pessimist hat on, please remember they're a first cut! I really like the new installer interface in most ways; it's a big improvement on the old one. It's pretty inevitable, though, that an installer interface we've had one release to tweak will have some rough edges compared to one we had over 18 releases to tweak. It's also inevitable it'll be somewhat more buggy. The most complex bit of an installer, in both design and functionality terms, is the storage management stuff, and while we squelched a whole bunch of issues in the storage handling of F18 between Alpha and Final, it still has some bugs we'd have liked to get out if we'd had infinite time, and some UI design issues that we're aiming to improve in F19.

So please do try out F18, but maybe don't let it loose on a system which contains a lot of important data on a complex partition layout without backing up first, and please do read the common bugs page and the overview of the new installer, which has some useful notes on the overall design, the design of the storage section, and some limitations that are commonly encountered (most of which are just the result of running out of time to get them written, and they'll be fixed in F19). And please bear in mind that this is the very first build of the new installer, it's only going to get better from here! I think for a v1.0 of something as complex as an OS installer it's a really great achievement for the time we had to work in, but please keep in mind it's a v1.0, and the F17 installer was somewhere around v25.0.

The documentation team did an absolutely amazing and sadly unheralded job of somehow updating the entire Fedora 18 installation guide - all 322 pages of it, holy cookie - for the completely revised F18 interface. While we were re-designing it on them every week. I have no idea how they managed that, but they did. Keeping a copy of the install guide handy while you install F18 is probably a really good idea; it's a great reference, and will help you out if you're not quite sure how any bit works.

Come work for us: Red Hat is looking for a(nother) Fedora QA community person

Are you interested in working for Red Hat? Got experience working in F/OSS communities - particularly the Fedora community? Do you have or can you convincingly fake some kind of understanding of or at least interest in QA, or at least QA as practiced by Fedora (which bears little resemblance to anything you can study in QA School, so don't worry if you don't have the qualifications)? Then you might be interested in applying for this job!

We're hiring a Fedora QA Community Co-ordinator, who will be helping us with maintaining and growing the Fedora QA community. This will involve getting involved in our processes and making sure they're community friendly, helping to organize and publicise community-focused events like Test Days, and lots of things along those lines. I did a lot of it when I first came on board - my official title is still QA Community Manager, though I've been doing more of a 'team lead' job lately - and I'd like to get back to doing more of it quite soon, but we'd also like to have another person to help with the load.

The job is officially based in Brno in the Czech Republic (where RH has a large, growing and by all accounts really cool, though I haven't been there yet, office) so ideally you should be based there or able to re-locate there, but for a really great candidate, working from another office or remotely may be possible - it'd be harder for us to organize, but we're willing to make the effort if we find a really great candidate.

RH is of course an equal opportunity employer!

A plea to regular Linux journalists: include an email address

I'm noticing a worrying/annoying trend in Linux journalism lately: quite a lot of authors don't have an easily-discoverable email address any more. If their articles include an email address at all, it's one that obviously gets routed through a feedback desk at the publisher of the particular article in question, which is often useless when it comes to freelance content.

It used to be far more common for mostly-freelance, specialist tech journalists to include an email address in their articles. This is invaluable.

The 'modern alternatives' appear to be comment sections and Twitter. Neither of these is acceptable. The 140 character limit on Twitter inherently precludes meaningful feedback or conversation. I cannot explain the subtleties of your misunderstanding of secure boot in 140 characters, tech journalist. Please don't expect me to try. Most journalists don't read their own comment sections, especially when they're freelance, and posting in comments is inherently more public than private email feedback: there's stuff I can say to a journalist in a private off-the-record email that I wouldn't necessarily want to put out in a public comment section. (Not to mention that in these degenerate times, by posting a comment, I'm usually assigning copyright to that comment to some sort of outsourced third-party comment system provider...)

If you want feedback from the people who work in the sector you're writing about, folks, please include an email address. Accept no substitutes.

KDE 4.9 Test Day today!

Anything the Desktop team can do, the KDE team can do as well, it seems: following on the heels of last week's GNOME Test Day, today is KDE Test Day! If KDE is your desktop of choice, today's the day you can join the KDE packagers to test out KDE in the upcoming Fedora 18 release and make sure it's working well and smoothly polished.

Full test instructions are on the page. You can test with a live image, which you can download from the page - you don't need to install Fedora 18 at all. You could also do much of the testing in a virtual machine.

We also encourage testers to join the #fedora-test-day channel on Freenode IRC: members of the Fedora QA team and KDE team will be around to help you with the testing process and also to take a look at any bugs that you might find. If you’re not a regular IRC user, you can find instructions here, or you can simply click here to join the channel (chat room) through a Web front end.

Even if you’re not a Fedora user, please consider joining in - as with GNOME, Fedora's KDE is a clean build from upstream and fixes for any bugs found at the Test Day will land in upstream KDE for all distributions that use KDE to benefit from. Please do come by if you have time! Thanks.

GNOME 3.6 Test Day today!

It's that Test Day time again, folks! Depending on where you are, tomorrow or today - Thursday 2012-11-08 - is GNOME 3.6 Test Day. We'll be testing various areas of GNOME to ensure the desktop is working smoothly for the upcoming Fedora 18 release. If you have some time to drop by and help GNOME continue to get better, please do!

Full test instructions are on the Wiki page. We also encourage testers to join the #fedora-test-day channel on Freenode IRC: members of the Fedora QA team and graphics development team will be around to help you with the testing process and also to take a look at any bugs that you might find. If you’re not a regular IRC user, you can find instructions here, or you can simply click here to join the channel (chat room) through a Web front end.

Even if you’re not a Fedora user, please consider joining in - Fedora's GNOME is a very clean build from upstream and fixes for any bugs found at the Test Day will land in upstream GNOME for all distributions that use GNOME to benefit from. Please do come by if you have time! Thanks.

There is no internet

Today in random philosophical pontification...

I think now is the point at which I should use my power as Grand Poobah of Everything to declare that everyone except network engineers should stop talking about 'the internet', please.

There no longer is 'an internet', except at the physical / protocol level which smart network engineers maintain, virtually thanklessly, so venture capitalists can make lots of money on top of it. And when people talk about 'the internet', they're not talking about that.

Lots of talking heads like to pontificate about 'the internet', even still - is it ruining our social lives? Is it good or bad for society, politics, free speech, knitting, $WHATEVER? I think the reason these types of discussions tend to be ridiculous is that they're built on a false premise. There just isn't 'an internet' any more.

There was a long period from the 1970s to the 1990s when there was clearly 'an internet' like this, which the few geeks who used it could sensibly talk about in a monolithic way. Pretty much everyone on 'the internet' used the same small range of services (which didn't come to include the WWW until the 1990s, of course) - email, usenet, telnet, ftp, maybe a bit of IRC. It was the network layer which still exists with a thin veneer of basic services on top that a fairly small number of people used for broadly similar purposes.

There was then a very brief period from the late 1990s to the early 2000s in which a whole lot of people were using 'the internet', but they were still really only using a pretty basic range of services - many of the same ones that the geeks used, especially email. The new bunch of 'regular people' internet users went gaga for the WWW, of course, since it's nice and easy to use and permits the publication of lolcats.

Now - and, really, for at least the last five years or so - there is no 'the internet' any more in terms of a recognizably unitary culture or set of services. 'The internet' that I, as someone who started using the internet at the end of the 'geek' period, use is utterly different from 'the internet' that a typical high school kid uses, which is utterly different again from 'the internet' that a 63-year old guy in Dakota uses...and so on.

I think I've visited tumblr twice. I filter the crap out of facebook and treat it as an analog of MSN Messenger, which may not make any sense to Mark Zuckerberg or an internet marketing graduate (these things now exist) but makes perfect sense to me. I only have a Twitter account for sending out QA announcements.

I run a desktop mail client and an IRC client and still use an ftp client occasionally, which makes sense to me and maybe most of my readers but is an alien language to someone who first used 'the internet' on a smartphone. (Let's not even get into the servers.)

Even the WWW has now diverged enough that any ten people might have a completely different set of top-ten most visited sites (aside from Google). Does it make any sense to say that someone who visits gadget blogs, someone who visits political news sites and someone who visits petcare tip sites are all using the same internet? I'm not sure it does.

This really crystallized for me with the Hurricane Sandy coverage. There was loose discussion everywhere about how it was being followed 'on the internet'. Well I was reading about Hurricane Sandy 'on the internet' but to me, that meant I refreshed the CNN front page about once every five hours. I wasn't subscribing to the mayor of NYC's Twitter feed or following photo feeds or anything like that. How is it possible to talk about a single 'Sandy on the internet' experience?

My internet is not your internet, and all our internets are now different. We're all routing our traffic over the same network but there is no monoculture any more. It does not make any sense to talk about 'the internet' in that way, and I would like for everyone to stop, please. Thank you!

Another 'living in the future' moment

So I just saw this post in the planet Fedora feed, and by the time-honoured romance language trick of 'all the words probably mean something a bit like a similar word in one of the other languages' decided it was probably interesting. So I fed it to Google Translate, and got a perfectly decent and intelligible translation of the original, with only one or two obviously awkward translations and a single missed word.

This is one of those things you can do and not think much about it, and then you think, holy crap, a computer just translated entirely unknown human writing for me, from one messy, sloppy human language to another messy, sloppy human language. And I could read it. A computer did that. That's pretty nuts. And all it takes is a giant international evil tentacle farm (that's Google), a mind-boggling amount of slurped data for the computer to learn from, and some pretty smart engineering boffins living off whatever organic mulch they're serving in the Google cafeterias this week. boggle

Active Directory integration Test Day Thursday (2012-10-18)!

It's Test Day time again, folks. This Thursday, 2012-10-18, is Active Directory integration Test Day. This Test Day is focusing on the Active Directory integration feature, which combines several improvements that together aim to ensure that Fedora can act as a member of an Active Domain out of the box.

This is one of those Test Days where you're either interested or you're not, and if you are, you know it :) Some of you might be very interested in having Fedora systems act as AD members, some of you might not care at all. If you're in the former group, it's in your interest to come along and take part in the Test Day - we need this kind of real-world testing to ensure Fedora 18 and future releases can work well in an AD setup.

As always, the event will be in the #fedora-test-day channel on Freenode IRC; this will likely be a chat-heavy event where you'll be able to chat with the developers about your configuration, so please do join IRC for this one. The main developer on this feature is Stef Walter, who'll be in IRC as stefw. All the basic test instructions are on the wiki page.

So if this is an area that interests you, please try and find a bit of time to drop by for the Test Day, it'll help us to help you! Thanks.

Thoughts on Iain M. Banks' 'Use of Weapons', as I have nowhere else to put them

I realize this is the second time I've somewhat randomly dumped a large splurge of thoughts on an Iain M. Banks book, but hey, it's my site. :) There is nothing to do with Fedora or Red Hat here. But I suspect there's a reasonable degree of overlap between F/OSS people and Banks fans, so that should be fine.

I finished re-reading Use of Weapons (the third Culture book) for about the sixth time today, and was interested in looking at some of the available criticism of it, so I looked through a few online reviews. I found them, frankly, frustratingly shallow and incomplete; I think the sheer brilliance, subtlety and depth of the book is not widely appreciated. (I was also incredibly annoyed that there was a Guardian book club event, a detailed in person discussion of the book with Banks, a measly two bloody months ago, which I entirely missed. Gah. Would've flown to London for that.) Most of what follows was written as comments on various reviews, so it's not really structured as a standalone review and starts out quite randomly, but I wanted to pull it all together for my own page, I wasted too long writing it all to just dump it. Note that UoW has a very spoil-able plot and these comments comprehensively spoil it, so don't read this if you have not read the book. Go read the book, because it's one of the best books of the 20th century, then come back and read this. Mainly this is for people who do what I do, poke around the internet randomly for UoW criticism, in the probably forlorn hope that it might help to shine a light on what a great book it is, in my mind far superior to almost all his other books (the only that's close, in my mind, is Inversions, which is similarly brilliantly subtle and similarly poorly understood.)

There is one element to the 'why a chair?' question (active in several discussions of the book, with many people missing the point entirely...) that seems never to be brought up: in one of the scenes of the four children, there is a point where they each learn some sort of manual skill.

"Livueta wanted to take up metalwork too, but her father would not allow her to; it was not seemly. She persevered. He would not relent. She sulked. They compromised, on carpentry. The boys made knives and swords, Darckense pots, and Livueta the furniture for a summerhouse, deep in the estate."

It's not just that Cheradenine finds Darckense and Elethiomel having sex on a small, white chair; it's a small, white chair that Liveuta built with her own hands. It is one of the objects that ties all four of them together, in a horribly twisted way. Throughout the book, any object or event that ties all four together has unique significance to Elethiomel/Cheradenine.

Some people consider it an act of sheer psychosis or plain evil, but that's not it at all. Elethiomel constructing a chair from Darckense's bones and sending it to Cheradenine and Livueta is...intentionally, on Elethiomel's part, a quite terrible irony; as some reviewers have pointed out, one of the crowning examples of Elethiomel's talent, the 'use of weapons' that recurs throughout. It is a precisely targeted psychological weapon. (Also note that by the time Elethiomel sends it, Elethiomel and Cheradenine between them have destroyed all the physical remains of their shared past; this recreation of just one solitary element of it is clinical.) You have to take into account the sheer desperation of Elethiomel's position in the war at this point. It's made quite clear that his position is virtually hopeless, as bad as his position in any of the scrapes he later gets into. He is besieged in a permanently beached ship, in control of that ship and a single city. All on Cheradenine's side are quite convinced that the war is won, the remaining issue is only how to achieve Elethiomel's surrender at minimal cost. The chair is a quite terrible weapon, but it is uniquely effective, as potent a demonstration of Elethiomel's 'use of weapons' as any in the book; it almost wins him a hopeless war. Just killing Darckense would not have had the same effect, nor would torturing her, threatening her, even sending her body to Cheradenine would not have 'worked'. The chair is a precisely calculated move. Elethiomel is not psychotic, though he may well be sociopathic (but then, so is Cheradenine, possibly more so).

Unintentionally on Elethiomel's part, it's a sad mockery of creation, the only form of creation of which he is truly capable (witness his attempts at poetry, later). Again, the childhood experience is important here; the two girls both choose creative pursuits, Elethiomel and Cheradenine choose to make weapons. This is echoed by the lives of Elethiomel, who makes a career of war, and Livueta, who makes a career of healthcare.

There's far too much focus in UoW criticism on the Culture; the book is about Elethiomel/Cheradenine, it is his story, all the rest is backdrop. As Banks noted in an interview, he wrote the first version of UoW before any other Culture book and originally invented the Culture simply as a backdrop for Elethiomel's story, the Culture were intended to be unambiguously 'good' guys to focus the narrative on Elethiomel. This is still the function they serve in the final book. If you read the other Culture books it's easy to attach too much importance to the Culture in UoW, but UoW is not really about the Culture. If you want a book that's much more about the kind of stuff people try to discuss when discussing UoW, the Culture's interventionism, you want Inversions or The Player of Games. In UoW the Culture are just the good guys. The point of the byzantine games the Culture plays in several of the episodes, particularly the Hegemonarchy one, isn't really to make you question the Culture's motives or philosophy or achievements, it's to highlight elements of Elethiomel's character and motivations; it's necessary to our understanding of Elethiomel, for several reasons, that he is put in the position where he is expected to lose, while believing he is expected to win - and more than once, at different stages of development. UoW is a novel of character and motivation; the plot is incidental, the background is incidental, all other characters (besides the four key ones) are incidental or exist only to reflect on Elethiomel. Yes, even Sma.

I could write a book on it, maybe I should, but to dip into just a few things that seem to be completely missed by most criticism:

  • The careful design of the 'past' chapters. They are not random vignettes from Elethiomel's past. They are a quite linear story told in very precisely constructed incidents. Almost every one involves Elethiomel, a war, a woman, and a chair. The fact that this motif remains central, unchanging, while the nature of the war and the circumstances in which it takes place vary wildly, is quite intentional.

  • The extent to which the trope of the 'use of weapons' prevails throughout the book. Elethiomel himself is a weapon, a fact of which he is quite aware, for instance. Virtually every action he takes, throughout all his attempts at reinvention and whatever else, can be understood as a 'use of weapons'.

  • The question of what Elethiomel is trying to achieve at various key points in the book. The scene just before the twist, where he abandons the Hegemonarchy, is usually either entirely ignored or briefly skimmed over, but it's a key scene, and a slippery and difficult one. So much criticism seems to take Elethiomel's own explanations of what he is doing as gospel, which is strangely unsophisticated. Elethiomel never quite understands what it is he is doing; he's quite open about this at various points through the book. He is not really attempting to prove that he can be a good person in contrast to his past, he is not attempting to atone for his sins or anything as simple as that. Everything that happens in the book is important to Elethiomel only as a reflection of his past, he spends the entire book attempting to understand it, to understand himself. Many critics seem to consider the structure of the novel only in the context of the 'final twist', but it's far more complex and subtle than that. The explicit self-descriptions of motivation that are usually quoted all come from early in Elethiomel's personal history. The truth comes at the end, but is not stated, as he abandons his sole talent, his use of weapons (the point in his walk at which he refuses to kill a soldier is critical), and attempts a form of self-destruction before he goes to meet Livueta. There's so little attention paid to the actions Elethiomel does not take. He never takes the Culture's offer to just stop doing what he does, to live with them or anywhere else as a normal person. At the end of the Hegemonarchy segment he doesn't simply let Sma take him away, or keep his morning appointment to be taken away, but he isn't simply suicidal; he completes his walk and then signals the Culture to extract him (they don't take him away against his will). His walk is a renunciation of his methods, his skills and abilities, and a strange attempt at self-punishment, self-destruction, before his meeting with Livueta; he has the idea, without understanding it, that he must be damaged when he meets her.

God, this is getting long, but some quick hits, tricky questions to keep in mind when re-reading which illustrate the depth of the book:

  • What is Elethiomel's 'obsession', which he tests with the torch in the century ship? Why is it related to his abortive suicide attempt? Why does he not know how to die when the torch goes out? Why does he always court death and never quite embrace it? (The answer to this question changes throughout his history, and the true answer is different from the various answers Elethiomel comes up with himself.)

  • What are the implications of the surprisingly numerous scenes in the book in which different characters look through windows at snow?

  • How do Elethiomel's methods change throughout his career, viewed strictly in a linear way? Particularly with regards to the way he interacts with and uses other people? What are the implications of this? How about his understanding of his own capabilities, and of the nature of conflict?

  • What exactly is the significance in Elethiomel's mind of the stone chip in his chest, and the relationship between the stone ship on which they test the gun and the Staberinde? Why the repeated emphasis, in Elethiomel's remembrances, on the potency of the Staberinde (in implied contrast to the stone ship)?

  • Why does Elethiomel think about killing the frozen woman on the century ship? Why doesn't he do it? Why did Banks include the scene at all?

  • 'Why did Banks include the scene at all?' is a good question to ask at all points in the book, in fact. It's not for the plot. It's never for the plot.

  • What exactly does Elethiomel's late career mean? Remember, it's pretty packed: he retires, he comes back for another job, he goes rogue, and then he comes back without any kind of a fight when Sma asks him to. Why? What's his actual frame of mind? Why does he only start asking the Culture to find Livueta at that late stage?

  • Not to get Freudian, but what's the significance of Elethiomel's parents, and the Zakalwes'? What's the impact of the uneven grouping, that there are two boys and two girls but three of them are siblings? What is the relationship between Cheradenine and Livueta, between Elethiomel and Livueta? What's the importance of the rather neglected childhood episode where Elethiomel nearly kills Cheradenine?

  • What does the contrast between the two poems written by two women about Elethiomel tell us?

  • Not a question, but note the neatness of Skaffen's almost unthinking, instinctive 'use of weapons' - oh, such weapons, far beyond even Elethiomel's capabilities - "to do good" at the very end of the main narrative.

I really should stop there :) But it's a great book, most of the things identified as flaws in it are not flaws, and it is far deeper, subtler and more consciously structured and designed than it is given credit for. I've read it six times and I'm nowhere near the bottom of it yet. I don't know if I know the answers to half those questions yet. It goes all the way down. And it's all about Elethiomel.

(This part in response to a debate in one comment thread on whether the Culture knew Elethiomel's 'secret' history): I don't think any part of the Culture knows Elethiomel's true identity or personal history at any point before Skaffen discovers it. The way Skaffen's part of that chapter is written seems to make it quite clear that Elethiomel's origin planet is unresearched by the entire Culture, not just by Diziet and Skaffen. The Culture finds him during the war on the tabular icebergs, a war in which it is implied that the Culture is interfering. They recruit him on the basis of that experience, not on the basis of his further past. My reading, anyhow. The other reading is possible but not really supported by any part of the text as I see it. Anonymous wrote "After a cursory scan the ship informs Skaffen-Amtiskaw that Zakalwe has been dead for many number of years. So what puzzels me is this, would'nt special cimcumstances backtrace a possible recruit's past before they are considered for employment? Especially if it were as easy as the said part indicates.", but it's only easy for Skaffen because he's on the planet where it's easy to access the history. It's clearly stated that Culture had never previously done a serious analysis of the planet. The Culture's powers are considerable but not infinite; they can mess with a planet from a distance of at least one light year, but there is definitely a limit on this range, and the universe of the Culture series is huge, the Culture travels and acts across far far far larger distances than it can directly affect. It seems pretty clear the Culture can't fiddle with a planet from, say, a thousand light years away. It's also fairly clearly implied that Elethiomel never told the Culture where he was from until quite late in the timeline the novel covers; as I wrote, the Culture found him in the tabular iceberg war and there are various points in the book where it seems to be made clear that they don't know where he was originally from, probably until he starts asking to meet Livueta for the first time at least.