The BileBlog

Now, I’m not one to regularly hop onto the latest and greatest framework like some cheap two bit whore, as many are wont to do. You won’t see me dangling from the rafters screaming like a deranged baboon about how great spring/struts/tapestry/whatever is. Still, I realise that all these things have their users.

What is baffling however is the inevitable cry of ‘make an Eclipse/IDEA plugin!!’ that is guaranteed to occur from one in every 15 users.

Now, I’d like something explained to me. If your framework is THAT complicated, THAT confusing that you need to paint it in with pretty pictures, don’t you think something has gone wrong? If the config file format is so unintuitive that you cannot just type it all out after doing it once or twice, then you’re screwed from the outset, and you need to reconsider your choice of framework in the first place.

Of course, having that ridiculous plugin console could mean that that’s a tacit acknowledge that the underlying configuration was designed by developers whose parents made a point of dropping them on their soft little heads as often as possible, and as a result are unable to coherently express themselves without stumbling around the point for hours before finally (accidentally) arriving at their destination. So I guess there IS an argument to be made for struts to have a console, for example. God knows that the poor spastics forced to use it need all the help they can get.

What’s amazing is that without exception, all of these consoles are basically just a more awkward way of editing text. They all require you to be intimately familiar with the underlying configuration mechanism, and more often than not expect you to drop down to plain text editing if you need to do anything vaguely non-trivial. it’s a layer of abstraction nobody asked for.

I understand that there’s some sexual pleasure to be derived out of writing a little IDE component. Certainly, you webturds seem to relish the odd scraps of swing work tossed your way, but really, can’t you channel this naive childish curiosity into something a little bit more productive or useful? Do you honestly expect anyone to strictly use your shitty little swing dangleberries without being forced to know all about the underlying configuration file and how to edit it manually?

UPDATE For some reason, this bile entry is now showing up as the latest one. To get the latest (and for an explanation on this bizarre behaviour), use the bileindex on the right to find your bile of choice.

JRoller is one of those impressive applications where the version number consistently marches in the opposite direction to usability and functionality.

The new version is rolled out, and it is astoundingly bad. I don’t know who the current set of geniuses behind it are, but kudos to you guys, you really seem to know how to fuck up a good thing.

Where does one start? Why on earth were the font sizes changed? Why must titles on the front page be so massive? Is Rick Ross going blind in his old age, and forcing his need for font magnification onto everyone?

Speaking of the front page…What on earth has happened here? What sort of sick drug induced haze would make this seem like a good idea? If there’s ever proof that developers can’t design a functional UI then this has to be it. Previously, you could view all recent entries without scrolling. Now, you just see 3 if you’re lucky. Previously, you weren’t forced to read nazi shite or foreign gibberish, you just saw the title and author and quickly averted your eyes. Now you’re regaled with this wanky content whether you like it or not.

Now this might be a surprise to some, but most JRoller blogs are, on the whole, terminally dull and tedious. Most people have next to nothing useful to impart. Why oh why do we need brief blurbs for every entry?

Next up, we lose the useful and professional blog name in the hot blogs list, to be replaced by usernames. Who the fuck cares that this is the ‘fate’ blog? The author is irrelevant, what matters is the content and blog name.

The tabbed UI look is also astoundingly braindamaged. Most people are read-only users, so seeing a single tab with ‘main’ is a complete waste of space.

The search functionality is still amusingly useless. Returning any number of duplicate hits, and managing to be the only search functionality on any website in 2004 without paging support.

The edit weblog pages are equally pathetic and depressing. Fields are laid out in a haphazard manner, their positions seemingly decided by a random number generator. Where previously you could see how many comments a particular entry had right on the main edit page, now you have to manually count them in the edit entry page. The spastic calendar date picker does not work on Safari (it pops up, but never goes away).

Not only that, but we’re also presented with a list of nonsensical options for formatting. ‘Email scrambler’ and ‘bookmark linker’ might seem like obvious intuitive names as alternative formatting options to ‘textile formatter’, but to us dummies, it’s utter gibberish. An html link for each option to some sort of help or guide to the differences wouldn’t be considered too rude, you fuckwits.

The edit user page brings its own brand of humour, by insisting that I am in fact Albanian, and live in Eirunepe (in the US, apparently).

How hard can this possibly be for you people? I find it stunning that someone could have such an incredibly bad design sense. It’s almost as if whoever developed this UI did it as a cruel joke, and went out of his way to be spiteful and inconsiderate. Does your day job suck that much that you need to hurt others in your spare time, just to alleviate your inner torment?

UPDATE: Yes, I know it says comments have been disabled. Needless to say, that is not my doing, and my config clearly says that comments are enabled. Complain to the cocksuckers that wrote this pile of dogshit. I assure you that if any of them worked for me, they’d have been fired a long long time ago. I can only hope that they’re not allowed anywhere near any sort of production system in their day jobs.

UPDATE 2: Grr, comments have to be explicitly allowed per entry. Another stroke of genuis by the UI experts.

A week or two ago yet another Jakarta commons project was inflicted on a largely unsuspecting world. This release didn’t so much as arrive with a bang or make an entrance, it sort of guiltily snuck in hoping no one would notice.

If you were to have a look at the contents of this latest abomination, you’d realise why it’s guilty and ashamed from the outset. Let’s take a brief walk through this remarkable insight into the mind of a a happy idiot java beginner. Yes folks, most of you too were this stupid and ignorant once, so let’s not judge these genuises too harshly.

First up, we have the standard open source pointless class syndrome. This seems to be a vital requirement for any open source project; a bunch of classes that explicitly do sweet fuck all. In commons-io, these take the form of True/FalseFileFilter. Wonderful file filters that always pass and always fail, just what your project needs.

Stepping back a little, one has to wonder at whether the world really, really needs a file filter package whose sole purpose is to bring together the built-in FileFilter and FilenameFilter. These works of wonder take up 13 classes of the 34 in the whole jar. The whole premise is somewhat ridiculous, anyone who cares a whit about performance, encapsulation, or readability would NOT build up file filters by contorting their code into bizarre shapes just so they can do something like new OrFileFiler(new NameFileFilter(), new AndFileFilter(new TrueFileFilter(), new NotFileFilter(new FalseFileFilter(), new SuffixFileFilter())));. At the risk of being controversial, might I suggest doing what any sane person would, and just writing your own goddam filter that does exactly what you need it to?

If only the horror ended there. When these clever kids were taught Java, nobody bothered explained the standard API to them, or how it contains an awful lot of handy-dandy classes that one need no re-implement. Witness the genius of BufferedOutputStream! commons-io of course ignores that, and includes the inspired DeferredOutputStream. What does it do? Why, it buffers, but only once! Latter chunks are no longer buffered! Still, they should be forgiven, as this functionality is something that all of us regularly make use of, I mean, who DOESN’T need a stream that has one buffer chunk then starts streaming?

Unconvinced yet? Well, here’s the crowning glory of this package of congenital defects. IOUtils.toByteArray(String). The more perceptive and java-savvy amongst you might remember that String already has a getBytes method. Oh well, I guess that’s what they call in the implementation. Not so! They instead creating a byte array output stream, write the string into it, then return the contents of the stream as byte[]. Why oh why weren’t these people taught that strings are immutable? Why did nobody think to point out to them that if you modify the byte[] of a string.getBytes(), the string itself is unaffected? Of course, let’s not shame these Jakarta folk any further by being pedantic and mentioning that converting a String to byte[] is pretty silly if you’re not going to take encodings into consideration (needless to say, they’re blissfully encoding ignorant all over the API).

It’s easy to laugh at these wankers and dismiss them out of hands. That’d be irresponsible though. As adults and contributors to Java it is our job, nay, our duty! to educate and help these poor souls. So do the world a favour. Find the nearest Jakarta committer, and beat him like a red-haired bastard child he most likely is, while explaining to him the java fundamentals. Maybe this time round they’ll stick.

Oh and for those of you desperate for more JBoss goodies:



From: "Nathalie Mason" 

Received: from galaga ([65.81.161.44]) by imf15bis.bellsouth.net

JOnAs motherfuckers. At least when the JBoss/BEA killer thread got posted on
the ServerSide we didn't have to stoop so low as to post it ourselves.

"We think JOnAS...is...maybe..er.a better app server than JBoss."

If they are going to use negative PR tactics, they are going to have to  come
up with stronger language than that

What one of you guys needs to do is make up an alias on Serverside and  post
something like
"JOnaS rules, JBoss sucks big donkey dicks"

That way they'll get the choirboys after them too

As a result of the JBoss expose here and elsewhere, various trade rags have picked up the story. It made it onto slashdot. The comments there of course are equally hilarious and depressing. Amazing how many people fail to grasp the difference between an anonymous post and a fake persona post. Still, that’s to be expected from a bunch of spotty gentoo tuggers anyway.

Next up we have eweek, which ran a very professional sounding story which managed to put JBoss in a hilariously bad light. The best quote of course had to be the ’so-called professional’ line. Delicious.

Not to be outdone, JDJ had to jump in the fray. Needless to say, I strongly approve of their article. I’m hopelessly biased against JBoss so that couldn’t possibly surprise anyone. Cameron buying me so many drinks has also clearly affected my judgement, but oh well.

However, the real fun starts if one were to look at what else is going on at JDJ now. On the one hand we have a sane sensible article about happy community building and a mild waggle of the finger in the general direction of JBoss, and on the other hand we have….immature BEA bashing.

This farce started a couple of days ago, when JDJ ran an article about BEA ‘firing’ sys-con and electing to go with another publisher for an ‘official’ Weblogic magazine. BEA pointed out that sys-con’s Weblogic magazine is unaffiliated with them and suchlike and so forth.

At this point, it’s pretty clear that various head honchos at sys-con all simultaneously soiled their panties and suffered a fairly embarrassing bout of serious backdoor leakage. What follows is an amateur campaign of insults that would be quite newsworthy if it wasn’t trumped so thoroughly by those JBoss tards.

JDJ is now merrily in the process of running at least one or two disparaging BEA articles a day. These articles are hilarious. It’s like watching a three year old spastic child shaking his fists and expressing his anger by leaking from all orifices. Pretty it is not.

The contrast makes the whole thing very funny. On one hand, a sane article that doesn’t make the reader squirm and feel mildly embarrassed (unless said reader is JBoss employee, of course), and on the other hand we have petty vindictive jabs. Far be it from me to tell you how to run your business but…oh fine, I’ll tell you, since you asshats don’t seem to know. Have the right hand have a little chat with the left, and decide once and for all, are you to remain the object of mockery, or should people actually start taking you seriously?

Does the fun ever stop with these guys? It turns out that theserverside.com forums now has an interesting new feature that many of you might not be aware of. If you click on a particular user, you will see all the other users that have logged in from the same IP.

Obviously, this method is not foolproof, and can be easily misinterpreted to mean that two people behind the same proxy are indeed one and the same.

Having said that though, it is with immense glee that I see that once and for all, the JBoss wankstains can be seen for the hypocritical, conniving, underhanded, petty and insecure little fuckwits that they are.

So, let’s pick a user at random! Say..ohhh I dunno…Marc Fleury. Who else does the slimy little fleury have hidden away in this too-large-for-one-person personality of his? Why, none other than our friend Arun Patel! Arun, for those of you unfamiliar with TSS, posts incredibly offensive polemics that happen to exactly mirror the unspoken thoughts of a certain JBoss cult. To be fair, many of the posts are too eloquent to be His Royal Garlicness, what with his penchant for multiple exclamation marks, rambling sentences, and all round freakish usage of full stops.

So, who else posts from that IP? Well, we have James Hardy. James’ posts are often of the ‘I’m sane but lets face it, JBoss rule’ variety, as opposed to another on that list of deranged psychopaths, Chip Tyler. Chippie here will eagerly pipe in in any number of threads to say how much CDN suck, as well as how every move JBoss makes is intelligent and wise. There are literally dozens of other accounts that show how widespread this behaviour is. The only thing they have in common is a surprising love and admiration for all things JBossy, and disdain and abuse for all things non-JBossy.

Needless to say, Ben Sabrin and the majority of the JBoss folks are all on the trail too. So while it’s impossible to actually draw lines between the fakes and their puppetmasters, it’s very very easy to spot the group of nefarious rumprangers who have embarked on this laughably incompetent marketing exercise. Having said that, some of the linkages are very clear and easy to follow to individuals who happen to not work in the same turdfactory as the fleurys do. Bill Burke has the highly dubious honour of also being Joe Murray, famous for making noises to the effect of ‘Mike Spille doesn’t exist!’. Let’s see you uuhmm and err your way out of this one Billy! Marc’ll have you back in that gimp suit pretty sharpish if you keep being this sloppy.

It all makes an awful lot of sense, if you think about it. JBoss is very very little beyond a whole lot of smoke and mirrors. I was actually surprised at how many people at TSSS for example smirked when I asked if they used JBoss. The silent majority pretty much views that server as a bit of a joke, and rightfully so!

What is stunning is how they seem to have next to no capacity for learning. Come on garlicboy, how fucking hard can it be to write a classloader? Do you really NEED UnifiedClassLoader, UnifiedClassLoader2, AND UnifiedClassLoader3? Didn’t your mother teach you how to name classes? The thing is, it’d fine if you lot were just crapping out hilarious code. Sadly, it isn’t enough for you, instead you feel the need to go out of your way and create a million fake accounts everywhere just to spread the word and hijack any potentially intelligent conversation. Are you that scared of innovation and a world where JBoss co-exists with the rest of humanity?

The saddest part about all this is that the most likely outcome is for these posters to now ensure they use a different IP when posting, to disguise the trail more effectively. I’m sure the very notion of ‘gosh, maybe we should let our software do the talking instead of using underhanded tactics like these’ is heretical in that camp.

So, the next time you see someone posting anything positive about JBoss, rest assured that it’s just another one of their employees desperate to justify their pathetic existence.

UPDATE: Removed Corby Page’s name since he is real. Sorry Corby, the ONE name I didn’t bother double checking turned out to be real, pah!

UPDATE 2: More dups! You can actually find which employees work at the ‘office’ and which work remotely. Juha-P Lindfors is also Thomas Mattson (a VERY prolific JBoss wanker), C R, and Cam Teegar. Even that fat fuck Andy Oliver seems to join in on the fun (which makes his innocent ‘why must you all hate us!’ mewling all the more hilarious), either masquerading as or living in sin with Steve Lewis.

One thing that struck me at TSSS when talking to ‘normal’ people was how many were using EJBs, running in lard-ass containers, no less. Everyone was excited about spring, hibernate, and whatnot, but very few non-adventurous types were actually using them. I heard more than one person complain about the fact that all this opensource fappery was nothing they could actually use in their enterprise environment, and thus somewhat irrelevant from a practical perspective.

The people actually using this cutting edge type stuff were often those who had their own blogs, weren’t afraid of joining mailing lists, and were pretty much part of a disgusting incestuous java community that has long since swapped common sense and practical results for feeble genitalia tugging.

The problem arises of course when this tugging of unmentionables is lauded and admired and viewed as some kind of astounding achievement. Even worse, those who do not indulge in this filthy habit are mocked and taunted and looked down on. The whole thing of course is reinforced by the fact that most are happily shielded from interacting with the ‘normal’ people, they’re able to simply dismiss all their fellow employees as ‘idiots’. It’s no coincidence that most of these tuggers will almost uniformly state that ‘my entire team is a bunch of idiots! I’m the only smart guy!’.

Of course, human psychology dictates that we must always clearly establish superiority over anything ‘different’. We’ve been doing it for millennia. So the genitalia obsessed for example won’t think ‘My job is terminally boring so I try to spice it up with dumb shit that others are too old to be amused by’, instead they’ll think ‘they’re all dummies!’.

Another requirement for this filthy circlejerk club is to have the memory of a guppy. You must, under no circumstances, learn from your mistakes. Instead, the self respecting java circlejerker will build hacks around the mistake to try and paper over the gaping holes and herculean logic flaws. In any other language, open source products are allowed to die the miserable death they so richly deserve. In the java world, nothing dies, it just gets refactored.

So look around, do you consider yourself in the smart minority in your team? Do you think it’s a matter of time before everyone else realises your genius and adopts your ways? If the answer is yes, then it’s time to get up, slowly pull the thumb out of your anus, and casually let go of the nearest guy’s penis, and perhaps, with truly open eyes, look around.

The third day (yes yes, I’m a day late posting about it) was a bit of a non-event, as far as I was concerned. Unlike the evil supermen who conspired to keep me up for obscene numbers of hours on all previous nights, I turned out to be some freaky mixture of mortal and zombie for the third day.

Needless to say, I missed all the morning sessions. I did however make it to the Eclipse tools session. The session was great, really really useful for me personally. The presenter’s voice was a soothing monotone that seemed almost engineered to put one to sleep. Needless to say, I was powerless against this astounding weapon and promptly fell asleep in my chair for the entirety of the session.

What DID wake me up however was Sam Pullara, managing to ask yet another awkward question. Sam has this disturbing ability to ask very innocent sounding questions that invariably most presenters either don’t understand or look foolish answering (usually the latter). I thought it was accidental until I realised he specifically came in to this session to harpoon the poor guy. The question (heavily paraphrased, needless to say, Sam is too polite and accurate for his own good) was if Eclipse had any plans to switch from SWT to Swing now that Swing beats the living poo out of SWT, and that Swing evolves much faster and gets much better with every release, whereas SWT is essentially a dinosaur idea that has far surpassed its extinct-by date. The answer consisting of the usual shuffling about, some feeble arm waving gestures, which I think pretty much anyone could have reduced to ‘umm, no. We’re sticking with swt, we’re screwed’.

The next session I went to was Scott Ambler’s Agile database stuff. Scott is by far the slickest speaker I’ve seen in this conference. It’s not all smoke and mirrors though as he clearly knows his stuff and manages to bridge the gap between the app guys and the data guys very well. His ‘agile’ talk though just ended up making the Thoughtworks guys look like a bunch of young boys staring down at a small protrusion between their legs and giggling at the fact that if they paw at it in an appropriate manner, they can illicit some sort of reaction.

So, onto the Symposium aftermath. Here’s a rough guide to it all:

Here we go, Marc Fleury is about to wow us with his genius. We’re finally going to find out what those ten million dollars buy.

Well, it doesn’t look like they got much value for money, to be honest. It seems to have bought them a few ludicrous constumes and some bad face makeup. Marc’s Joker smile looks more like a Vegas hooker’s lipstick on a Sunday morning than Jack Nicholson’s thespian perfection. Marc indulged in a few foolish leaps. On stage though (sans costume, but still clothed, thankfully) is Bill Burke, merrily giggling to himself (no doubt interjecting the peels of laughter with the regular spate of uhms and errs that seem to accompany him whenever he sees a stage).

Our friend starts off with some feeble joke directly at yours truly (sadly, he pretty much mumbles it out so I have no ideas if it was actually funny). This was followed by an enema reference from the Batman movie, likely a branching out from a homoerotic obsession with the caped crusader and his sidekick Bill^H^H^HRobin.

The JBoss propaganda machine is in full force. Marc is clearly derisive of the crazy AOP people, or the academic aspect people. Nono, JBoss does it differently, it uses…AO! Yes folks, the P has been unceremoniously ditched. They probably found out that they only own 66% of AOP so dropping the P seems like a good investment, as it allows them to own 100% of AO!

Next up we have a nice little blurb about how great EJB3 is. As expected, he has a competition with himself to see how many times he can say JBoss and EJB in the same sentence.

The obligatory ‘arm waving’ joke is tossed out (sorry Marc, Gregor did it with infinitely more panache and style), and for some bizarre reason, our new fearless leader chooses to insult Sun by being very dismissive and vague towards the JBI spec.

Next up we have possibly the worst diagram I’ve seen in the last two days. Trust me, I’ve seen a LOT of bad diagrams the last couple of days. This one has a dazzling array of circles, with various blocks, cylinders, and lines tossed in for visual effect.

More tired analogies of better moustraps, the usual adulation directed at Microsoft, Marc is still a good speaker, he just says such ludicrous things with amazing disregard to content and reality.

In turns out that the new version of JBoss, the one that’ll take everyone to the stratosphere of application development, is going to be an orgy of aspects (sorry, AO) that makes all current aspectfests seem like a bunch of servlet weenies.

It turns out that to do this, the JBoss group now, in its entirely, all now THINK IN AOP! Presumabely a few pressed of Marc’s control device and the appropriate instructions were beamed into the brains of such hapless victims as one Gavin Fleury.

It is now time to reveal a dirty secret. AO is, apparently, not new at all to the JBoss camp. JBoss 3.0 and JBoss 2.0 both included AO! That’s right, JBoss is not jumping on the AO bandwagon for cheap PR - they’re driving it and have been at the wheel for years, behind the scenes, scheming away in some private cvs tree (yay open source!) that has to this day remained hidden. Mike Spille is familiar with this tree, I’m sure.

What aspects have they been using? Simple you fool - the logging aspect! Yes - the feature that sets JBoss apart from those silly fools from BEA and those suited morons at IBM is the 90 pages of logging that JBoss spews out when starting, some call it logging, but to them, it’s a logging aspect.

OH HORROR OF HORRORS! There’s a huge error on one slide! True to JBoss style, sloppiness rules the day, and this slide has, emblazoned on the top, almost taunting poor Bill Burke (who is still looking uncomfortable) with those three letters, A O P. That P! How dare it! Out damn P out!

Setting aside the usual slipshod crap they’re throwing up, we are also presented with some ideas of highly dubious merit. For example, the whole concept of a web request in its entirety being in a transaction. Gosh, I’ve always wanted to roll back my transaction after returning a result page telling them everything worked. If JBoss has learnt anything, it’s the importance of a false sense of security to give to everyone they deal with. Credit though for not only applying that approach to their marketing, presentations, talks, and posts, but to the very code itself!

The slides really are the gift that keeps on giving. They’re degenerated to such an extent that the current one actually says ‘FAST TIME TO MARKET’. There also seems to be a bit of an obsession with things of 3. We have titles like ‘etc etc etc’, ’simplification simplification simplification’, ’simple simple simple’.

There’s also a continuing battle of wills between Marc and Bill to determine which one actually wears the pants in this presentation. Bill gets up, sits down, Marc accosts him to explain something, Bill gets up, shuffles about, demands Marc explain something. Perhaps a few rounds of practice might have paid off.

At this point, something seems to have gone horribly wrong as Marc resorts to making…..fart noises. Yes folks, the man who just managed to scam a VC out of ten million dollars is up on stage, with cheap lipstick, making fart noises. There is something sick yet fascinating about this.

Finally, we’re regaled with JBoss’ stunning success, with pictures of bigger and bigger circles denoting increasingly amazing things. The next logical step would simply be for them to delcare JBoss an independent country.

We’re finally at the end, and nobody has any questions. Marc clearly wanted to say something about open sourcing java but I think he badly miscalculated when most people responded that they do NOT think that Java should be open sourced. His feeble recovery effort was nothing more than a whimper, “I know what that means. OK, I don’t know what that means.”.

That’s that folks, an entertaining keynote, albeit not deliberately so. At least, not in the manner it came across. I’d also like to think the 5 other mean horrible people who contributed insults through the wonderful magic of Rendezvous (a group bile, if you will). Most importantly, I’d like to thank the genius nearby who realised I ran out of battery power and rapidly whipped out a spare and handed it to me. Now THIS is a Java community!

Mark Hapner is pretty much an old school guy. He’s boring, he’s a techie, and he generally knows whereof he speaks, he also doesn’t seem to be the sort to resort to cheap circus tricks when presenting (ie, it’s boring as fuck to watch this).

However, this is appropriate in this talk. SOA usually involves a series of complicated arm flapping maneuvers that are rarely seen outside of AOP/TDD/XP/Agile circles. Consequently, the attendees are also mostly serious boring people. No OSS tuggers or framework fappers here!

Interestingly enough, Mark manages to produce a diagram that actually explains what SOA is in somewhat clear terms. Basically SOA seems to be a layer on top of all the other layers we have now. It’s a platform that is almost too obscene to be anything but a very strange joke. There are clouds involved. Service clouds, policy clouds, federation clouds, and many other clouds with nary a silver lining in sight.

In fact, it’s all so horribly boring I might hop next door to Jason Carriera’s WebWork 2 talk. The problem with doing that of course is that I’ve promised to be well behaved about it, and I don’t trust my more bestial instincts. If I know me, I might just lose it completely if I do go there, so maybe I shouldn’t tempt fate.

A service in the SOA world seems to contain the usual layer upon layer of gunk. At the lowest level we have our meat, the bit that actually does Stuff. This can be EJB or whatnot. Next up we have BPEL (I dozed off briefly so I really don’t remember how the two tie together), then above that we have that old dead horse, web services.

As if one SOA cloud wasn’t enough, we also have the option of many clouds of SOA thingybobs coming together, whereby each SOA cloud somehow automagically picks up the fact that you want a service that happens to need to cross the cloud, and it’ll automatically manage talking to this other cloud and making friends with it and suchlike.

The real trouble with this talk is that it’s so very…abstract. There’s absolutely nothing accessible in the whole idea. If anything, the expert group seems to be struggling to actually pin down what SOA actually entails in a ‘real’ sense, instead of the furious flapping as trademarked by the likes of Gartner.

Even setting aside the abstract nature of the concept as a whole, every one of the little (!) components are equally ambigious and fluffy. For example, you have stuff like ‘integration brokers’ and ‘policy directories’ and other such ideas. The JBI (Java Business Integration) spec wants to provide SPI’s for all these guys. Worst of all, Mark keeps saying ‘wizdle’ and ‘bepple’ instead of ‘W S D L’ and ‘B P E L’, and it’s hugely distracting. In time though I suppose it’ll provide a lot more amusmenet: ‘if you don’t stop investigating your unmentionables I’m going to have to bepple your wizdle in front of everyone’.

One note of personal interest for me was the general discomfort directed at BPEL. I’ve always vaguely considered spending some effort at it for the sake of OSWorkflow, but the general concensus is that BPEL is just too complex and unwieldly and uncomfortable to use in javaland.

Verdict? SOA is very, very boring, and it’s hard to find two people to agree what it actually is in non-arm-flapping terms.

I’m using SubEthaEdit to write biles. This means that anyone with an OSX machine can download it and run it, and join the edit session to see biles being written ‘live’. It might be an interesting experiment anyway, and needless to say that if people misbehave or cheat or do anything that I deem to be uncool they’ll be turfed off. So, download it, run it, click rendezvous, and you should see an edit session for ‘bile.txt’ by me, if you’re a sensible distance away (I just tried it with jiramike and he’s in a different room in another session).

This morning’s first talk was provocatively titled ”From EJB to Hibernate to EJB3′. The promise at least is rather delicious, that all those turdy little hibernate asshats will now suddenly migrate to the next cool thing, EJB3, because one of their fearless leaders told them to.

Sadly I don’t hold much hope for that being the take home message, given the bad taste the EJB3 talk yesterday gave me.

The first half of the talk goes over some background and history. Gavin is an earnest and genuine speaker, but he’s clearly rather nervous and hesitant. He’s no Bill Burke, but there certainly were one too many umms and the odd mumblemumblemutter instance.

Next we go over Hibernate and how great it is. Blahblahblah, hibernate blahblah POJO flimflam CGLIB fapfap.

The Hibernate fappery ends up with a short but sensible list of Hibernate shortcomings. While he doesn’t explicitly say so, there was kind of implicit ‘EJB3 will fix these!’ air to the whole thing. Yes, I realise I’m grasping at straws at this point, but the situation is dire and desperate.

We’re given an example of an EJB3 entity bean. While on the whole it looks sensible, I was mildly horrified to see the table name jammed right there into the source code. I suppose it’s easy enough for vendors to treat that as a hint, and still allow some sort of configuration file that takes priority over what table the source says it wants itself jammed into.

One persistent problem with this presentation is that it’s pretty clear than Gavin thinks all this is astoundingly obvious stuff (which it is), but his argument for proving that a particular approach (eg, named parameters in queries) is….’and named parameters are great because….they’re good’.

Gavin also frequently mixes ‘I’ and ‘we’ (as in the expert group). What’s funny about this is that he’s pretty much screwed either way. If he says I, the EJB whiners will complain of him hijacking the spec. If he says we, Hibernate users become defensive at losing their fearless leader. Anyone at the conference can see this hilarious conundrum at play by talking to Jason Carreira. Poor old Jason is furious that EJB3 gives him stuff that Spring and Hibernate already does. What’s baffling about this is that I have yet to find any laws which prohibit using Spring+Hibernate. Of course, if EJB3 did NOT build on what those tools already did, Jason and his ilk would have panties in a bunch about how EJB still sucks. There’s no reason to be so agitated, just bugger off and use your frameworks and stop showing up at the EJB talks.

The whole psychology and personal investment aspect of EJB is interesting. For some reason it’s become a religious issue. People have built careers out of whining about EJBs. Without EJBs sucking so much, they’d find themselves looking foolish and outdated. Alongside that, there are Hibernate/Spring users who are resentful that they might no longer be the cool kids on the block. On the other side of the argument of course are the people who explicitly rejected Hibernate/Spring (yes, such brave souls do exist in a secret underground cult), and now feel resentful that the ideas they aren’t buying into are being shoved down their throats anyway.

What is clear from all this is that Gavin spends an awful lot of time working on the new spec. What remains to be seen is how that affects Hibernate’s future (if at all), given that it’d be somewhat of a natural evolution for him to start advocating EJB3 instead of ‘old’ Hibernate.

AOP panel

The final tech event of the day was the AOP panel. The story of AOP seems to be somewhat similar to the story of linux on the desktop. THIS IS THE YEAR FOR IT!!!!

Gregor Kiscales was the first speaker, who immediately endeared himself to me by offering to flap his arms about wildly to try and convince the world that AOP is a whole new paradiggum shift. To his credit, he does fully acknowledge that one could live one’s life quite successfully WITHOUT using AOP. It’s a different approach that one can train oneself to think in and develop in. That’s an argument I can easily stomach. AOP is indeed another (interesting) approach to solving certain problems. Are we all doomed to die a horrible death if we don’t switch to AOP as soon as is physically possible? Nope. Are we doomed to have disgusting filthy code with poo littered every which way, with the twin demons of inflexibility and unmaintainability ruling supreme? Nope!

Rod Johnson was up next. Smarting from being shat on every which way by the EJB 3 talk, he went out of his way to see how many times he can pimp Spring while still remaining on-topic. There was much talk of flexibility, non-invasiveness, crosscutting behaviours, and vital POJO stuff. Full marks for buzzword compliance!

Adrian Koyler is up next. Apparently he’s an IBM chappie who now works on AspectJ. Poor old Adrian never had a hope in hell after admitting that he’s responsible for the initial EJB implementations in Websphere. He dutifully quoted the one study that insists that AOP is the best thing since sliced bread, while simultaneously being the simplest thing since Fred Grott’s brain. Pretty much a yawnfest, with the only redeeming feature being that Adrian had a decent British accent and seemed mildly amused by everything he said.

At this point it was becoming very difficult to remain awake (damn you crazybob!). The next speaker is Ron Bodkin, the man responsible for the obscene examples of AOP earlier today. I have no idea what he’s saying, but he managed to convince me earlier anyway that his brand of AOP is best smiled at politely and ignored.

Next up is Cedric Beust. Cedric seems perpetually amused by something that he won’t quite reveal. The trouble is, it’s just SO much more fun to sit here in the back and giggle and snicker with whoever happens to be around than to actually listen.

Bob Lee (crazybob) is next. Much to my glee he really does sound fucked up. Much to my chagrin, he quickly becomes comfortable and starts sounding coherent. Bob’s blurb basically involves him discussing dynaop, and the practical guff surrounding it. Oddly enough, he does seem to just get bored halfway through and just starts mumbling to himself, then perks up when he tells us of his new discovery, abstract advices, which he ALSO gets bored of and descends into mumbling. Much giggles all round.

Bill Burke is up next! Many uhms and errs. Some pitiful faffing about trying to justify JBoss doing its own AOP thing (how’s that CTS compliance going, billy?), followed by even more ahming and erring. He does the company line thing and tries to say JBoss every other sentence.

Finally, the moment we’ve all been waiting for. The head Fleury himself! It’s nice to see that they’re putting their 10 million dollars to good use and forcing poor TSS to allow two of those genuises on the panel. To Marc’s credit, he is actually is able to speak clearly. I can make out most words and he’s clearly well practiced. What he said though was a whole bunch of words that didn’t manage to actually say anything. Even worse, there wasn’t a single suggestion from him that various people do inappropriate things to his genitalia. Most unexpected!

Ted Neward next asks an interesting question about holding state in aspects. Ted is a Microsoft lover though and is here as an evil spy to try and get everyone to use .net, so I’m convinced his question was just his way of suggesting that we all switch to .net.

I’d like to cover the questions in more detail. The trouble is, AOP IS SO GODDDAM BORING. These bastards have been saying all this crap for over a year now. Enough already. One interesting fact about the questioners is that they seem to all be people who write code for other developers or ‘leader’ type people (Evil Ted, Howard Lewis-Ship), which I think does speak volumes about how much traction AOP has. Teehee.

I went to the EJB 3.0 talk expecting to be very pleasantly surprised. From all I knew, it was going to be an annotation fapfest the like of which hasn’t been seen since xdoclet.

I wasn’t dissapointed on that count. What was stunning was the utter dominance that Gavin King seems to have on the expert group. We were shown the list of expert group members a couple of times, and I think what’s basically happened is that everyone except the new joiners (read, Gavin) found the whole thing terminally boring and basically decided they can’t be bothered to argue or debate much.

End result? EJB 3.0 is basically a bizarre subset of Hibernate, for the CMP portions. The session bean bits are…Spring! While it’s funny seeing the spring folks have it stuck to them up the dirtbox, it’s still mildly galling to see how streamrolled this expert group has been. They really seem to have taken the servlet group approach: those who yell the loundest get their words immortalised in the spec.

Sure, the current state of the spec is promising. Linda (nice as she is, blahblah) came across as a bit of Gavin King sock puppet, with every other line involving what Gavin will say to us tomorrow, or what Gavin did, or how Hibernate works. It’s somewhat astounding how much they’ve chosen to toss out and basically decided to roll over and play dead when confronted with of the fists of fleury.

Here’s hoping the other members of the expert group wake up sometime soon, and realise that if all they do is try to play cathup with Hibernate and/or Spring, they’ll become a laughing stock (AGAIN) in 12 months when those two projects have moved on and Sun is pathetically farting out a spec based on a bunch of worthless whiners more impressed with how much penis they can tug than deploying enterprise applications. Asshats, one and all! I spit upon the EJB expert group!

This talk was given by Gregor Hohpe (who happens to be a thoughtworks chimp, but I won’t hold that against him as he looks quite serious).

The first half of the talk goes over threads and the awkwardness and pitfalls of using them. While many of the arguments are valid, at this point it’s not really clear that you can do away entirely with threads (of course, much easier to do if you’re abusing them in the first place, which I suspect is the point of this talk).

Irritatingly enough, we’re regaled with the usual ingratiating disclaimer that all thoughtworkers proclaim when they’re trying to sell anything. ‘Well I don’t know about you guys but I’m not smart enough to do X or do it correctly all the time so instead I’m going to do Y nudgenudgewinkwinkI’myourfriendreally’ is insulting and pointless. Note to all TWers: If you actually do think you’re dimwitted, have the decency to be ashamed of it, not proud.

We’re at the halfway point of the talk right now and I’m starting to feel a bit cheated. I came here expecting to learn about a new development paradigm and all I got was this lousy worker queue solution. Most of the talk so far pretty much talks about issues, pitfalls, and benefits of using a work queue with a thread pool managing the work.

The second half talks of how one would go about implementing this toy project using java.util.concurrent (basically Executors and Futures) with callbacks tossed in for returning results. Fun academic exercise I suppose, but if you’re going to build your whole architecture, you’re going to need a crapload more boundary case handling and error handling that this trivial example. To the guy’s credit, he does say that if you wanted a real worker queue/thread pool type architecture, there’s a lot of other crap you have to worry about.

In a way it’s hard to decide who this talk is aimed at. Many basic concepts of asynchronous calls are explained and spelt out at great lengths. Yet we also have many pointers on how one would go about writing such an asynchronous library with queues and threads and suchlike using JDK 1.5.

To be fair, this talk was actually pretty good. Sure, I didn’t learn anything, but that’s a minor issue. He seemed to have an understanding of the issues and APIs that was oddly…grounded. There certainly weren’t too many ‘let me wave my arms about to con you into agreeing with me’ situations, which is a pleasant and welcome surprise.

Next up is the Enterprise AOP talk. This promises to be particularly fun as AOP fans are basically those too lazy to do any thinking or design up front.

The talk starts with a couple of slides talking about modularity in Tomcat. Now, admittedly, I know next to nothing about practical AOP, so I am, at this point, utterly baffled by what I’m looking at graphs that have something to do with code in tomcat.

Shortly thereafter, we’re regaled with a slide pointing out ‘the crosscutting problem’. It dawns on me as this point that the first few slides were basically trying to say that Tomcat has crosscutting concerns, and things like logging are scattered everywhere. The usual suspects are trotted out to advertise the whole thing: non-AOP is ‘inflexible’, ‘unmaintanable’, ‘messy’, and ‘non-modular’.

Ohdear, here we go. The furious arm waving telling us that AOP is as much of a step as OOP. Why oh why must people be so obsessed with the sound of their own flatulence? Why can’t it just be ‘this is AOP, it does some stuff that might be useful’. Why do advocates of this shite need to feel like they’re messiahs?

After the bizarre arm waving interlude, we finally have the basic AOP concepts explained to us. Maybe someone should have pointed out that, amazingly, ‘crosscutting concern’ is in fact gibberish outside of the AOP world, introducing that term up front left many people looking befuddled and sad.

The example trotted out is also misleading. They pick on servlet buffering (basically, response.setBufferSize()) as a good candidate for a rampant round of AOPisation (using AspectJ no less!). Now, maybe I’m crazy and old fashioned, but this strikes me as an exceptionally braindamaged shoehorning of AOP. There’s NEVER any reason to use AOP with servlets. The spec lets you specify filters which can be applied pre and/or post request. No extra jars needed, no fuckfaced precompiling of freaky AspectJ sources, no bytecode genitalia tuggage, no AOP fappery anywhere. This really seems to be someone’s sick idea of Enterprise AOP.

After what seems like hours, the next genius idea introduced is using AOP to prevent servlets from setting the buffer themselves. Again, cruel disregard for the spec and the fact that that functionality is already trivial to achieve by simply using a ResponseWrapper that overrides setBufferSize().

Next construct to be introduced is how to limit the scope of the aspect and only have it apply to a limited set of servlets. Hmmm, yep, spec thought of that too! Welcome to filter url mappings, a modern day wonder provided for free to you by those crazy servlet kids.

Finally, they had the decency to try and highlight with the furious arm waving AOP contrived example with the servlet filter approach. The first slide manages to have many lines which allegedly say why AOP is better, yet somehow, doesn’t manage to actually SAY anything that doesn’t boil down to ‘AOP is better!’. The final ‘proof’ is that the AOP example is all in one class. Needless to say, it ignores the fact that declaring paths and applying filters in web.xml is a hell of a lot nicer than applying a filter to hardcoded classnames. Ultimately I think everyone felt a bit scammed by the ‘AOP is just better!’ argument. Ultimately all the time wasted on the example can currently be trivially done using 3 lines of filter code, and a one line response wrapper.

Of course, no AOP talk would be complete without a logging example. So to illustrate a ‘real world’ enterprise usage, we’re basically insulted with the ‘transfer money from account A to account B and log errors’ example. The only excitement added is some piddly exception handling. So what did we learn from this? Put very simply: avoid AOP, half the people involved in it are asshats who have a disturbing logging fetish.

Here we go. Loud music, thumping bass, and a RED background light, snazzy videos. It’s show time! Ohyeah, I was also wrong, this morning’s keynote is by the TSS monkeys. Fleury madness will take place at some point later on.

First we have a blurb advertising last year’s symposium. Hmm. More bass. Floyd walks on to some Eminem (!). Poor Floyd spends his time telling us how great J2EE is, while looking awfully uncomfortable. Next up we have the CEO and COO of The Middleware Company. Just when everyone stops paying attention, they pull out a silly skit where The governor of California face pops up with oddly misplaced lips, applauding TMC and suchlike. This is mildly amusing, but the effect is somewhat ruined by two things: 1) the guy behind the lips has very very different skin colour than Arnie, 2) the accent is sometimes russian, sometimes Indian, but almost never Australian. Credit though to the rampant lip licking, it adds a good dose of the ridiculous to the proceedings, and the final suggestion to use ‘lots of body oil’ was rather amusing too, although I think many missed the necessarily visual

Next up…err…we have Homer Simpson…with the SAME Russian/Indian accent! Homer does manage to get in a swipe at the JBoss people and all their hundreds of fake TSS accounts they use to pimp their piddly little server.

I never thought I’d say this, but it’s possible to be a little TOO self-deprecating. There are one too many .net jokes, too many ‘where is TSS headed’ jokes.

All this is followed by astounding amounts of waffle, telling us all about TMC is ‘repositioning’ itself, these poor guys are trying so so very hard to give a non dotcommie explanation the whole thing and failing rather miserably.

The final bit is going through some of the deck of cards. Mostly involved making fun of various pictures (not particularly hard given how few people sent pics in!).

So far, mildly amusing intro, sadly too waffly to actually feel too strongly about either way.

So here we are at TSSS at 8:30am. Conferences are funny things. There are invariably huge swathes of people stumbling about like lost little sheep. Some hesitantly poking their way through in the vain hope of finding a mother sheep to guide them towards better things. Most just mill about pathetically looking broken.

Still, a free breakfast is a free breakfast, and I’ll be damned if I let a little running-around-vegas-until-5am-with-crazybob-who-is-actually-crazy get in the way.

The first ‘event’ is a book signing. TSSS has roped together alll the poor schmucks who have written or pretended to write any old cruddy book and gotten them to sit on a ridiculously shaped table trying to look comfortable and natural (most seem to be failing). Nice to see that Jason Hunter is STILL milking his book though, the java world needs such permanent fixtures to give us a sense of stability and appreciation of the past.

What is surprising though is how many people have just slumped by the nearest walll and are sitting there with sad little faces looking for all the world like befuddled children unsure of quite what went wrong. Now, as any conference veteran knows, this isn’t surprising in and of itself, what is surprising however is that the damn thing hasn’t even started and the casualties are already piling up. Most disturbing.

The first fun packed event of the day is going to be Marc Fleury’s keynote speech. I will hopefully provide live coverage, while trying to stop myself from running up to him and humping his leg uncontrollably as my own special way of expressing my admiration for what he does.

Well, if TSS can manage to rustle up 52 losers to make up their ‘Who’s who in J2EE’ list, I can certainly match that and rustle up 52 asshats to make up my ‘Asshats in java’ list.

If anyone on the list is tempted to feel proud to be there, just look at the other people on it and see if you can honestly say that you’re proud to be on the same list as them. You should all be ashamed of yourselves, you rampant underachieving asshats.

Also, no jokers, nobody deserves to be singled out, I wouldn’t want anyone to feel special or honoured, or particularly insulted. Each category has 13 entries, in no particular order. Any name can go on any number card.

So Without further ado, here’s the list.

Comically inept