Smalltalk Tidbits, Industry Rants

Dare Obasanjo has a few words for the complexity oriented:

Actually, a lot of the stuff he talks about with regards to SLAs, monitoring business processes and regulatory issues are all things we face as part of building Windows Live. However it seems Jeff missed my point. The point is that folks building systems in places like Yahoo, Amazon and Windows Live are building systems that have to solve problems that are at the minimum just as complex as those of your average medium sized to large scale business. From his post, Jeff seems to agree with this core assertion. Yet people at these companies are embracing approaches such as RESTful web services and using scripting languages which are both often dissed as not being enterprise by complexity enterprise architects.

He goes on to give an example, comparing RSS to WS-Eventing:

Just because a problem seems complex doesn't mean it needs a complex technology to solve it. For example, at its core RSS solves the same problem as WS-Eventing. I can describe all sorts of scenarios where RSS falls down and WS-Eventing does not. However RSS is good enough for a large number of scenarios for a smidgeon of the complexity cost of WS-Eventing. Then there are other examples where you have complex technologies like WS-ReliableMessaging that add complexity to the mix but often don't solve the real problems facing large scale services today. See my post More on Pragmatism and Web Services for my issues with WS-ReliableMessaging.  

The bottom line is, you don't want too much of that enterprisey stuff on you.

I like reading Ted, but I think he makes a common mistake here - it's one you see a lot of people make in the software development business, while discussing Cedric Beust's take on Ruby and the mainstream:

Cedric's definition of mainstream includes being appealing to Visual Basic and PHP programmers. That seems to be the backdrop of his first two points, that Ruby and Rails are too hard for these folks. I can see some of these points - folks in our reading group have been somewhat mind bent by some of the Ruby concepts, and they are Java/C# folks, which would put them higher on the food chain than VB and PHP programmers. I think that some of this is just unfamiliarity as opposed to difficulty, but there's not doubt that there is a learning curve there.

I find that a little funny, because of the conversation I had with Joshua Bloch a couple of years ago after his API talk at OT 2004. When I asked him about various Java things I dislike ("final" being the main one), he asserted that Java developers are down near the bottom of the developer food chain, and needed crutches like that - Smalltalkers and Lispers were up near the top, so they didn't.

The only thing that differs about these thoughts is where various programming cliques belong in the food chain. Just about everyone thinks that VB developers are some kind of lower life form, but they differ about where other people "belong" in the list. The thing is, I've dealt with a number of VB developers, and I don't accept the idea that they are lower ranking. They found a tool that works for them, and solves the kind of development problems they face. Heck, for lots of common tasks, VB has been a great tool - it can solve a number of problems faster and easier than Java, Smalltalk, or Ruby. Giorgio Ferraris recognized this awhile back, and has been building tools to make Smalltalk a player at that end.

One of the larger blind spots in this industry is silver bullet-ism. We Smalltalkers are as guilty of it as anyone else. Java developers think Java is the end all, be all. So do Smalltalkers. So do Lisp developers, and the Ruby buzz is filling the Ruby-ists with the same triumphalism. In an important sense, many VB developers may actually be more sane about this - most of the VB developers I've met don't think VB is the uber-answer.

Mike Austin points out an interesting issue surrounding the learning of Smalltalk - the perceived irregularity of the syntax. As is pointed out in his comments, it's not irregular syntax, but the consistency of the object model that's at work. However, Mike has a point in terms of how newbies see things:

In one aspect, Smalltalk is a very small and elegant language. But I find there are some hidden intricacies when you actually use it. One example is the difference between using the messages whileTrue: and ifTrue:. whileTrue: expects a Block, while ifTrue: expects a Boolean. In the following, at first glance it's not understood why one uses square brackets and the other parenthesis:

[count < 10] whileTrue: [count := count + 1]
(count < 10) ifTrue: [count := count + 1]

Once you understand that the control messages are just that - messages, not operators - it makes sense. I can see Mike's point though - to a person just learning Smalltalk, that might be seen as confusing. I'll have to think on this.

I've just posted some information on the upcoming releases of Cincom Smalltalk - both summer and winter. See the summer info page and the winter info page. The summer one is a bit sparse at the moment.

Here's an update on the StS 2006 Coding Contest:

The Coding Contest has been extended until the 19th of April.

While there was a lot of interest in the contest, many people complained that they could not do it in the timeframe alotted. Unfortunately, this means we only got a few submissions on the 9th of April.

Because of this, the STIC Board has decided to extend the contest until the 19th of April. If you wish to participate, just drop me an email (michael.lucas-smith@softwarewithstyle.com).

Those who did complete have been granted finalist status and will compete at Smalltalk Solutions. These two finalists, of course, are allowed to continue to refine their program.

This means there is one slot left open - so get in and compete for this last position. If you were unable to complete the contest due to time constraints, now is your chance.

Best of Luck,
Michael Lucas-Smith

Good luck!

There's still time for Advance registration, and by using the Smalltalk code of ST135, you can save even more money. You'll get access to all the talks, both LW/NW and Smalltalk. Here's the sort of Smalltalk session you'll have access to - Smalltalk in the food sorting industry:

Ever wonder how the food you eat gets sorted, the good from the bad and the ugly? This is an experience report describing how Key Technology's latest machines (and the people behind them) do just that, using Linux, Smalltalk, and Test Driven Design. The presentation will cover the challenges, solutions, and wins using these technologies. We'll probably show some of the UI/control system, some footage of the machines in action, and the auto test and build environments in action. The intent is to "tune" the presentation as we go based on the interests of the audience.

See you in Toronto!

Today, Joel Spolsky hits the sweet spot of what he does best - explain the mechanism of software development at a business level. It's a great column, and managers everywhere would be well advised to read it. The summary explains via an analogy - read the whole thing:

Nobody expects Dolly Parton to know how to plug in a microphone. There's an incredible infrastructure of managers, musicians, recording technicians, record companies, roadies, hairdressers, and publicists behind her who exist to create the abstraction that when she sings, that's all it takes for millions of people to hear her song. All the support staff and management that make Dolly Parton possible can do their jobs best by providing the most perfect abstraction: the most perfect illusion that Dolly sings for us. It is her song. When you're listening to her on your iPod, there's a huge infrastructure that makes that possible, but the very best thing that infrastructure can do is disappear completely. Provide a leakproof abstraction that Dolly Parton is singing, privately, to us.

That's how a productive business entity runs

My daughter was disappointed about a school field trip - there were a couple of possibilities, and she was hoping to get the trip to the Spy Museum in DC. Didn't work out that way, so I'm taking her down there tomorrow. It's spring break, so we can do it on a weekday. It should be fun; I'll be out most of the day.

James McGovern asks to be engaged:

Hoarding of knowledge also comes about in that most folks in other enterprises work in walled gardens where they are not allowed to communicate with outsiders. I am the polar opposite. Would the community consider engaging in a face-to-face conversation at any of the events I plan on speaking at? You may find it interesting that I will be on a panel regarding web 2.0 at the Infoworld conference hosted by Jon Udell where I hope to represent and encourage others within the enterprise to not eschew but to embrace web 2.0.

I'd be more than happy to engage Mr. McGovern in a reasoned concersation. Other than OOPSLA, none of the events he mentions are on my radar at the moment, but if he wants to set up a panel discussion at one of those events, I'll happily look at it.

We held a User Conference in Franfurt (Germany) in December 2004. We decided at the time to hold a new conference every other year, with the location to be determined.

Well, it's now 2006, and we are looking at where to hold the December conference. It will be in Europe, and we are casting about for locations. So - suggestions? We are open to ideas. I'll have a new survey up shortly.

Dare runs into a problem with the .NET APIs that I'm sure every developer has hit, with pretty much any non-trivial API they have to work with. I know I've had my own private "WTF??" moments inside VisualWorks while working with Silt and BottomFeeder. There are subtle disconnects between library developers and library users, and every so often we land in the seam between them.

More good pragmatism from Dare Obasanjo:

MIME types and HTTP content negotiation are good ideas in practice that have failed to take hold on the Web. Arguing that this fact contravenes stuff written in specs from last decade or from findings by some ivory tower group of folks from the W3C seems like religous dogmatism and not fodder for decent technical debate. 

That said, I don't think MIME types should be retired. However I do think some Web/REST advocates need to look around and realize what's happening on the Web instead of arguing from an "ideal" or "theoretical" perspective.

Sometimes, I feel like web specs are like battle plans - they don't survive contact with content providers any better than battle plans survive contact with the enemy.

We are looking to finalize the location of our bi-annual User's Conference (last held in 2004, in Frankfurt, Germany). We are looking at a number of locations in Europe, and have put together a survey to guage interest. Please give us your feedback here.

Thanks!

Google Calendar sounds like it should be an interesting service - I've never had the personal patience required to use a tool like Outlook (or, back in the day, various other PIM applications). Heck, I have been hard pressed to manage a paper calendar (so much for my organizational skills). However - having a calendar app that ties in via RSS to my aggregator? That sounds interesting to me, and like something I could actually make use of. I'm setting some stuff up there now, and we'll see how it goes.

The conference location application works now - I deployed it at 3 am last night, which was a mistake :)

Here's the dumb thing I did - I'm saving the results to a simple BOSS (serialized object file), as a running collection. Simple enough - but here's how I was getting that collection:

existing := self getExiting.
existing := existing add: newInfo.
self saveToFile: existing.

Now, if you know the way the collection APIs in Smalltalk work, you'll have spotted the error: #add: answers the thing you pushed in, not the collection. So I was saving only the new entry. Worse, the second usage of the application blew up, since #add: through an MNU. Sigh.

Never deploy an application at 3 AM...

There's still time for Advance Registration for Smalltalk Solutions 2006 - you can save even more money with the STIC discount code. You'll want to attend, so you can attend talks like Avi's - which should have some DabbleDB details:

Every startup needs a secret weapon. Ours was Smalltalk. Learn how we got an acclaimed web service out the door with no investors, no capital, and no experience, staying profitable the whole time.

See you in Toronto!

Digg may be trying to be the new slashdot, but they don't need to copy the historically low signal to noise ration. Here's an example: I just ran across this Digg item, talking about DMCA extensions that would make Firewalls illegal. I thought it sounded familiar - and sure enough, the link ran to a 2003 Register story.

I took my daughter down to DC yesterday - I had promised her a trip to the Spy Museum. Alas, I have no photos from there - they don't allow any (kind of appropriate in an atmospheric sense :) ). There was a long wait to get in, so we had time to stroll the neighborhood before our 3PM entry time. So, we walked down to Ford's Theater, where Lincoln was shot in April, 1865:

There's a small museum in the basement, which we visited. It has various exhibits from the time of Lincoln's death, including the nasty political cartoons of that era. The various people who think that politics are too mean spirited now have no sense of history, that's for sure.

We didn't get into the theater itself - it was restored back in the '70's, and is a working theater now - there was a group rehearsing while we were there. We took a look at the old boarding house across the street though - that's where Lincoln actually died, after being carried across. There was a long line, so we didn't go in - but I did snap a photo of the plaque:

We also walked down to the Navy Memorial, which is beautiful. Victoria took a few shots of that - I might post a few once she gets around to downloading them off her camera.

This article has a lot to say about raging fanboy-ism, but also correctly identifies what Nintendo is up to: they aren't after the hardcore, willing to spend 12+ hours a day gaming addict. Instead, they are after the casual gamer, and especially after the handheld game market. They have a great thing going with the gameboy/DS space; the Revolution is a small upgrade to the GameCube, intended to attract those currently not in the market for game systems:

The oddest thing about the fanboy fantasy of Nintendo being King of Video Game Mountain again is that Nintendo doesn't share that fantasy. Nintendo knows where its profit is coming from, and that's handheld gaming. No one console is selling as well as Nintendo's combined GameBoy lineup (DS and GBA and all the various other incarnations) and no game division is raking in the cash like the mountain of licensing fees Nintendo collects from handheld software.

They didn't spend huge developing a new machine - the Revolution is just a souped-up GameCube. They're not taking a loss on the manufacturing - that would require some kind of initial investment and the danger of losing it (Microsoft took a 10-digit loss on the original XBox). They took no such risks because they didn't need to. Anything they make off the console market is gravy to them.

The beautiful thing for Nintendo - in business terms - is that no one else is competing for that space. At all. Microsoft and Sony are both after the "serious" gamer market. They each lose tons of cash on each console sale, as they continue to add better and better graphics (etc). IMHO, one of them (and I think it's going to be Sony) will say "uncle" eventually. The interesting thing then will be whether the survivor has any interest in encroaching on Nintendo's turf.

It's an awful lot like the early "browser wars", actually - remember when, for a few years, MS and Netscape constantly pushed new and better stuff out? What happened when Netscape cried "uncle"? It wasn't a golden age for browser fans, that's for sure. Rather, it was a long slog of stagnation until Firefox appeared to push things again. When one of the two hardcore systems dies, I expect to see a few years of stagnation in the console space as well.

The Enterprisey theory of development is still very prevalent in the industry - witness Robert McIlree's take on development:

The underlying theme behind the anti-EA, and moreover, the "Web 2.0-Saves-Humanity-As-We-Know-It" crusade is this: you, the user, can have it better, cheaper, and faster if you [ fill-in-the-programming language-or-kewl-technology blank here]. Most of us who have been around the information technology business for a substantial length of time know that, eventually, this mode of thinking reinforces a number of serious and detrimental issues, particularly in the complex corporate and government environments where most of us ply our trade.

Here's the thing: Most applications just aren't that complicated. The propeller heads would like you to think they are - too many analysts want you to think they are - and too many vendors want you to think they are. As Chris Petrilli said today, an 80% solution delivered quickly is far more valuable than the Enterprisey solution that takes years and millions of dollars. The (supposedly) highly scalable, buzzword compliant solution doesn't help anyone while it's busy being late.

Reminds me of a situation related by a friend of mine awhile back. He was learning about various development projects that were ongoing at his new firm, which does consulting to a government agency. He was hearing about one project, that had set itself up to use a three tier architecture, Oracle as the DB, Enterprise Java Beans to connect to that, and a browser on the front end. He asked about the number of end users, and the answer - at the height of deployment - was "fewer than 20". He suggested that they just implement a simple Access front end to the data and be done with it. They branded him a heretic and sent him on his way.

I get the impression that McIlree would have been excited about the buzzword compliant enterprise architecture - even though it was going to take well over a year to build. Had they taken my friend's advice, they could have had a working 80% solution within a couple of weeks. But hey - it wasn't enterprisey enough, so down the garden path they went, led by people like McIlree.

There's another problem too - the large development job that takes N years to deliver is probably outdated by the time it does manage to get delivered. Those are the real wages of Enterprisey-ness - late solutions that cost tons of cash, and end up being outmoded to boot. Heck, this next bit from McIlree is more or less proud of that:

We work in environments where IT budgets are in the tens of millions, and in a number of cases, hundreds of millions of dollars. While there will always be some wasted money and failures financed by budgets in that range, part of our role is to insure that the systems designed and deployed with those monies provide value and cost control to the organization beyond the scope of any individual system or project. As JT notes, "Every architect and customer must understand the REAL business problem and functionality we are solving for." Not only is that true, but I would add that a message like this must be clearly communicated to executive management, both line and IT. If you do not have the complete support of your CIO, for starters, you're working with a minimum of one hand tied behind your back.

Translation: "You bozos have no understanding of the really important (expensive) job we're doing here. Leave us (and our large army of favored consultants) alone so that we can deliver a scalable enterprise (extremely costly and immediately obsolete) solution"

The real answer: you don't want any of that enterprise stuff on your fingers. Deliver the 80% solution now, so that the actual business of your company can move forward. What McIlree - and too many IT people, to be honest - forget is that they are just plumbers. Important, yes - no one likes clogged pipes. An actual center of profit? No. IT enables profit, but it doesn't actually create any of it.

Well - it looks like the reality of offshoring/outsourcing savings differs from the hype - the savings are 10% - 15%, not 60%:

Outsourcing of information technology and business services delivers average cost savings of 15 percent, a survey found on Thursday, disproving market claims that outsourcing can reduce costs by over 60 percent.

After professional fees, severance pay and governance costs, savings range between 10 percent and 39 percent, with the average level at 15 percent when contracts are first let, according to outsourcing advisory firm TPI.

Well. The question you then have to ask yourself is this - is a 15% savings worth the hard to measure, but real annoyance your customers face when dealing with disempowered support staff they can barely understand? Proving that management is often immune to reality, the article goes on to state:

Cost reduction remains the primary motivation behind current outsourcing contracts, but an increasing number of companies are outsourcing primarily to improve quality, at 21 percent now versus 11 percent in 2004.

*Cough*. Yeah, I've always felt that I get better service when I deal with a remote call staff. I have to repeat everything I say, and if my problem doesn't fall into the "is it plugged in?" bag, I have to escalate out of of their domain anyway. Which is always hard. Better quality my posterior.

Like Troy, I procrastinated on the whole tax thing. I just got done with TurboTax - or maybe it just got done with me. It installed nicely, offered to update itself - then crashed when applying updates. Tried again - same thing.

Sigh. Time to reboot. Rebooted, ran without updates. That worked fine until the part where I needed to file - then it demanded to ne updated. With trepidation, I went along. It worked. Then I recalled that I ran Civ 4 last night, and it always leaves Windows in a weird state.

At least I'm done with the filing of returns.

Looks like I touched a nerver - look at Robert McIlree's brilliant response in my comments - he even picked a great title for his comment: "Please stop blogging, you have another wife to beat and a dog to kick"

Feel free to give James or me a call after the crap you obviously develop leaks data to some identify theif or has various state attorney generals crawlingf through your systems before they hand what's left of your carcass back to the trial lawyers.

I do have to thank your content-free rants for one thign though, it drove a lot of traffic to my site. With luck, these folks probably have a few more brain cells then you do.

The Wizard of Oz didn't much care for the curtain being raised either. When you want a dysfunctional, but buzzword compliant, enterprisey answer - contact someone like McIlree. When you want something that gets delivered before the budget is busted, and without an excess of magic quadrant-ness, try considering the true nature of your business problem. I'd bet good money that it isn't actually as complex as some would have you believe.

Advance Registration for Smalltalk Solutions at LW/NW is still open - and STIC members can save even more money - send an email to Suzanne Fortman for details. Sign up now, so you can hear Georg Heeg extoll the virtues of Smalltalk:

In 1981 the first article series was published about Smalltalk. In the early 1990s many very successful Smalltalk projects were started. After a slow-down between 1996 and 2002, Smalltalk is once again picking up more and more momentum. This presentation looks at the question what the inner properties of Smalltalk are that it is still considered hot although almost 25 years old.

See you in Toronto!

Chris Petrilli defines Enterprisey for the clue challenged.

It's time for my weekly look at the logs - looks like BottomFeeder downloads ticked up to a rate of 331 per day last week - I'm never sure why these sudden burts (up or down) happen. Anyway, the details:

Platform	BottomFeeder Downloads
Windows	784
Sources	491
Linux x86	167
CE ARM	146
Mac X	140
Update	117
Mac 8/9	86
Solaris	69
HPUX	59
AIX	59
Windows98/ME	49
Linux Sparc	46
SGI	35
Linux PPC	31
ADUX	19
Source Script	18
CE x86	1

Off to the HTML page accesses for the blogs, by tool:

Tool	Percentage of Accesses
Mozilla	68.2%
Internet Explorer	19.7%
Other	4.6%
MSN Bot	3.8%
Megite	1.5%
Everest/Vulcan	1.2%
Opera	1%

After a one week absence, Everest/Vulcan is back. I wonder if that means testing is underway again, or if it's gone into deplyment? Off to the RSS pages accesses:

Tool	Percentage of Accesses
Mozilla	26.3%
BottomFeeder	14.2%
Net News Wire	11.5%
Other	9.7%
BlogLines	7.7%
Safari RSS	5%
Internet Explorer	4.5%
Google Feed Fetcher	3.5%
BlogSearch	2%
RSS Bandit	2%
SharpReader	1.7%
NewsGator	1.7%
Planet Smalltalk	1.6%
Magpie	1.6%
JetBrains	1%
Liferea	1%
MSN Bot	1%
Feed Demon	1%
Java	1%
Attensa	1%
News Fire	1%

Apparently, back in the 1920's and 1930's, the war departments of the US and Canada had a little too much free time on their hands - both developed cross border invasion contingency plans.

Via Steve Rubel comes news of another brilliant theory of effective management: blocking RSS feeds:

In case you hadn’t heard about it, some companies have begun blocking RSS feeds at the firewall. The rationale for this short-sighted, counterproductive bit of paranoid stupidity ranges from bandwidth worries to productivity concerns. The first I heard of this was from a reader of my monthly email newsletter. I’ve been cajoling my 2,500-or-so readers to switch to RSS for well over a year now. This particular reader wrote back saying he’d be happy to give RSS a try but for the fact that his company has banned RSS.

That just brilliant. So when product managers and product marketers want to monitor what's being said about their products - their management instead throws a cone of silence over them. I love the rationale behind this - it's to improve productivity:

The rationale behind monitoring employees, according to Newman, is that a computer at work is a corporate tool for enhancing the employee’s productivity. Because some people abuse that privilege by sending personal e-mail and viewing movies during working hours, employers feel they have little choice but to monitor what their workers are doing.

Ok, here's a tip to every manager and IT staffer who's ever had that thought - lie down until it goes away. If you have people who are not doing their jobs, then there's a simple procedure: document the problem, and - if it doesn't stop - terminate the employee(s) in question. Punishing the whole class instead of having the guts to address real problems simply lowers morale and productivity. Yeah, that's a brilliant management strategy. To follow it up, I suggest holding $100 bills up and lighting them on fire.

The same sort of stupidity is blocking mp3 downloads. Yes, I understand the legal issues, given the current state of cluelessness at the RIAA. At the same time, your marketing department might well want to monitor podcasts covering the industry you're in - there could be good news to tout, or negatives to counter. Of course, there's also the cone of silence approach. That works so well in negative PR situations.

Dale Wolf (a colleague of mine at Cincom) explains how syndication technology hits the b2b sweet spot in terms of attracting the attention of busy people.

With taxes out of the way, it's time to register for Smalltalk Solutions 2006 - Advance Registration is still available, and Smalltalkers can get the STIC discount code by contacting Suzanne Fortman. Registrering will give you access to all tutorials and all sessions - the Smalltalk tracks and the LW/NW tracks alike. Get in now, so you can save money and learn about the latest in Wiki development from Lukas Renggli:

Web applications and wikis are often built using string-based approaches to parse and generate the resulting web-pages. While such approaches work well for simple applications, they hamper the customization and adaptability to end-users with more sophisticated needs such as different output formats, user-interfaces, management tools, application logic and security policies. Pier (formerly called SmallWiki 2) is the second version of an industrial strength application framework built on top of Seaside. Pier is written with objects from top to bottom and it can be easily customized to accommodate new needs. Pier is based on a powerful meta-description called Magritte, that allows one to create user-interfaces elements, queries and persistency in a declarative way.

See you in Toronto!

I suspect that China's new "no email server without a license" law will have a few unintended consequences - it'll probably play a role in any offshoring decisions. Why? The simple cost (monetary and bureaucratic) in getting a license, and the need to monitor otherwise inocuous conversations:

China's new rules also prohibit use of email to discuss certain vaguely defined subjects related to 'network security' and ' information security', and also reiterate that emails which contain content contrary to existing laws must not be copied or forwarded. Wide-ranging laws of this nature have been used against political and religous dissenters in the past.

Ignore the politics of free speech for a minute, and just consider the notion of having a few developers in China - and having to consider which technical issues are safe to discuss.

Mini-Microsoft:

Moderation. So I know you have Clay Shirky talking about it and folks like me and Scoble living it, but what is the solution? It seems that blogging as a communication medium is prone to entropy the more successful a blog becomes... perhaps comments should be tiered so that there is always a secondary page one can go to for all submitted comments and elevated comments ( either by the owner or readers ) can make their way onto the main blog page to ascend next to the main post's text. Kind of like Slashdot, just without the one-liner noise of each filtered message.

It's simply human nature at work. As any communications forum becomes popular, the number of troll comments rises - it's happened on old-time BBS systems, USENET, Slashdot... etc. Not a big shocker that the popular blogs have the same problem. Heck, look at any popular political blog and you'll see one of two things:

• No comments allowed (i.e., link from your own darn blog)
• An ever growing, more and more useless comment section

It's just the way things work

Rick Bradley is setting up a dynamic languages group in Nashville:

I've set up a Google Groups mailing list for the Nashville Dynamic Languages group "which is a pretty informal social and technical group of people (predominantly located in Nashville) who are interested in dynamic programming languages ( Smalltalk , Ruby, Lisp, Io, etc.). We had a get-together earlier in the week to test the waters and had a really good turn out. There's a heavy Ruby (+ Ruby on Rails) bent, but there's a lot of interest as well in other languages.

There should eventually be a website at nashdl.org.

Sam Ruby is looking for consensus on the handling of content:encoded and the description entry in RSS feeds. One se of thinking is to treat the description as a summary, and the content:encoded as the full text:

As to content:encoded , if people can come to a consensus on to the precedence rules regarding description and content:encoded , the Feed Validator will honor such consensus. 

The trouble is that "in the wild", usage is all over the map. I haven't looked lately, but back when I added support for content:encoded, it was typically the same content as description - it was just explicitly encoded. Given that, I treat them as the same in BottomFeeder, and have content:encoded override description. I suppose I could add a preference, but I don't see a compelling reason to do so at the moment.

Now, maybe if there was a real spec for RSS, this - like other issues in RSS - would go away. We all know that there's one man who, for utterly inexplicable reasons of his own, thinks that more specificity would be a bad thing.

With tax day over, there's no reason to put off registering for Smalltalk Solutions at LinuxWorld/NetworkWorld. Advance Registration is still available, and Smalltalkers can get an additional discount by contacting Suzanne Fortman prior to registration. There are lots of great talks, like Emil Markow's on regression testing tools:

Testing at Ontario Teachers' Pension Plan has for some years heavily relied on BRITE (internally developed, now open-source application) to run acceptance and regression tests. BRITE is used to test applications written in Smalltalk and applications written in other languages, most notably Java. This session talk about BRITE's main features, including test data management, test execution, reporting and documentation management and Web Services testing.

See you in Toronto next week!

If you're coming to Smalltalk Solutions this year and looking for things to do Sunday night (April 23), come down to Pure Spirits at 8:30pm where a bunch of us are meeting to catch up on things that may or may not have anything to do with Smalltalk.

Directions to Pure Spirits here.

Lighter fluid? Gasoline? Gas grills?

Nah, Real Men cook with Liquid Oxygen :)

If the RIAA folks had a clue, they would realize that this story is another nail in their asinine theories about music ripping:

Don’t feel so bad that your iPod contains illegally-obtained music, because US President George Bush has also been stealing music. Check out this video, where he talks about his Beatles songs on his iPod, and of course, Beatles music is not yet available online. That means he must have ripped them from a CD. Last February, the RIAA said that ripping CDs is illegal. Welcome to the band of thieves, Mr. President.

The trouble for the RIAA is that holding to that absurd opinion - that simply ripping a CD to a music player is itself illegal - nothing else they say gets taken seriously. Perhaps the top guys there should read the tale of the little boy who cried wolf...

Smalltalk Solutions is getting closer - it all starts next Monday at LinuxWorld/NetworkWorld. Advance Registration is still available at a discount over onsite registration - and Smalltalkers can get an additional discount by contacting Suzanne Fortman for details. Register now - there's a lot of cool stuff going on, and the price of admission covers all talks and all tutorials - both Smalltalk and LW/NW. For instance, Thomas Stalzer is talking about dynamic web applications in Smalltalk:

Dynamic web applications are the future of the internet. Currently the majority of web applications are more or less just forms which may be filled by the user. In dynamic web applications the granularity of changes is at a much more detailed level; e.g. changes in an entry field which is linked to an attribute of a server object may be reflected immediately including verification and dependencies. The seminar will discuss a possible new solution to combine classic web architectures with modern dynamic content behaviour.

See you in Toronto! And if you arrive on Sunday, join us at Pure Spirits.

I just posted a new (4.2) development build for BottomFeeder - not a lot of new stuff in this, just a collection of recent bug fixes. I'm running it here (I always eat my own dogfood this way).

I like Civ4. However, it does not like my notebook. Every time I play, it makes Windows wonky. The most common problem? The File dialog - the standard Windows file dialog - won't open properly after I play the game. It's just bizarre.

Here's a post that the enterprisey folks ought to read over and over, until it sinks in:

Yes, overengineering is common among us smart people. (Ed note: Has anyone else noticed that we need an international emoticon for irony?) The smarter we are, the more likely it seems we are to overengineer. Until you get to the really smart people, who are typically the ones offering you the chewing gum and pennies. (The brilliant people look at the solution proposed by the smart people and figure out how to implement it only using the same chewing gum and pennies that the very smart people offered thereby solving both the simple and general cases with no incremental effort.)

Words of wisdom

This Calvin and Hobbes cartoon should be the blogger's manifesto :)

More evidence that the USPTO should get out of software - this Burst.com suit against Apple's iTunes:

After being approached by Burst.com in late 2004, Apple had filed for a declaratory judgment in January that it isn't infringing on Burst's patents, but Burst is going ahead with its lawsuit, filed Monday in federal court in San Francisco. Burst is asking for royalties as well as an injunction, it said in a press release.

Yeah, because those bits that happen to be music download so very, very differently than the bits that happen to be an HTML document.

Update: Is this what you call "Innovation by lawyering?"

Burst.com is represented in the action against Apple by San Francisco law firm Hosie McArthur, who also represented Burst in its successful litigation against Microsoft Corporation. In March 2005, Microsoft settled that litigation by paying Burst $60 million for a non-exclusive license to Burst’s patents. Burst has also expanded its legal team in the Apple litigation to include attorneys from the Seattle office of Susman Godfrey, LLP, as well as Houston-based intellectual property firm Heim, Payne & Chorush, LLP. Also representing Burst is Palo Alto-based intellectual property firm Carr & Ferrell, LLP.

When you see "expansion of the legal team" in the headline, you can be pretty sure that there's a lot of horse hockey being tossed around...

I just love the way newish technology seems to impede logical thinking. Take the movement to make cell phone usage (without a headset) illegal while driving:

Addressing what safety experts say can be a deadly distraction, states are scrambling to impose restrictions on cellphone use by drivers. Twenty-six states and the District of Columbia have written legislation on the issue, mostly since 2003. This year, other legislatures are tackling the subject, and two states have passed laws on it.

The guy with the cell phone in his ear bothers me a lot less than a number of things. For instance, the guy one lane over eating a Big Mac. Or the other guy doing 70, but who seems to be reading a book. Or, the woman coming up from behind who's applying mascara. Or the guy two lanes over who's shaving.

What I want to know is, in what way are cell phones worse than any of the examples I just gave? Or tons of others I'm sure you can think of?

Dare Obasanjo thinks that Dave Sifry is playing fast and loose with blog numbers. I don't know - it's understandable that Dave wants to limit discussion to what he has facts on; it might be nice if he added a few caveats, sure. Dare's biggest beef:

It's now general knowledge that services like MySpace and MSN Spaces have more blogs/users than Technorati tracks overall.

I'd be interested in knowing why Technorati doesn't track those services.

We are still looking for feedback on the location possibilities for the December (2006) User's Conference. It will be held in Europe, but the specific location is not determined yet. Let us know what you think! If you tried the survey last week, there was a bug - that's been fixed, and your answers will be looked at.

One of the nicer things about Smalltalk is the consistency of the object model. A fair amount of that is due to the fact that everything is an object, and that dynamic typing allows for developers to just "do what works". I have a pretty simple example of this - let's look at the method #factorial

result := 10 factorial.
result inspect.

That gives us:

Ok, that doesn't seem exciting - we ended up with a SmallInteger object. However, now let's try this one:

result := 100 factorial.
result inspect

The nice thing here is that we didn't have to do anything special - in the process of getting the answer to the second question, the SmallInteger object got promoted up to a LargePositiveInteger, and I didn't need to do anything - no setting up of interfaces, no casting, no need to ensure that all factorials produce LargePositiveInteger objects - they get created when they are needed. Developers can do that themselves, btw - the library does this kind of thing with numbers, you can do similar things with your own objects as needed.

The bottom line - Smalltalk stays out of your way, and lets you solve the problem at hand. Instead of having the satisfy the anal retentive needs of the compiler.

Over the weekend, we tought we had lost an audio cable and an S-Video cable in our family room A/V hookup - sound was getting distorted, and the picture was gone for both the DVD player and the Replay feed to the TV. As it happens, the audio cables did need replacing - one of them had problems. The video? Just loose.

However, there was still the standard fun of unplugging the cables and trying to figure out why audio worked for the VCR, but not for the DVD (or vice-versa). Mind you, it's a unitary DVD/VCR player, and why the output labelled "DVD/VCR out" plays audio for the VCR, but not for the DVD - while a separate "audio out" does play the DVD audio - is a complete mystery to me.

It's even worse than what a bunch of us talked about over the weekend, when we noted that no one could understand anyone else's A/V setup anymore. Heck, I have problems with my own :/

Smalltalk Solutions at LinuxWorld/NetworkWorld is coming up fast - there's still a few days to save money with Advance Registration, which will buy you access to all the tutorials and sessions. Smalltalkers can save more by contacting Suzanne Fortman and getting the STIC discount code. There's a lot of great stuff this year, like Martin Kobetic's talk on Cryptography in Smalltalk:

This presentation will introduce cryptographic hash functions and public key algorithms and discuss some of their applications and practical aspects of their use. It will continue in the spirit of an earlier talk about secret key ciphers presented at Smalltalk Solutions 2004. The talk will include demonstrations using the VisualWorks security library.

See you in Toronto!