Smalltalk Tidbits, Industry Rants

I'm not normally a fan of class action suits - the biggest winners are always the lawyers - but I'll make an exception for the RIAA. I'm all in favor of giving them a taste of the rectal probe:

Like a shark smelling blood in the water, the latest round of investigations has attracted the lawyers. Prominent California attorney William Lerach has now launched a class action suit against the labels on behalf of consumers who have allegedly been overcharged for music. This in itself is not particularly surprising given the ongoing federal investigation into the same topic, but the lawsuit does contain some interesting tidbits. For instance, the suit claims that the music labels fought tooth and nail against the arrival of online music stores, and that they did so by launching their own poorly-conceived (on purpose) online ventures.

I had not heard about this suit before, but it looks like Microsoft's decision to settle - leaving Sony to sight the battle against a patent for game controllers with feedback - is going to bite Sony:

Sony's struggle with Immersion dates back to 2002, when Immersion came after Sony and its DualShock vibration feedback system for controllers. Immersion also pursued Microsoft and its controllers, but Microsoft settled with the company and entered into a licensing agreement, leaving Sony to fend for itself. In September of 2004, Sony lost a jury trial and was ordered to pay US$82 million in damages for infringing on Immersion's patents. Half a year later in March of 2005, Sony lost an appeal and damages were revised to nearly $91 million.

Worse, it looks an awful lot like Sony tried to buy off the guy who filed the patent:

In this latest round, Sony argued that Immersion was holding back evidence and requested that the original verdict be tossed out. They argued that inventions of Craig Thorner—once a consultant for Immersion—were not fully and properly disclosed. Sony argued that the full body Thorner's work on haptic feedback reveals weaknesses in Immersion's patent claims, and that such weaknesses are grounds for a new trial.

US District Judge Claudia Wilken has sided with Immersion. The problem is Mr. Thorner. While Thorner did once work for Immersion, he has also received a $150,000 payment from Sony for royalties and a purchase option on another patent. Although the money in question appears to be technically unrelated to Thorner's testimony, Wilken wrote that Thorner's testimony was suspect and that it was quite possible that he viewed his testimony as a favor to Sony. Since Thorner's testimony serves as the basis for Sony's new attack on Immersion's patents, Wilken's ruling effectively puts this line of appeal to rest.

Regardless of the merits of the patent, I doubt the court will take that payment lightly. I think Microsoft just got another leg up in the console space.

I've read plenty of criticisms of Microsoft's employee review system, but this one caught my eye:

Microsoft employees are growing more and more disillusioned with stagnating salaries and an increasingly contentious review system that they say is unfair, according to a recent report in WashTech News. That's led to more defections by senior engineers and growing dissatisfaction among rank-and-file workers, the report said.

Until I got to the next paragraph:

The publication is affiliated with the Washington Alliance of Technology Workers, a labor union affiliated with the AFL-CIO that has tried to organize Microsoft workers in the past. At issue is the company's performance review system, according to the report. Microsoft employs some 38,000 workers in the U.S. alone.

Now, never mind what you think of Microsoft, or this union, or unions in general. Just consider the concept of a "news story" that uses a source that was unsuccessful in its attempt to deal with the subject of the story. Nah, there wouldn't be a conflict of interest there, hmm?

Maybe CNN should just cut out the middleman and host press releases from advocates - employers, unions, political parties. That way I wouldn't have to go all the way to the second paragraph in order to evaluate the usefulness of a story.

From LWNW / STS 2006 comes this call for Student Volunteers:

We offer students of colleges and universities the opportunity to volunteer at the show. This entails assisting speakers in their session room and a variety of other duties photocopying handouts that sort of thing. In the time in between the specific job duty they are free to attend any seminar they choose. Or they can request that they volunteer in the specific seminar rooms that have the sessions they want to hear.

Please send an email to desiree@plumcom.ca and request a student application. We need at least 40-50 students.

See you in Toronto!

Apple is one of a number of companies that are going to get a crash course in globalization shortly - it looks like France is going to try and force them to open the iTunes store up:

It would no longer be illegal to crack digital rights management -- the codes that protect music, films and other content -- if it is to enable to the conversion from one format to another, said Christian Vanneste, Rapporteur, a senior parliamentarian who helps guide law in France.

"It will force some proprietary systems to be opened up ... You have to be able to download content and play it on any device," Vanneste told Reuters in a telephone interview on Monday.

Some people seem to think that Apple will shut down the store in France, but that's going to be hard to do - France, like the rest of the Continent (outside the UK) uses the Euro. If you use an ISP based across the border, what are they going to do? I suspect that they'll have to let it happen.

This reminds me of "daytime running lights" on cars. A few years back, Canada mandated that. Manufacturers started shipping cars that did that automatically, so that drivers wouldn't have to remember to turn them on manually. Given the easy border crossings, and the commonality between the US and Canadian car markets, most cars in the US now follow the Canadian convention, even without the matching law.

Bottom line, companies operating globally have to pay ever closer attention to the legal environment everywhere they operate.

I found this "I want to be alone" website to be truly amusing. I especially like the IMolatr :)

Here's one of the downsides of outsourced hosting that I'm not sure is solveable - if the thing being hosted is critical to your business, who cares about it more: you, or the hosting company? If there's a problem, who will be more highly motivated to fix it? That seems to be the sentiment behind this story from CNet:

In a meeting here with reporters on Friday, Gianforte said reliability and cost issues mean the company isn't interested in managed hosting services, including the $1-per-processor-per-hour Sun Grid.

He tried turning over his servers to a managed hosting company seven years ago, he said, and the move was a "miserable failure" that has since been reversed. Managed hosting companies want control over computers, but RightNow needs to be the boss in order to keep its equipment running around the clock. "We need control to get that kind of reliability," Gianforte said. Nothing has changed in the last seven years to change his mind, he added.

There's also the cost thing:

It doesn't make financial sense, either, Gianforte said. Running his own data center, including engineers and other staff, costs 6 percent of revenue, and he expects that to drop to 4 percent in the next two to three years. One of his top competitors, SAP, pays IBM much more than that to host its software-as-a-service offering, he said.

That certainly sets a target for any outsourcer who wants to get business, now doesn't it?

If you follow the Public Store, you're about to see a change in the RSS Feed. Back when I first started publishing that, the feed generator didn't produce a GUID for the items. Most aggregators simulate one under those circumstances, but the information that most of them use (often the link) wasn't going to work out so well here - the link for each item in this case is the same.

So, I've gone ahead and added a GUID. The downside is, on the first round of updates, you'll see a bunch of "new" items that aren't actually new - but it should behave better over time.

There are only 4 days left for early registration discounts for Smalltalk Solutions 2006 - advance registration discounts last until April 17. You'll want the early discounts - full registration pays for all sessions, including tutorials - like the "Using AJAX from Seaside" tutorial with Andrew Catton and Avi Bryant:

Why shouldn't your web apps be as dynamic as your language? This tutorial is for intermediate to advanced Seaside users who want to use client-side Javascript to add a richer user experience to their web applications. Learn how to add autocompletion, drag and drop, visual effects and instant feedback using Seaside's AJAX object model, usually without writing any Javascript code by hand.

See you in Toronto!

Murphee points out some of the difficult issues that Sun faces with JSR 292 - adding dynamic language support to the JVM. The upshot is, it won't be a small change:

Reading the overview of JSR 292 brings up another interesting planned feature: updating class structures at runtime. It‘s already possible to use HotSwap to update the contents of method bodies of loaded classes, but structural change of the class definition is impossible (causing the dreaded debugger dialog “Hot code replace failed”). One annoying restriction here is the it‘s not even possible to change method signatures or just add/remove methods from a class.

As he points out later, it's simple to do that kind of update in Ruby, and I've pointed out that I do exactly that kind of live update to the server code running this blog. Heck, I did that again over the weekend. I think the people responsible for the JVM are going to end up re-inventing everything we already did in Smalltalk, Lisp, and Ruby if they decide to take on 292. If they are true to past form though, it will involve a lot of really "interesting" syntax changes :)

Cees has a brief report up on the Smalltalk party that took place in Brussels over the weekend. Apparently, they are planning more of these in various European locales over the next few months.

Via Brad Wilson, I see that Ryan Tomasko ripped Gosling a new one over his clue free ramblings - about which Ryan said:

Minds changed. Respectful debate, honesty, passion, and working systems created an environment that not even the most die-hard enterprise architect could ignore, no matter how buried in Java design patterns. Those who placed technical excellence and pragmaticism above religious attachment and vendor cronyism were easily convinced of the benefits that broadening their definition of acceptable technologies could bring.

The people who are still unconvinced are those that just don’t care or are too lazy to spend a small amount of time researching and validating the arguments, which brings us back nicely to James Gosling’s recent statements.

There's a lot more as well, and it's all worth reading.

Update: More piling on here, at Developer Journals:

It really does seem that we're beginning to emerge from the 10 year long Java nuclear winter, when excellent dynamic languages such as Objective-C or Smalltalk were kicked out of the mainstream.

...

In contrast what underwhelms me about the design of Java is how it feels like it was done by someone who never really programmed in Smalltalk or Objective-C, and so they just left out any sort of introspection, left out meta-classes, left out dynamic method redirection via #doesNotUnderstand, and left out open classes that you could add methods to via categories. It was obviously designed by someone who had tried C++, found it over-complicated and designed a simpler alternative with a runtime derived from the Pascal p-code interpreter.

Read it all.

Gosling's outburst is continuing to demonstrate an "out of ammunition" posture at Sun. I spotted this retort to Gosling tonight:

Here’s a good measure of performance. A certain sorting algorithm that determines the most active conversations. 100ms in Java, but a painfully slow 500ms in Ruby (5x). Once you add database query and networking, the difference between the languages is about 10%. 10% is not a big difference.

About the size of it. Compared to dynamic languages, Java is a huge pile of premature optimization.

Ayende Rahien spotten my post on cooperative scheduling, and had a comment:

Cooperative threads relies on programmers' disipline to yield often enough to make sure that other threads are not starved. I can see several cases for doing this in SQL Server, since this can ensure that you'll not be pre-empted before you finished. I would guess that this is a good way to reduce locks in certain situation, since in this scenario you know that you will not be interrupted until you are ready. Windows 3.1 and Mac OS before X proved that this just doesn't work in the general case. A single ruoge application can take hold on the whole system.

Well, in Cincom Smalltalk, this model gives you predictability - you know exactly what a thread is going to do. The issue with runaway threads rarely comes up for a simple reason - most processes end up pausing for I/O (user input, db access, file access, sockets - what have you). That wait for I/O state is what prevents a problem from arising. Sure, if you create a CPU bound process, you can hose a system - I even blogged about that kind of issue here, in terms of a CPU bound process taking up too much time in this blog server.

What it looks like the SQL Server team spotted is something common to an awful lot of applications - CPU hogging is fairly rare, while processes getting into an I/O wait state are pretty common. As to worrying about thread management - I've worked with a lot of customer code, and I've dealt with application level threading extensively in Silt and BottomFeeder. With the exception of a few expensive operations in the Silt server, I've not had to devote a lot of thought to the problem. And even there, the problem had already been solved for me, by some Smalltalk library developer a long time ago.

We watched the finale of BSG and Stargate Stargate Atlantis last night. After watching Atlantis, it seemed to me that the story lines of SG-1 and Atlantis are going to have to merge for awhile - the way they ended left me with little doubt of that. Maybe Rodney will finally get his chance to one-up Carter? Assuming he and Ronon can get out of the bind they are in, that is.

BSG ended with a bang. The writers set it all up a couple weeks back with Six and Eight surviving the cafe bombing on Caprica - we saw how that would play out and it really does reset the series. I have to say, I have no idea how they get out of this one. On SG-1, there was definitely time travel foreshadowing, which provides an out. BSG doesn't play those tricks, so I'm left just wondering. I'm not displeased with their direction - but I sure hope that they haven't painted themselves into a corner.

Sounds to me like James Gosling hears footsteps, and he's getting nervous:

"There have been a number of language coming up lately," noted James Gosling today at Sun's World Wide Education & Research Conference in New York City when asked if Java was in any kind of danger from the newcomers. "PHP and Ruby are perfectly fine systems," he continued, "but they are scripting languages and get their power through specialization: they just generate web pages. But none of them attempt any serious breadth in the application domain and they both have really serious scaling and performance problems."

I'm sure he knows that due to the vast well of experience he has with Ruby, PHP (etc). Apparently, in Gosling's world, Ruby is only used for web pages. Sure James - and Java is only used for applets. How often do they let him out of his lab? His cluelessness abounds:

PHP (for example) is able to make things simpler because it's 100% aimed at web pages, Gosling explained. Whereas with Java, he said, "We have a balancing act: we need the simplicity but we also need power."

Hmm - I see that the phrase "best tool for the job" isn't part of his lexicon. Simplicity? In Java? Yeah, how about that implementation of generics, hmm?

David Heinemeier Hansson has related thoughts.

tech.memeorandum has taken notice of DabbleDB, in a story linked from TechCrunch. DabbleDB is Avi's latest venture.

Every so often, someone asks me why Cincom Smalltalk doesn't use native (platform) threads - the assumption being that lightweight (i.e., managed by Smalltalk) level threads just can't cut it. Someone should tell Microsoft, because they have SQL Server using the same kind of model that Cincom Smalltalk uses:

Ken Henderson profiles the User Mode Scheduler (UMS) in SQL Server 2000 that requires developers to write code that runs efficiently, and yields often enough in the appropriate places. UMS provides more control and allows the server to scale better than it could with the Windows scheduler.

Read that last sentence a few times :) Then have a look a little further down in the article:

An important difference -- in fact, probably the most important difference—between the Windows scheduler and the SQL Server UMS is that the Windows scheduler is a preemptive scheduler, while UMS implements a cooperative model.

Just like what we do in Cincom Smalltalk. Read the article, and then look at the Smalltalk process model we use in CST. You'll see a lot of similarities.

It's still all about Dave!

I had to decide whether this belonged under 'security' or 'stupidity' - I decided that the latter was far more descriptive. There's a report out on a fairly serious loss of data by Citibank (and a bunch of other banks) - someone hacked into a system and stole a bunch of card data - including the PIN numbers - for a set of debit cards. The stupidity is in this quote:

"This is the worst hack ever," Litan maintained. "It's significant because not only is it a really wide-spread breach, but it affects debit cards, which everyone thought were immune to these kinds of things."

Unlike credit cards, debit cards offer an additional level of security: the password-like Personal Identification Number, or PIN.

"That's the irony, the PIN was supposed to make debit cards secure," Litan said. "Up until this breach, everyone thought ATMS and PINs could never be compromised."

Who exactly is "everyone"? The dumber flacks in the PR department? And to cap it off, here's her non-solution solution:

"Security is tight at the ATM, but point-of-sale is a whole other story," said Litan. "Look at your [debit card] account on a regular basis, and don't use a PIN-based debit card at point-of-sale," she recommended. "I never do."

Yeah, I'd much rather carry my checkbook with me everywhere, like it was 1978 all over again. Another thing - a regular credit card at the POS won't be any safer from a fraud standpoint - if the stores aren't careful with the data, then it doesn't much matter.

Time for my weekly look at the logs: BottomFeeder downloads went at a rate of 261 a day last week, a nice little pickup from the prior couple of weeks:

Platform	BottomFeeder Downloads
Windows	706
Sources	328
Update	234
Linux x86	140
Mac X	123
Mac 8/9	75
CE ARM	55
HPUX	35
Windows98/ME	35
AIX	29
Linux Sparc	26
Solaris	26
Linux PPC	6
SGI	5
ADUX	3
Source Script	3
CE x86	1

Next up, the HTML pages accesses, by tool:

Tool	Percentage of Accesses
Mozilla	52%
Internet Explorer	24.9%
MSN Bot	11.4%
Everest/Vulcan	4.3%
Megite	3.4%
Google Bot	2%
BottomFeeder	1%
Other	1%

Looks mostly the same as always, except for that MSN bot. That's a lot of crawling, IMHO. Off to the RSS tool accesses:

Tool	Percentage of Accesses
Mozilla	24%
BottomFeeder	16.2%
Net News Wire	9.1%
BlogLines	8.6%
Other	7.3%
Safari RSS	5.4%
MSN Bot	3.3%
Google Feed Fetcher	3.2%
Internet Explorer	2.8%
RSS Bandit	2.6%
Planet Smalltalk	2%
Feed Reader	1.9%
Magpie	1.5%
BlogSearch	1.4%
NewsGator	1.3%
Java	1.3%
SharpReader	1.1%
JetBrains	1%
Liferea	1%
FeedFlow	1%
NewsOutlook	1%
News Fire	1%
Feed Demon	1%
Attensa	1%

Looks about like always there.

I find this whole mess around voting machines (in Maryland, where I live, and elsewhere) just fascinating. The Maryland state legislature just voted to ditch the touchscreen machines (they just went in over the last 2 elections) because they couldn't produce a paper trail:

The state House of Delegates this week voted 137-0 to approve a bill prohibiting election officials from using AccuVote-TSx touch-screen systems in 2006 primary and general elections.

The legislation calls for the state to lease paper-based optical-scan systems for this year's votes. State Delegate Anne Healey estimated the leasing cost at $12.5 million to $16 million for the two elections.

The whole voting machine mess goes back to the 2000 presidential election, which - like the 1876 election, was disputed. In the aftermath of that, laws were passed in an attempt to apply a technical fix. This ended up looking a lot like the sort of management fad that blows through IT shops from time to time. You know the sort: "Everyone else is using Java, we have to!", or "XML is the best thing ever, we should use it everywhere!"

So here in Maryland, the state grabbed the touchscreenn machines in 2002, thinking that here was the technical fix for any potential problem. Well, not so much. The machines were different from the old optical scan system we used to use - better in some ways, worse in others (a lot like those management fads). Which takes us to the present - like the IT shop that decides to drop everything and start using (insert fad here), the state finally looked at the level of complaints (no paper trail) and the amount of money spent (a lot).

And where are? Right back where we were in 2000, with Optical Scan machines. In the technology industry, we call this kind of thing "management by magazine". This episode shows that the transmission method might be different (warring press releases from political advocates), but the end results look a lot alike - trying to fix a supposed problem by applying a "new technology" - without doing an actual evaluation of the costs/benefits - usually ends up failing. Badly.

Oh my. I haven't watched the finales of Atlantis or BSG yet, but I just got through the season finales of SG-1. I won't give any real spoilers, but for you Star Trek fans - it reminded me a lot of "Wolf-359". My thought at this point? Where the heck do they go from here?

I spent the day bogged down in a syncronization exercise. Steve Kelly has been making updates to the Silt server, and we were trying to create a unified version that would load the same on 7.1 and 7.4. The issue? I extended the DES class, and between those two versions it migrated between namespaces. This made life difficult. For now, I'm back to maintaining parallel versions. However, the server will be moving to 7.4 soon, and I'll be able to kill off the older version.

The good news is, I'm off to play Civ IV and Caylus this evening!

Via Instapundit, I found this interesting story on the crossover between video games and military action:

But there's another reason, not often talked about, for the success of CROWS. The guys operating these systems grew up playing video games. They developed skills in operating systems (video games) very similar to the CROWS controls. This was important, because viewing the world around the vehicle via a vidcam is not as enlightening (although a lot safer) than having your head and chest exposed to the elements, and any firepower the enemy sends your way. But experienced video gamers are skilled at whipping that screen view around, and picking up any signs of danger.

So maybe all that time on things like "Call of Duty" aren't just entertainment - I can see my daughter's retort to get off the GameCube already: "But dad, I'm doing this for the good of the country!"

I have to take this announcement from Microsoft - that they will not be supporting EFI bootup with Vista, and will instead rely on the creaky old BIOS - as something of a good sign.

Why? Well, stop and think about it for a moment - it means that Microsoft actually considers Apple to be a viable competitor. The only reason for them to make dual boot a pain is to keep Apple software off of systems that have Windows. They realize that dual booting into Linux is irrelevant - Linux on the desktop is a non-starter. Apple though? Clearly, they think that might be a problem.

I get Jonathan Schwartz' complaint about the default use of Windows by so many organizations - both public and private. On the other hand, I can deal with Word docs on my Mac without a problem, and my Mac also handles Windows Media player files just fine. Interestingly enough, my Mac is a better citizen on my local LAN (from a Windows file sharing perspective) than the other machines are; one of my wife's machines sometimes just disappears from the LAN without losing connectivity otherwise. Go figure.

I suppose I should get to the ironic part - Schwartz is out advocating for the ODF format. Meanwhile, Corel just backed away from ODF in WordPerfect (reported by ComputerWorld this week). If Microsoft wants to really put egg on the face of the ODF backers, they'll ship support for the format in Office 12. Now that would be amusing.

Travis Griggs explains how Smalltalk is simpler - in looking at some new C# features, he says:

First of all, I'm amused that people continually get more excited about language feature creep than better libraries using the language. It's always "bolt some more syntax on."

...

Variations that use arrays and or optionally send messages ad nauseum. It's all in the messages. No new language feature necessary.

Anonymous types are next. Apparently, they're missing ad hoc structure defs. What to say. I'm sorry. This kind of stuff just isn't a big deal for us. I could make a proxy object that would do something like what it does. Again, it's all in the objects and messages. No syntax necessary.

Here's news from STIC:

Dear Professional,

Fantastic Early Bird rates are on for LinuxWorld and NetworkWorld (LWNW) Conference & Expo--but only for another week!

LWNW is a unique conference and tradeshow that gives you more ROI than any other event of its kind. And until March 17th, you can get even more value: up to $200 off on-site admission rates.

The 3-day conference of 80+ sessions is completely non-commercial. All paid sessions are presented by authors, analysts, IT consultants and end user IT specialists, demonstrating real-life examples and case studies to ensure you get the maximum educational experience. To read more about the conference details, speakers and sessions, visit http://www.lwnwexpo.plumcom.ca/conference.cfm.

New this year: We are proud to be hosting the Smalltalk Solutions Conference. In addition, we are offering the ITIL Foundation Certification course with bonus admission on the last day of the event.

Be an Early Bird now--customize your curriculum later. Even if you haven’t picked your sessions yet, you can still register before March 17th’s Early Bird deadline to take advantage of huge savings. Your itinerary can be selected or modified until the event date.

Tradeshow registration is included with all conference packages. Just want to see the tradeshow and keynotes? Pre-register today--no charge for admission and you’ll avoid the line-ups.

If you have any questions about sessions or what the best package is for you, please contact our hotline anytime at 1-888-823-7586 x211.

STIC members don’t forget to contact Suzanne Fortman for your 25% discount code.

See you there!

STIC would like to have Smalltalk Solutions be the venue to help connect Smalltalkers looking for work with employers looking for Smalltalkers. If you are an employer looking for staff, then we want to help - you can send your information to Suzanne Fortman, who will coordinate for STIC at the show. What we'll need:

Company Name (or appropriate contact info if that's not feasible)
Opportunity Location
Position type: Fulltime or project, expected duration if the latter
Preferred Smalltalk Dialect(s)
Relocation Required, assistance availability
Specifics of the Opportunity - short summary
How to respond (email/phone/web)

All of that will end up at the STIC booth at Smalltalk Solutions, and on the general LW/NW conference board.

I came across this post from EZBoard's CEO this afternoon - Ezboard is currently built in Cincom Smalltalk, as you might be aware (there's a success story on our website). In the post, he talks about a new application architecture they are building, and in that discussion, he said the following:

ezboard was built on an older technology that has reached its end of life for web site development. It is time to replace ezboard's software platform so that we can bring you the new features you want while the improving security and stability of the service. It is not possible to attach new technology to ezboard's software in a reliable way. This is why we are building a completely new message board platform called Yuku.

Well, I need to address that statement, because it conveys a false impression about Cincom Smalltalk. Our current release of VisualWorks (the product they are using) is 7.4 - you can see the release page here. As you look through that, you'll see that we have a full platform suite supporting many things, including web development. Heck, the site you're reading this on uses Cincom Smalltalk as the engine, and I'll be giving an experience report on the technology at Smalltalk Solutions.

EZboard started using VisualWorks back in the 3.0 timeframe. Back then, the non-commercial product (which is what they started with) did not include the web server frameworks, which was, quite honestly, a mistake made by the previous owner of VisualWorks. The upshot is, they built their own HTTP application server framework, and their own object storage framework using serialized object files. I can well believe that it is difficult to move forward; this post I made yesterday has a lot to do with my own issues in carrying forward an application I wrote 3 years ago that uses serialized object files for storage. Today, I'm in the midst of an exciting data scrubbing mission based on that application's architecture.

Bottom line, there's nothing EOL about Cincom Smalltalk, or any application built with it.

Via Chris Petrilli, I came across something that made me laugh out loud - Go read Amy Hoy on her more recent issues with Internet Explorer. I've not pushed the edge on browsers the way she clearly is, but I do run across oddball issues with the CSS we use on the blog server from time to time. Read the whole thing, but this is just priceless:

And yes, we are both using the exact same version of IE, right down to the millionth decimal and little spatters of text indicating the XP service pack revision. (I'd never really thought it possible to make unusable version numbers before, but Microsoft proves, once again, that you can achieve anything if you just try hard enough.)

You'll have to visit the post to read the background for that comment :)

Remember AutoLink? You might have forgotten, because after many months of overheated hyperbole, it's still part of the Google toolbar, and the net hasn't devolved into the morass of rewritten links that some people were absolutely convinced was about to happen. Here's an example of the overheated rhetoric that was being deployed - The Register write an article titled "Google AutoLink: Enemy of the people?" That article was written in March, 2005.

Things have gone awfully quiet since then - almost as if the loudest objectors noticed that it was not, in fact, the end of the web as we know it, and decided to stop writing on the subject. This illustrates a problem that blogs and online media share with their older cousins in print, TV, and radio - it's very easy to declare disaster, create a blogswarm of posts agreeing that "something must be done". After awhile, people start to notice that the sky isn't falling, so - in their best Emily Litella voice they mumble "never mind" - and move along to the next pseudo-disaster.

The main difference on the web is that dissenting voices get a chance to object - something which is mostly not possible when the mainstream media has one of their drive-bys.

Sci Fi Wire reports the good news!

SCI FI Channel announced that its hit original series Battlestar Galactica will return in October with a full 20-episode third season.

Tim Bray posted a brief set of comments on Ray Ozzie's ETech talk:

He’s pumped about RSS as “the DNA to wire the web”; the connective tissue between active websites. He talks about “Composite Applications”; for example, Unix pipes are weaving together a composite app. Claims < is like “copy”, | like “cut”, and > like “paste”. Then the question arises: “Where is the clipboard of the Web?”

The demo was “Live Clipboard” (integrated with Windows clipboard). “Great way to bridge from Web to PC”. Smart, structured, tagged data on the clipboard. You can paste in a feed object, which then remains dynamic.

Granted, it was pretty well pasted-together vaporware. But the idea might have legs.

I have to admit to being a huge skeptic about the mashup idea. Sure, there have been a few interesting demos on the web, and I'm sure that some mashups might even be useful. However - it's still a house of cards, as far as I'm concerned. Why?

• Who's going to create a mission critical application that depends on some foreign (i.e., not under your control) component that is remote? What happens if it goes offline, or if the people controlling it change the API?
• Non-demo applications tend to have non-trivial APIs, which tend to require intense communication to hammer out. I don't see that happening amongst disconnected groups.

For some social applications like wikis and blogs, sure - this will work fine. For other stuff? I'm completely unconvinced.

Tim Bray on the "Attention Economy" at ETech:

Their answers took two forms: “I don’t get it” and “Yeah, sounds like you might be able to build some cool stuff with it”. I’m in the first camp: I freely admit to Not Getting It. But I can report some of the more compelling things that were said. I omit some presentations because I missed then and others because, as far as I could tell, they consisted entirely of vacuous hand-waving.

I have to agree on the "not getting it" part. Most of the verbiage I've seen spewed on Attention makes my eyes wander immediately. As Tim indicates, beyond counting links as votes, I'm not sure what really works.

Well, I finally managed to get through a task that I've had a mental block on all week. I was on the hook to prepare some documents for a set of upcoming meetings, and they wer by no means exciting :) Useful and important for the discussion, yes. Deadly dull to work on as well. I finally got them shipped out though, so I can get along to the data migration task that's sitting in front of me.

We are migrating the services on cincomsmalltalk.com to a new server, and part of that is an upgrade of our PostgreSQL database. As it happens, one of the applications I wrote against that DB (helpfully labelled "temporary", as I expected to need it for weeks - it's been deployed for over 4 years now) has some serious data problems. Tomorrow is data scrubbing day.

Smalltalk Solutions 2006 is coming up fast - speakers have to get their presentations in by the 15th of this month, and the first day of the show is April 24. Go register now, so you can attend presentations like "I have Nothing to Declare but my Genius" from Brian Foote:

With mainstream language design mired in ennui and retreating into formalism, the field has been effectively ceded to a ragtag, de-facto coalition of old-school dynamic stalwarts, scripting language designers, and ad-hoc domain specific API architects. A generation of research in this area can be distilled down into three overarching ideas, the rest is filigree. This talk will explore these ideas, examine how and why these currents are converging, and show why the large scale, dispersed, heterogeneous, polyglot world of 21st century computing demands nothing less than this degree of commitment to dynamism.

See you in Toronto!

Dave Winer points out a patent application by Apple that is truly, truly stupid - pretty much every extant aggregator provides an example of prior art for all or part of what Apple is claiming to have invented. I think it's past time to get patents out of software.

There's a long thread on whether to comment code in comp.lang.smalltalk - reading it, I think there's a bridge not being crossed by any of the proponents. The case for comments was made by Vassili on his blog here; read the cls thread, and you'll see plenty of the other side - this being a good example.

On the one hand, Vassili points out that comments are another form of communication between the current developer and future developers - and more communication is usually better than less. The caveat I have is the matter of bit rot. Over time, code gets changed for a variety of reasons. Maintenance, refactoring, what have you. The comments almost never keep up. What started out as a well intentioned comment with high communicative value often ends up as a misleading marker to the past. I have no idea how to fix that problem, either. The decisions to update code are made over a long period of time, and whether a given developer "has time" to muck with the comment is typically an ad-hoc decision.

Perhaps a development standard, taken on by the whole group would help. I don't know - if it were easy, that long winded cls discussion wouldn't be happening.

Bob Congdon, speaking about Steven Yegge's Tour de Babel post:

Note that Steve wrote this in 2004 which, based on Steve's estimate of expansion rate, means that Amazon may have over 100 million lines of C++ code by now. As a point of comparison, Notes/Domino R6.5 (a complex beast) was documented as containing just under 20 million lines of C/C++ code. Compare that to some other estimates of size such as Windows XP's 40 million lines of code. What exactly is in Amazon's 100 million lines of C++?

That's a brick wall being approached at a pretty high rate of speed, IMHO. Forget the language - 100 million lines of anything are simply incomprehensible.

Rogers Cadenhead discovers that some bumper stickers fall into the "be careful what you ask for" category :)

Looks like my cable modem has served it's last bits, or my signal suddenly got weaker for no good reason. I'm reduced back to *gasp* dialup, while I await the princely arrival of Comcast's technician - which won't be until Thursday. Oh, the joys of great customer service from the local monopoly...

Update: I should know better than to take the word of Comcast phone techs. He said that "there was no problem in my neighborhood" - which certainly explained the truck parked by the cable box up the street, with two technicians replacing a circuit board. The good news is, I don't need to wait for the service tech now.

Dare Obasanjo has a nice summary of a scaling talk given at ETech - lots of good stuff, but this bit at the end is especially interesting:

One big lesson learned about database scalability is that 3rd normal form tends to cause performance problems in large database. Denormalizing data can give huge performance wins.

Step one: Defeat the Architecture Astronauts :)

Boy, do I ever identify with this post from Joel Achenbach of the Washington Post:

My house is littered with partially read books, moaning for someone to put them out of their misery. I approach the basement library like a doctor assigned to perform triage. What book shall be given life, and picked up anew? And what consigned to oblivion? "The Known World," by Edward Jones, is one I'll finish, and "The Little Friend," by Donna Tartt, if I can just figure out where I put it. But already the stack of stuff I'm supposedly reading is getting so high that I fear it will fall on me. All that unread material may literally crush my spirit.

Well, "crush my spirit" is too strong - but I sure recognize the problem. My bedstand is littered with partially read books...

This was inevitable - astroturfing campaigns in the blogosphere, leveraging the meme trackers and Google. This one is political, but the same thing could easily be used for anything. Hat tip Scoble.

Hmm. Winer doesn't like Wikipedia because of the lack of editorial control. Now he doesn't like sites (like Yahoo) that use DMOZ (editors). The beauty of the web - as opposed to print media or TV - is that we as consumers get both, and can evaluate them against each other. The more the merrier, I say.

Torsten says:

What if they would take the next step and directly start using Smalltalk. I'm sure they will enjoy the wonderful world full of objects.

Come on in; the water's fine!

Who knew? The Russians pack heat on every space flight, as part of their post-landing survival kit:

“In 1965, two cosmonauts overshot their touchdown site by 1,200 miles and found themselves deep in a forest with hungry wolves. That's when Russian space officials decided to pack a sawed-off shotgun aboard every spacecraft. It took Russian search crews more than two hours to locate the spacecraft and another two hours for helicopters to get support crews to the landing site.”

Heh.

If you grabbed the latest development build of BottomFeeder, you may have noticed that the blog client suddenly disappeared from the plugins menu. That's a glitch in the way I did the build; it will be back when I do the next build. In the meantime, you can execute this code in a workspace (see the "System" menu) to bring it back:

| plugs |
#{RSS.RSSFeedViewer} ifDefinedDo: [:cls | cls
		registerPluginClass: BlogTools.PostingTool
		startupMessage: #openWith:
		label: 'Bottom Line'].


plugs :=  Array with: (RSS.RSSFeedViewer getSingleInstance class plugins last).
RSS.RSSFeedViewer getSingleInstance addPluginMenuItemsFrom: plugs.

To get that to execute on every startup, just save that into a file called ".btfrc" (without the quotes) in your BottomFeeder directory.

The VUB held a Dynamic Languages day in Brussels last month, and they have the presentations, and some of the demos available for download. They also have quicktime movies of the presentations available for download. For more info, check out the website.

Via Dave Winer, a link to this post from a woman who's tired of hearing "it has to be easy enough for my mother to use":

I'm so tired of people talking about how their mother wouldn't understand something. I've been hearing this for 20 years, and it's sexist and ageist, and wrong and unfair, and how about let's get rid of this offensive idea. I'd never say that about my mother, who has a PhD, and is pretty smart. I certainly wouldn't want to encourage her helplessness! At one point I leaned over to Tara Hunt and expressed this sentiment. Then I realized that she's a mom, and said so. I wonder how many mothers were in the room and how they feel about always being held up as the paragon of cluelessness.

It's not about being smart or stupid. It's about the level of usability that is typically not built into software. Recall this post I made on our excellent adventures with the Media Center PC we bought. My wife and I are both software developers, and were willing to go through the pain of setting that up. It's not that "my mother" wouldn't be smart enough; it's that she would actually be smart enough to call BS on the usability and send the thing back. I linked to this post from Doc Searls a few days ago, where he related the following:

Add to this the sad fact that audio/video sales showrooms are a confusing mess. One guy who works for one of the big-box retailers recently told me the return/swap rate on flat screens exceeded 50%, because too many people are baffled beyond endurance by the complications of hooking them up, and the results afterwards.

Are all of those people idiots? No, they just have better things to do with their free time than ponder a confusing mass of cables and ports. Software developers - and apparently, most hardware manufacturers as well (with the notable exception, IME, of Apple) seem to have a "the more the merrier" attitude about possible configurations.

The menu on the archive pages for the blog server has been slightly non-functional for awhile (it didn't start on the current date, and it didn't remember your selection). I had been ignoring that for awhile, and Steve Kelly went ahead and fixed it. Thanks Steve!

I've posted updated BottomFeeder development builds - these are under the "Dev" label on the download page. Bear in mind, this is early development, so don't grab unless you are willing deal with potential problems :)

The funny thing is, this article by Steve Yegge on programming languages and how (and why) different ones succeed and fail reads a lot like "The Thirty Years War" that I'm reading. One of the primary goals of many developers is not to use the best tool for the job at hand, but to stamp out heresy:

During those years, I wondered why Python wasn't as popular as Perl. It seemed like a much stronger language than Perl. That's just my opinion, of course, and there were certainly things I missed from Perl, so I'm not claiming that Python is the be-all, end-all of language design. But it seemed like the best thing out there.

Why wasn't it more popular? It seemed to be getting crushed by marketing forces -- by fiery-eyed Perl zealots who went around and gained converts, one at a time. Perl was acting like a virus, and spreading rapidly, while Python sort of limped along, growing much more slowly. Richard Gabriel, of course, had already pointed out that C and Unix were virus-like in his famous short essay, The Rise of ``Worse is Better''.

I'm not sure whether I should be amused or depressed, actually.

eWeek's Peter Coffee took a look at Cincom Smalltalk VisualWorks recently, and had a talk with Suzanne Fortman and I. He's got it written up in this week's issues, which should be out today. He has some nice things to say about the product, which he installed using the NC CD:

Version 7.4 of VisualWorks became available at the beginning of this year, offered in both a supported commercial version and an unsupported but full-function noncommercial version that can be downloaded at smalltalk.cincom.com.

VisualWorks offers developers a capable tool set with a remarkably long list of contributed components and utilities that are supplied under varied license terms.

Versions of the 7.4 release on Microsoft's Windows XP and Apple Computer's OS X delivered a polished development experience in eWEEK Labs tests on both platforms, with well-paved pathways for interactive graphical interface development and efficient packaging of finished applications for convenient deployment.

I got a couple of quotes, and BottomFeeder got a mention as an example of cross platform deployment:

After one of the Labs' past development "Shoot-Out" competitions, we reported that watching a Smalltalk developer recovering from an error was like watching someone perform brain surgery on oneself. But Smalltalk has since become somewhat safer for journeyman developers with the addition of such features as namespaces, as we learned from Cincom product manager James Robertson, in Cincinnati, during a conversation with the Labs late last month.

Robertson brings a developer's perspective to his work. He's better known to the open-source community as one of the lead developers of the BottomFeeder news aggregator client for the RSS and Atom protocols.

Very nice stuff. I'm located in Maryland though :)

At the moment, the 4.2 (development) download of BottomFeeder is pretty much "bleeding edge" - so if you download it, be aware that there may well be odd problems. I'm still adjusting to the sub-system framework in VW, and the RTP's usage of it.

Phillip Greenspun on Java as a choice for building web applications:

After researching how to do bind variables in Java (see the very end of http://philip.greenspun.com/internet-application-workbook/software-structure), which turns out to be much harder and more error-prone than in 20-year-old C interfaces to relational databases, I had an epiphany: Java is the SUV of programming tools.

A project done in Java will cost 5 times as much, take twice as long, and be harder to maintain than a project done in a scripting language such as PHP or Perl. People who are serious about getting the job done on time and under budget will use tools such as Visual Basic (controlled all the machines that decoded the human genome). But the programmers and managers using Java will feel good about themselves because they are using a tool that, in theory, has a lot of power for handling problems of tremendous complexity. Just like the suburbanite who drives his SUV to the 7-11 on a paved road but feels good because in theory he could climb a 45-degree dirt slope. If a programmer is attacking a truly difficult problem he or she will generally have to use a language with systems programming and dynamic type extension capability, such as Lisp. This corresponds to the situation in which my friend, the proud owner of an original-style Hummer, got stuck in the sand on his first off-road excursion; an SUV can't handle a true off-road adventure for which a tracked vehicle is required.

Not only are Java developers using the wrong tool for the job at hand, they are using the wrong tool for pretty much any job they might come across. It's an old column by Greenspun, but it's still highly applicable.

Phillip Greenspun addresses the "why aren't there more women scientists" via the more general question: why aren't there more scientists, period. He gives an example:

How closely does academic science match these criteria? I took a 17-year-old Argentine girl on a tour of the M.I.T. campus. She had no idea what she wanted to do with her life, so maybe this was a good time to show her the possibilities in female nerddom. While walking around, we ran into a woman who recently completed a Ph.D. in Aero/Astro, probably the most rigorous engineering department at MIT. What did the woman engineer say to the 17-year-old? "I'm not sure if I'll be able to get any job at all. There are only about 10 universities that hire people in my area and the last one to have a job opening had more than 800 applicants."

That addresses the problem of why so few people (in general) enter the field. It doesn't speak specifically about women though. All I have is anecdotal evidence - my sister went into aerospace (not academic - she got a job in industry straight out of university). The pay was good, the hours no worse than for any other professional (like, say, cs). She left the field completely after less than 10 years in. Why?

She had her first child, and decided that she couldn't stand having her daughter being partly raised by other people (day care). She and her husband ran the numbers, and decided that they could afford to have her quit and stay at home. Is this something all women do? Not hardly - a quick glance at the number of women working will tell you that. However, it's still a much larger number of women who downshift from work after having a child, compared to men. For the purposes of wondering why there aren't more women in some highly competitive fields, it doesn't even matter why that's the case - given that it's true, more men are going to rise to the top in those fields simply on a numerical basis.

Greenspun's examination of pay and working conditions do play a major role. If you're planning to have a family, pay does matter, and so do the working hours. My brother in law got himself a PHD a few years back in the bio-tech field, and we watched how he got treated (by his professor) while he was doing his research. Believe me, it wasn't well. The hours he put in were long, and the pay was beyond insulting - it came pretty close to serfdom, in my opinion. Greenspun elaborates on the pay problem:

Even a public schoolteacher actually does better than a scientist. Consider the person of unusual ability who takes that bachelor's in science and decides to become a schoolteacher instead of going to graduate school. At age 22, the schoolteacher is earning a living wage and can begin making plans to get married and have children. By age 30, when the scientist is forced to start moving around to those $35,000 per year postdocs, the schoolteacher is earning $50,000 per year. By age 44, when the scientist is desperately trying to switch careers, the schoolteacher is making more than $90,000 per year for working nine months (only the better school systems pay $90,000 per year, but remember that we posited a person with a high IQ and motivation sufficient to get through graduate school in science). Being a public employee and a member of a union, the schoolteacher cannot be fired but may at this point in his or her life begin thinking about a comfortable early retirement and some sort of second career.

I'd question his "living wage" for a teacher at 22 - I took that track back then, and I was making $14,500 a year at 22. I couldn't afford an aprtment of my own, and I was living in the New York suburbs, not in a big city. Even so, working conditions were better than what my brother in law faced, and - had I stuck with it - the pay would have reached a tolerable level well before I hit 30.

Bottom line, multiple things enter into the disparate numbers of men and women in fields like academic science. If I had to guess though, the working conditions of those first few years of doctoral and post-doctoral work have a lot to do with it. From the outside looking in, it looked a lot like what medical interns deal with at hospitals, only without the promise of better pay down the road.

So who does go into these fields? Read Greenspun's essay, and see what the decision making process looks like to someone from China or India (etc). Which explains why you find so many people from those countries in the sciences quite well, I think.

Read the rest of his article - I think he explains the disparity quite well in the summary part of his essay