Smalltalk Tidbits, Industry Rants

From the TSS Java Symposium:

Another tool, which I found was very impressive is CodeCrawler . Basically this tool is able to analyze source code (well not directly but that's not the point) and generates a graph of the classes found. Each class is linked to others and have color, width and height corresponding to number of line of code, number of methods and number of attributes. And the whole graph is dynamic, you can move things around. Pretty impressive... alas CodeCrawler in written in SmallTalk, but if you have to audit code, that's a killer tool.

I love the way he says "alas" - likes the tool, likes the way it works, cares way too much about the implementation. You might as well pull out Emacs and say "alas, not written in Java".

Scoble spotted this story about Boeing and Connexion:

Looks like Boeing is giving up on Wifi in planes because they've lost a billion and don't see that their investment will come back. Funny, didn't JetBlue just pay big dollars to add that to all their planes? 

I used the Boeing service on an SAS flight to Copenhagen and loved it. The problem wasn't with the Wifi. But there was a major problem elsewhere that'll keep people from using it: power.

My batteries in my laptop (and in most laptops I see on planes) last about two hours. Yeah, some models last four to eight, if you have additional "big" batteries. But most last about two hours the way you buy them out of the store.

Hmm. Most American Airlines flights I've taken over the last few years - both domestic and international - have had power at the seats. That hasn't been true for any other carrier I fly, and it's impacted my plans for long flights quite a bit. Any carrier that adds WiFi but doesn't add seatback power is just stupid - the phrase "lost investment" comes to mind. Scoble's right - why should I buy a network connection for an 8 hour flight if I only have 2-3 hours of battery time?

Friend of mine is looking for a job, and related this exchange:

[Person1] A fund nearby got in contact with me yesterday. They're looking for someone to write a portfolio/risk management system and heard about the one I wrote. They were very excited.
[Person2] person1: In Smalltalk?
[Person1] Until I told him I did it in Smalltalk. "Don't you think it's odd to write a system in a language like that?"
[Person1] That was his response

I love group-think. Offer to solve a problem, and anything that veers from the expected path raises flags. Sure, there are times that's a valid concern. But if you find a person to help you build a business - and you found them because of their previous expertise in the field in question - wouldn't actually listening to them make some sense?

Joi Ito talks about the loss of 6 months of information from his machine, and with it his World of Warcraft stuff:

I realize this may sound a bit high drama, but I'm sure I'm not the only one whose brain shuts down to almost all outside input during a broken computer incident. Now I'm running on a fresh install with very little baggage and it actually feels quite nice. This also means no World of Warcraft and possibly more blogging. ;-)

Based on the way I've seen that game discussed, there might be a case that it's approaching "Better than Life" from "Red Dwarf". I've never played, so that might well be over the top - comments?

James McGovern is a little over the top in his title for the linked post, but I think he touches on a good point with this anecdote:

My friends thesis was based on the fact that the enterprises who go down the outsourcing route tend to lower their expectations for individual consultant productivity when pursuing outsourcing arrangements. He stated once an American company has failed at attempting outsourcing to India, he gets to come in and pick up the pieces at a higher rate. He also mentioned that this allowed his 100% US firm to staff a lot lower on the food chain that prior to outsourcing. Clients generally don't do individual interviews anymore which has afforded him the ability to place less optimal resources on projects. In the past, he worked for one of the spinoffs from the big four consulting firms who had the notion of partner. While the partner would bill out at higher rates, they wouldn't necessarily bill 100% of their time to a client. He noted that the Indian outsourcing model had the same notion of a partner only that they stayed at a single client to work on relationship-oriented issues. He believes this is another opportunity for him to take folks who are losing their technical ability to not only make them billable but to do so at extreme rates.

I'd disagree that you can get away with sending sub-standard consultants in at high rates for any period of time. Clients will notice, and that will be that. On the other hand, the offshoring experience may well lower expectations, and James touches on that with this: "He noted that the Indian outsourcing model had the same notion of a partner only that they stayed at a single client to work on relationship-oriented issues." When the consultants are 12 timezones away, that's probably most of what gets done. There's not going to be any direct technical collaboration, nor is there going to be any communication between the developers and the end users. It's a complete return to the 1970's glass house of IT approach: toss the requirements over a wall, wait N months, and see what comes back.

That approach didn't work well back then, and I see no reason for it to work well now.

In an article titled "Java's new considered harmful", Doctor Dobbs describes Smalltalk:

The best example is an instance-controlled class, where the class writer wishes to restrict the number of instances. When the maximum number of instances is one, you have the Singleton pattern (described in Design Patterns by Erich Gamma et al., Addison-Wesley, 1995). Singleton is often used when the object corresponds to a unique item in the real world (such as the keyboard or computer itself), or when the object is the sole manager for some resource. Java's Runtime class is an example of the latter: Its single instance manages interactions between the Java virtual machine and native operating system.

In Smalltalk, simply override #new in your class, and have it hand back (or create if it's not there yet) the sole instance.

Sometimes only a few instances of a class are desired. Imagine a class representing ASCII characters. There are only 256 of those. If the class is immutable and doesn't carry any other information, only 256 instances of the class need ever exist--more would just clutter memory. For a more dramatic example, consider the Boolean class: Only two instances are ever necessary, one for True and one for False. In addition to saving memory, having only one instance per value also allows testing for equality using the "==" operator, instead of the slower "equals" method.

In Smalltalk, there are sole instance of True and False. Yes, there's VM support for that, but you can get the same effect in user code by overriding #new and #basicNew to not allow the creation of new instances.

For a more sophisticated and (currently) hypothetical example, consider the Integer class. There are many possible instances of this class -- too many to cache them all -- but we might conjecture that some values are more common than others. Probably Integer objects representing the numbers from -2 to 100 are more likely to occur than others. A clever implementation could cache objects for these numbers, saving the time and storage involved in their creation.

Heh. Smalltalk does this, with class SmallInteger. It's an optimization for smaller (2^^29-1) integers, with auto-promotion when you exceed that range. The article also points out the non-polymorphic nature of new(), and the various work-arounds used to deal with that. In Smalltalk, #new is just a method. Override it to your heart's content.

Nearly all the way down, the author gets to the nub of the matter - in Smalltalk, these problems simply don't exist:

In Smalltalk, where dynamic typing makes downcasting unnecessary and exceptions are not treated as rigorously as Java, class-based creation is convenient, elegant, and polymorphic (see SmallTalk-80: The Language and Its Implementation, by Adele Goldberg and David Robson, Addison-Wesley, 1983). In fact, it is the usual way to create objects in that language.

He describes all that a bit further down, and points to what he thinks is a flaw:

Smalltalk uses the class-based approach to creation. There is an object corresponding to each class, and instances are created by sending a new message to the class object. This is elegant and polymorphic. Indeed, it is common in Smalltalk programs to pass around a class object when polymorphic creation is required. But Smalltalk's mechanism suffers from a serious drawback: Initialization is not enforced. new creates objects but does not initialize them. The convention is to have an initialize method for each class that is called after new, but you can forget to call initialize, leaving the object in an invalid state. Why not combine creation and initialization by calling initialize from within new? Because new is polymorphic, so it must always take the same number of arguments (namely zero), whereas each class may require a different number of arguments for initialization.

Well, you aren't limited to #new. You can create your own class side creation methods (example: Point x: xVal y: yVal). As to initialization, I've had plenty of cases where I've not wanted my new object initialized, or initialized in non-default ways, so whether this is a flaw or a feature is a matter of opinion. I'm used to it, so I could easily be biased in the matter. Also, a new object wil all instance variables set to nil may not be invalid - that may well be the right starting state. It all depends on the use case.

In any event, it's a good article, and explores the problem. Read it yourself, and see what you think.

James Governor quotes Tim Berners-Lee:

This has to be the best first sentence to a blog ever:

"When I invented the Web, I didn't have to ask anyone's permission."

Oh to be able to say that. Al Gore eat your heart out. Heck- Dave Winer eat your heart out.

"Democracy depends on freedom of speech. Freedom of connection, with any application, to any party, is the fundamental social basis of the Internet, and, now, the society based on it.

Let's see whether the United States is capable as acting according to its important values, or whether it is, as so many people are saying, run by the misguided short-term interested of large corporations.

I hope that Congress can protect net neutrality, so I can continue to innovate in the internet space. I want to see the explosion of innovations happening out there on the Web, so diverse and so exciting, continue unabated."

He's almost got it, but then misses at the end. Note the call to have Congress protect the net. Hmm. That's worked out so well for radio and tv, hasn't it? You mark my words: If Congress passes a "net neutrality" law, it will turn the US based portion of the net into a "public utility". Once it's a public utility, then there will be a need to "protect" us from various bad things - after all, just like broadcast TV, anyone can see the net. So content regulations "for the children" will pop up - call George Carlin about the 7 forbidden words. The kinds of campaign (political) restrictions you see on radio and tv will hop over too - suddenly, any advocacy for a candidate will be an "in kind" contribution.

You want "net neutrality"? Then don't advocate for Congress to create it.

Update: Lessig gets it wrong too:

One clue to this Net Neutrality debate is to watch what kind of souls are on each side of the debate. The pro-NN contingent is filled with the people who actually built the Net — from Vint Cerf to Google to eBay — and those who profit from the competition enabled by the Net — e.g., Microsoft. The anti-NN contingent is filled with the entities that either never got the Net, or fought like hell to control it — telecom, and cable companies. (The one clear exception to this is Dave Farber, who has been described as the “Grandfather of the Net.” I’ve never understood either what that description could mean, nor have I understood how he gets from the premises in his argument to its conclusions. But to be fair, this is an exception to the rule I’m describing.)

I have no doubt that the backers of net neutrality have their hearts in the right place. What they miss is that things won't stay pure. Once Congress regulates the net in the name of neutrality, we'll shortly end up with an "internet FCC". At which point the backers of neutrality will wail that they didn't have that in mind at all - but it won't matter.

I've spent the morning getting the internal blog server up and running - the server was actually the simple part. I'm no Apache expert, and futzing with the ProxyPass stuff was something of a learning experience :)

Wow. The last time I examined my subscription list to look at how many feeds had enclosures, and how many items with enclosures there were, it was a pretty small number - 3 or 4 feeds, iirc, and less than 20 enclosures. Well. Seeing as how I'm sitting on a plane with nothing else to do, I figured I'd have a look. So I opened up a workspace in BottomFeeder, and then ran this code:

dict := Dictionary new.
RSSFeedManager default getAllMyFeeds do: [:each | | items |
	items := each allItems.
	withEncs := items select: [:each1 | each1 enclosure notNil].
	withEncs := withEncs select: [:eachItem | eachItem enclosure notEmpty].
	withEncs notEmpty
		ifTrue: [dict at: each put: withEncs size]].

That gave me a dictionary with 26 feeds as keys - a fairly decent size increase (my subscription list is about the same size - 317 then, 320 now). The number of items with enclosures has exploded: 547. That's a big, big difference from just a few months ago. Simply amazing.

I've considered the idea of podcasting myself, but I'm not sure what I could do that would be compelling - I work in Columbia, Maryland, and all my co-workers (and customers, for the most part) are remote from me. If anyone has ideas, I'd be open to them.

I ran across this interesting Podcast - it's a regular Ruby On Rails podcast, but this one spoke with Josh Susser, who came out of Xerox (not Parc) with a Smalltalk application background. There's some interesting talk about how Ruby is similar to Smalltalk, and in how it differs. I'd be curious to know whether Josh has looked at Smalltalk recently - I'd love to hear his take on Seaside, for instance

It looks like AOL's management is having their staff actively engage in anti-marketing.

First up: The "why can't I just cancel my service?" call.

Second: The "Dead? Like that's a valid reason to cancel" call.

This is negative PR on wheels.

We head back home this afternoon, so we went to the beach before heading out - these two shots were taken this morning, at low tide:

As you can see, we pretty much had the beach to ourselves.

I might need to create a separate topic for this. So it turns out that changing the default view for reading mail (at least in non-inbox folders) is an option in the View menu. Yes kids, I'm blathering about Outlook again :/

Today's fun job is trying to figure out how to get the Exchange server our IT group uses to not retain mail on the server. I like to just keep mail around - disk space is cheap. There are limits on the server side storage though, which makes sense across a wide user base. So - the simplest thing would be to turn server retention off. That's easy if you use standards like POP3, but it apparently requires stepping through a forst of modal dialogs if you made the mistake of using the MS proprietary Exchange services.

It also seems to be painful if you already created a set of folders - the steps that my IS group is telling me to take just don't seem to exist in the dialogs that I'm getting. MS Outlook - it's teh Sux0r.

Charles Miller on the elusive ease of use of Linux on the desktop:

Around the Atlassian lunch-room table, during a discussion of the relative merits of operating systems, it was noted that Linux-use around the office seemed to correlate quite strongly with having a beard -- something that was certainly true of the small sample present at the time -- and that traditionally, the length of one’s beard is an indicator of Unix expertise.

“You know why that is?”, Mike asked. “The rest of us actually have time to shave.”

That's about the size of it. I much prefer managing applications on a Linux server, but I just can't see myself fighting with Linux on the desktop.

Cincom's Cincom Document Solutions group recently gave a demo at the Acord Loma Forum - there's video online for that here. That's Nic Carter giving the talk

Update: It seems that the direct link above doesn't work right. Here are the steps to get to the demo:

1. Follow the link
2. In the "Channels" list on the lower left, select "2006 Demos"
3. Scroll to pages 9-12 (control for that to the right of "Channels"
4. Select "Cincom" in the list.

Christine Rosen of TNR has a review up of "An Army of Davids" - and boy, she doesn't like it. The review is a hack job of the worst sort though - the "I'm an elite journalist and you're not" thing shines through like a fog light:

Glenn Reynolds is an unlikely visionary. Before he emerged as the "InstaPundit," he was just a law professor at the University of Tennessee, writing on administrative law and the Second Amendment for publications like Law and Policy in International Business and Jurimetrics.

Step one: denigrate the author as a small time hick from nowhere-ville. This is the oh so predictable response from *cough* professional *cough* journalists. Quick translation of her point here: "Pay no attention to this guy. He's a professor from Tennessee, for gosh sakes!". That works so much better than engaging his actual points...

She then goes on to make the claim that the book is all about blogger triumphalism:

Like many of his cyber-colleagues, Reynolds believes in a form of triumphalism: that his medium has transformed the exchange of ideas and information. When this triumphalism appears in the course of a 200-word blog post, it seems remarkably plausible. As bloggers never tire of reminding the world, they brought down Trent Lott and Dan Rather and powered Howard Dean's ascent. But, at book length, as the ideology's core assumptions and convictions are laid bare, the idiocies and dangers of this triumphalism become all too apparent.

The book is about trends in technology in general, and it hops all over the place - so this makes me wonder just how carefully Rosen read the book. There's blogger triumphalism to be sure, but it's actually a small part of the book - Reynolds makes the point that technology is empowering people across the board, with blogs and the internet being one piece of that. Her verbal sneer became clear in this section:

The little guy can be a poet or a pop star; with technology as his handmaiden, anything is possible. But this follow-your-bliss vision of individual fulfillment has little patience for the standards necessary for judging genuine talent, which is why Reynolds's book reads more like a middle-aged hobbyist's utopian manifesto than a blueprint for cultural renaissance.

No patience for us paeons from Rosen - no, we should stand back and let our "betters" decide on what's popular, what's trendy, and what works. Best not to get in the way - we might strain ourselves, or something. I've seen the kinds of people behind the "standards, Christine - they're the bozos at the RIAA and the MPAA, and thanks - I've had enough. Give me the amateur who isn't trying to outlaw future technology any day, any time.

Now, I have to say that I thought Reynold's interview of Kurzweil was too wide eyed, so I actually agree with Rosen there. I was able to get past that bit of triumphalism though, and see the wider picture Reynolds was painting. In her seeming desire to keep the hoi polloi away from the keyboard, Rosen utterly missed that. That's very clear with her closing paragraph:

Like his fellow techno-utopians, Reynolds dismisses criticism of these ideas as "the usual skepticism regarding the new." As long as individuals control the technologies, we should welcome them, he argues. As for the risk of catastrophic unintended consequences from our use of these technologies, Reynolds is sanguine. "You'd better hope that I'm right," he says, chuckling. "It's basically an unstoppable phenomenon."

There's the call for the amateurs to step back, and leave it to the *cough* professionals *cough*. Sure, there's a plan. I think I'll take my chances with Reynolds and the amteurs - the "pros" simply haven't demonstrated that they have better judgement.

Patrick Logan links to Steve Loughran talking about CORBA:

Gnome is built on Corba. You can tell that by changing your hostname and noting how you can't start any apps by double clicking on the icons. That is an ORB at work, if ever I saw one.

And then finds Miguel de Icaza thinking RESTful thoughts:

I have been considering the implementation of a system that would replace D-BUS and CORBA with a simple HTTP framework.

Heh. We tried that - SOAP. With any luck, take two will be smarter. On the other hand, if it runs true to form, it will start out simple and then end up like (insert your least favorite ditributed RPC system here)...

I have a hot tip for the mother here: how about you actually pay attention to what you kid is doing online?

A 14-year-old Travis County girl who said she was sexually assaulted by a Buda man she met on MySpace.com sued the popular social networking site Monday for $30 million, claiming that it fails to protect minors from adult sexual predators.

The lawsuit claims that the Web site does not require users to verify their age and calls the security measures aimed at preventing strangers from contacting users younger than 16 "utterly ineffective."

The best "security measure" possible would have been the mom actually paying attention. Computer and net access are no different than other activities - monitoring usage is no different than monitoring offline activities.

I just moved over to using Outlook, and I'm recalling why I stopped years ago. Eudora is hardly perfect, but it doesn't have as much baked in stupidity. Consider the default view in the inbox - just the title of the message (possibly with a small preview), and no immediate view of the message. I created some folders and rules, and messages display in this odd two pane mode - with the list being horizontal and left of the message pane - instead of above it. Not to mention that neither pane seems resizable in the vertical direction. Hell, selecting a message doesn't mark it read - I have to do that by hand. - Correction - there's just a slight delay, at least in non-Inbox folders. In the inbox though, it's manual. Bah.

This is actually unproductivity software. Which begs the question, why am I using it? Well, our IT group supports Outlook only - and I don't see any reason to run two email clients on an ongoing basis. Thus, all my mail is now organized there. I'd love to go back to Eudora, simply because it sucks a lot less.

Oh, and a note to Scoble and his cheering about how daring it was to introduce that stupid ribbon? Good gosh man, have the Outlook team make the rest of the experience not suck. When you define UI trivia as daring, while leaving the suckage setting on "infinite", you haven't achieved much.

It's a whole new Dr. Who cast:

London actress Freema Agyeman (the U.K. soap opera Crossroads) is rumored to be in line to replace Billie Piper as the companion to David Tennant's Doctor Who in the show's expected third season, the British Sun tabloid newspaper reported.

I realize that this is all in the spirit of the original, but I liked the first season's cast a lot. I hope the new one measures up.

Some people - even though they themselves have profited heavily in the entrepreneurial game - seem to think that the entire world should be rearranged to suit them. Witness this, on ad supported RSS feeds:

I decided it's time to unsub from feeds that have those annoying little web bug graphics in each item, masquerading as "useful" stuff -- they're really there to send messages back to the publisher that the item was displayed in my browser.

This means that every time I read the newest stuff, my machine sends hundreds of stupid messages to ComputerWorld, Fast Company, WNYC, USA Today and the Christian Science Monitor and other sites with these bugs.

Enough. I'm on a campaign to eradicate these Feeds Of Evil from my subscription list.

Well. I've said before that the advertising model is running into problems - click fraud on the one hand, and freely available opinion pieces on the other. However, that's not to say that advertising is dead, or that it should be. I do this via corporate sponsorship, for the purpose of promoting our product. Not everyone is in that boat though - not everyone is involved in product promotion, so they don't get a "free" billboard to post on. For those people who need to directly foot the bill for hosting content, what are the options? There's pretty much two: ads, or subscription fees. People are going to pick one or the other, depending on their target audiences. Decrying the whole thing as evil is just stupid. Without some form of sponsorship, a lot of the content out there on the web would just disappear.

That doesn't mean that every form of advertising is good, or useful - some of it (those irritating fly-over ads come to mind) does actual harm to the brand in question, IMHO. In any case, the ad model is not going to go away.

We headed down to the beach for a bit this afternoon; since then, I've been fighting with a new server configuration. The beach was much more pleasant. The sun was out, and the tide was high - here's today's view north:

And here's the view south:

It was a very nice afternoon. Now back to my server configuration...

Runar Jordahl explains it all.

I think the PS3 early offer price might be a tad too high:

The next generation consoles are on offer as "pre-orders" to Europeans for £550 ($1,020) at online shop, play.com.

Hmm. The lower end XBox 360 is $299. The announced pricing for the Wii is $250. So for the cost of one PS3, you could by two XBox 360 units, one Wii, and have $150 left over. Unless they can get that price down, I think Sony has a problem.

Dave Winer decries lock-in, and Nick Carr effectively rebuts his argument. Here's what Carr says:

It may not be honorable, but as far as "ways to win" go, lock-in is actually extraordinarily sustainable - much more so, in fact, than features, performance, and price, which all tend to get neutralized more quickly than lock-in does. Many of the greatest franchises in the history of the computer industry, from the IBM mainframe to Windows and Office to HP's ink cartridges to eBay to the iPod and iTunes, have been sustained by lock-in. And that's going to continue to be true.

Printers are certainly a great example of this - vendors make them cheap, and then hook you on their cartridges for the life of the printer. They certainly make many times the cost of the printer on cartridge sales, and your choices are pretty limited - there's no such thing as a universal cartridge - even though such a thing would be fairly easy to create.

There's lock-in that developers and their managers deal with all the time, too - their languages and development tools. Once a shop picks a language, it tends to happily stayed locked into it - even in the face of evidence that another language might work better for some particular problem. Various forms of inertia make the developers happy to stay locked in.

James McGovern asks about IT enterprisey types, and what people like me would do:

The comforts of a corporate job cannot be underrated. The benefits are there for a reason - to entice smart and productive people like you to stay and help the company generate more profits. Why do you think that crack dealers give young kids free samples? Don't get it twisted as I am not saying that IT executives are akin to crack dealers, but you have to admit that their means of persuasion are remarkably similar...

I wonder folks such as Chris Petrilli, James Robertson, and other entrepenueral types could help provide guidance on how to detox ourselves? I wonder if they were ever enterprisey folks themselves at one time but simply don't want to admit to it?

Well, I've never been in that position, or even close. I was a line developer for a few years, and then a trainer/consultant for a few more. After that I went into technical sales, and now I'm a product manager.

One thing I noticed during my time as a consultant is the "external halo effect" - it was frequently the case that internal developers at a shop I'd visit would know what was wrong, and how to fix it - but none of the management or architects/IT standards types would listen to them. In I'd come, with far less expertise on the problem at hand, but with the "external vendor expert" stamp. The amazing this was, I'd say the same thing that the internal guys did, and get listened to!

I've spoken to plenty of other people who are (or have been) in similar situations, and they report the same thing. So what advice would I give McGovern (et. al.)? Pay a lot more attention to your development staff. Stop listening to industry analysts as a first and only source of wisdom (which is not to say they don't have any - it's to say that you may well find the same thing for free in your own shop).

I went out on a paddlewheel boat cruise this afternoon with my parents and daughter. It was a four hour event, a few miles noth on the river, and then back. We started in Cocoa Beach. They served lunch and had a live band - they were a little too loud, but pretty good. Here's a picture I took on our way back, as we approached the BeeLine causeway:

I took this one of Merrit Island - the other side of that is the ocean:

All in all, it was a nice Father's Day out on the river.

Doc Searls recounts a few salient points about Flicr, and how their attitude toward the user base has been critical. The bottom line:

Here's the problem, and the opportunity: Every vendor involved in this - Flickr/Yahoo, Zoomr, Tabblo, and so on - will live and die by their relationships with their users and customers (and not just by customers alone).

He gives a brief example (by way of a link) to an example of not getting this - Home Depot:

Home Depot has delivered superb financial numbers in the past five years, with total sales growing an average of 12% per year and profits doubling. But the share price has dropped 24% during the biggest home improvement boom in history. And shoppers are getting grumpier. The University of Michigan's annual American Customer Satisfaction index shows Home Depot slipped to dead last among major U.S. retailers, 11 points behind Lowe's. Home Depot employees, who were encouraged to "make love to the customer" under co-founders Bernard Marcus and Arthur M. Blank now sometimes treat them like bad dates. "I don't want to say one bad apple spoils the bunch," says Curt D. Bridges, an electrical engineer from Decatur, Ga., who used to be a die-hard Home Depot fan. "But sometimes some [store clerks] almost blow you off."

Nardelli's strategy to expand into the contractor supply business, while cutting costs and streamlining operations in 1,816 U.S. stores, has pushed customer service down the company's priority list. Many full-time workers have been replaced with part-time employees, who now make up 40% of store staff. Meanwhile, workers' incentives for good customer service have dwindled, too. The profit-sharing pool for workers shrank to $44 million, down from $90 million the prior year, despite record sales.

An awful lot of "legacy management" thinks that it's all about the size of profit margins right now - so cutting costs is always a good thing. The trouble is, if those cost cuts make the customer experience worse - and if there's a competitor out there who recognizes the problem - then your cost cuts will end up being profit cuts as well. There are very few businesses that just don't have to care what the customer thinks (your local DMV comes to mind). There's no one willing to spend extra money like a customer who feels like he's been doted on. Forget that, and your "streamlining" may lead to unexpected problems.

A few groups inside of Cincom have expressed an interest in blogging on the intranet (i.e., the private Cincom network) - so I've arranged with one of the IT guys to set aside some space on a Unix server. I have the basic server ready to roll out - we should be able to get going next week. With any luck, this will help our internal communications as much as this site has helped the external ones.

The weather was perfect for a beach outing today, so after 18 holes of golf, it was off to the water. Here's a shot looking north:

The current was running south today, which is unusual. The waves were bigger than last night, and there were surfers out. Here's a view looking south:

There's almost no development on the shore going south - I jogged a good ways down that way last night. Finally, here's the waves coming in, and my daughter running out to jump into them:

It was a great day out there.

This is another shot, but by my daughter Victoria, who has a much better camera (5 megapixel) than I do:

This is looking north - on a clear day, you can see the coast jut out a bit at Cape Canaveral. If you look at the dunes on the left, you can see a segment at the bottom that looks dug out. My dad says that happened during a nor'easter type of storm during the early spring. If the space coast gets hit by a serious hurricane this summer, a lot of the houses on the beach are going to fall down - they are just way too close to the edge of the dunes.

Time for my weekly look at the logs - BottomFeeder downloads went at a rate of 152 a day:

Platform	BottomFeeder Downloads
Windows	338
Linux x86	149
Mac 8/9	107
Update	106
CE ARM	105
Mac X	85
Windows98/ME	45
AIX	30
Solaris	29
HPUX	26
Linux Sparc	17
Sources	12
SGI	6
Linux PPC	6
ADUX	3
CE x86	1

Off to the HTML Page accesses:

Tool	Percentage of Accesses
Mozilla	64.1%
Internet Explorer	27%
Other	3.4%
Opera	1.7%
MSN Bot	1.7%
Megite	1.1%
Jakarta	1%

Looks like Mozilla's numbers are back to the normal distribution for the site. Finally, the RSS accesses:

Tool	Percentage of Accesses
Mozilla	20.3%
BottomFeeder	18.6%
Other	16.6%
BlogLines	9.1%
Net News Wire	7.7%
Safari RSS	4.1%
Internet Explorer	4.1%
Google Feed Fetcher	3.7%
NewsGator	3%
RSS Bandit	1.6%
SharpReader	1.5%
Planet Smalltalk	1.5%
BlogSearch	1.2%
JetBrains	1%
RSS 2 Email	1%
Jakarta	1%
Feed Reader	1%
Java	1%
Lilina	1%
Liferea	1%

And that's a wrap for another week.

We headed out to the beach last night for a walk - it was a pleasant enough evening, and the threat of rain never showed up. Here's a shot of the water:

I also took this one, looking north into the distance - that's my dad walking ahead of us:

It was nicer out than it looks - it was close to dark, which is why the colors are odd.

Steve Rubel notes that real world stores are opening virtual world versions of themselves:

Advergaming in games like Second Life is beginning to really take steam. The 3pointD blog has discovered that American Apparel is opening a virtual store inside the role-playing game. American Apparel has 130 physical stores worldwide, including over 30 here in the US. In-game ads - which this is akin to - will hit $400M, according to CNET.

There are people making real money by selling virtual real estate in Second Life; I wonder if this outfit will be selling virtual clothing for real money?

John Collins draws some nice attention to the ValueModel framework in VisualWorks:

ValueModel was part of their VisualWorks application. As far as I know, that was the most highly evolved MVC-oriented GUI application programming framework devised in Smalltalk.

If you download the Application Developers Guide and do a search on ValueModel - you will get treated to some nice diagrams and code. Not a lot, but they are the right ones to examine. There is a page with all the VisualWorks documentation, if you want to browse everything.

It really is a nice framework, and applicable (as John explains) beyond Smalltalk.

Doc Searls on Net Neutrality:

The problem is, I like Sirius. I like the fact that I can hear Steve Gillmor, yours truly and other people I know on Sirius 102, right above Howard 100 and 101, where I get to enjoy Howard Stern, Scotty Ferrall and other exiles from terrestrial radio, where they were fined and censored off the air by the FCC, which found them "indecent".

Sirius, however, is a silo. It's private. It's not public. It's spaces are finite, but they're far more wide-open that the public airwaves have been ever since Congress, the Supreme Court and the FCC all agreed, long ago, that the First Amendment doesn't apply there . It's just the container cargo we call "content" (check the file name in the URL at the last link) and it can be regulated and controlled just like we do with trucking.

See, the private zones are relatively free, while the public ones are not. And private zones are the ones the carriers want clearance to create.

He makes the point about Sirius, which I think is crucial: The "public utilities" we have in the "content" area are heavily regulated, and free speech on those systems is limited (in multiple ways). If we regulate the net in the name of "neutrality", we'll end up with a new FCC for the net (or the existing one with new ground to regulate).

Thanks, but no thanks

David Rubinstein of SDTimes says that we should trust the professionals in media rather than bloggers. Let me start with his summary:

As more unnamed, untrustworthy writers enter the blogosphere, they will actually drive readers back to the traditional publishing sites, where the George Wills and Maureen Dowds -- and in our industry the Alan Zeichicks, Peter Coffees and Larry O’Briens -- have proven, over years of reporting and analysis, that they are the names you can trust.

Unnamed? If there's a rise in anonymous blogging, it's not in the areas I follow. But hey - no mention of the AP, Reuters, and UPI, who also push out stories with no bylines quite frequently (and whose stories get spread a lot more widely). And Maureen Dowd, or George Will? Those two are opinion columnists, not reporters. You don't read either one to get breaking news, and I'm not sure why either of their opinions on politics would be better than a blogger's.

Further back in his article, he mentioned WikiPedia:

Meanwhile, encyclopedia publishers employ a veritable army of fact-checkers to ensure the information they put out is accurate. Does Wikipedia use the same standard, or does it assume that the public at large is the army of fact-checkers, who jump in and correct errors they find? Well, what happens if the day I look something up is the day before someone with better knowledge corrects the very entry I relied on the day before? Who’s standing at the gate before this information gets disseminated over the Internet?

I addressed that here. Short answer: Wikipedia has churn primarily in controversial (current or near current) events. The fact checkers used by Wikipedia are reviewed in real time regularly - when are the fact checkers for a print encyclopedia reviewed? They make errors (and have bias) at the same rate, at least according to the checks that have been done. Rubinstein seems to believe in some mythic infallibility on the part of media pros. Umm, yeah. I have a few names for him - Steven Glass, Jayson Blair, Mary Mapes. Do they discredit the entire media field? Of course not.

By the same token, a few verbal bomb throwers don't discredit the entire blogosphere either. This article sounds a lot like others I've read on this subject - a deperate attempt to prove that the author is part of some group of near infallible experts, while the hoi polloi out here are just noisemakers who should sit down and shut up.

Tim Bray continues to try and fight his way through the Java breakage (he updated the linked post again) - he might be better off reading this

All static typed languages are broken.
Joshua Bloch, Software Engineer at Google Research recently wrote about the non-correctness of ALL mergesort implementations. His statement stands on the tottery shoulders of static typed languages. Joshua, get a decent programming language and don't tell non-sense!

Bonus quote linked from that page, from one of the authors of the Refactoring Browser:

Static types give me the same feeling of safety as the announcement that my seat cushion can be used as a floatation device. (Don Roberts)

Based on the comment stream from my last post on this, expect more desperate reaching for chewing gum and bailing wire...

Via Chris Pirillo, I ran across the lyrics to "You're Pitiful", a send up of "You're Beautiful". It was hilarious just reading the lyrics in time with the actual song - when I get off this plane, I'll have to grab the free mp3 from the site.

Posted this morning, since I forgot it last night...

Troy spots an amusing way to break Notepad. An explanation of the results can be found here.

With this news:

Microsoft announced Thursday that chairman and co-founder Bill Gates will transition out of a day-to-day role at the company, effective July 2008, to spend more time working on his charitable foundation.

Microsoft's active era comes to a close, and they enter the "big, but not influential" space that IBM entered when Microsoft came on the scene. Gates was able to turn the huge ship that is Microsoft, seemingly through force of will. He clearly has other interests now, and I just don't see Ballmer as the guy to lead Microsoft through the period they are entering. Ray Ozzie has vision, but I don't know that he'll be able to pull the troops along the way Gates - as founder - could.