Smalltalk Tidbits, Industry Rants

The NYC Smalltalk Group is meeting soon:

We are meeting on Wednesday (October 27, 2004). See:

http://wiki.nycsmalltalk.org

Charles

Dear members:

NYC Smalltalk will be meeting on October 27th where we will review VisualWorks Web Tool Kit. Visit our wiki for more detail: http://wiki.nycsmalltalk.org

Our meetings are opened to the general public. Invite a friend.

Early in the development of BottomFeeder, there was some issue with file url handling in the (then current) release of VisualWorks. I created a hack work-around for the problem that worked, and then promptly forgot about it. In the interim, the bug in the library has long since been fixed, and my hack has always had problems with relative file urls. This finally caught up with me today when Bob ran into a problem with some scarping code he uses to create feeds for Bf. He not only found the problem, he went ahead and fixed it - and I've got the new code available as an update. Thanks Bob!

If the Red Sox win, my father will put the "I" in insufferable, that's why :) On the Buckner play in the 1986 Series, I had to prevent him from doing real damage to the TV....

There's been a lot of buzz lately about podcasting. There's already been backlash, even - complaints about how posting of large media files won't scale, etc. That's not why I'm skeptical. I'm skeptical for time management reasons. In general, I'd much rather read someone's thoughts online than listen to them. Why?

• It takes a lot less time, and I can have music on while I read
• Written thoughts tend to be far more well organized than verbal ones

I can skim a 1000 word essay pretty quickly - it takes a lot more time to listen to the same thing. It also takes more effort on my part. Give me the written word every time - it's like the difference between longhand and typing...

I've done something right - my daughter wants to go to this convention over her birthday weekend (she's turning 11 this year). We went to a day of a similar convention last summer, and she played a game of Puerto Rico at a table with five adults. She did ok - came in a close third in a low scoring game. She liked it when the winner asked the 4th and 5th place finishers "How does it feel to lose to a 10 year old?" So, we are going to go to Timonium in November and play in the PR tourney. It should be a good time.

Now Scoble is asking why there aren't search feeds at Google. Dave Winer mentioned this earlier, but in his case, it's because he hates Atom. Mind you, I still think Atom is a waste of time and effort, but that doesn't mean I don't support it in BottomFeeder. I can find search feeds from Google, Amazon, Feedster, Blogdigger, Headline news, NewsTrove, and Yahoo right now. There are likely a bunch of others that I just haven't stumbled on. Some of them are Atom, some of them are RSS. At the end user level, no one cares, since most tools work with either format....

So what kind of moron generates spam for a Wiki page via editing? There's been a rash of it on the VW Wiki at UIUC and on the CST Wiki. I've added IP filtering to the CST Wiki, and Ralph says they have a solution for the VW Wiki almost ready to roll out. The question I have is, how stupid do you have to be to bother trying to do this? It's very easy to roll these changes back. A pain in the butt, yes. Hard no.

The car I normally drive is pretty old - a 1989 Mitsubishi Mirage. It's been a very reliable car - it's got 124,000 miles on it, and it still gets 30+ mpg in town. That's really the only kind of driving I do anymore - I think I drive less than 1000 miles a year now. So I was unhappy when it refused to go into gear last week - I thought "this is it, the transmission dropped". Well, I got lucky. It turns out to be a much simpler problem to fix, less than $200. Big sigh of relief. Of course, that couldn't be the end of the story though. Our other car, the minivan - needs brakes, two new tires, and an alignment. Not terribly expensive, but it does tend to add up. I'm getting the minivan dealt with this morning, which is why I'm sitting in Starbucks instead of sleeping. While they fix my car, at least I can surf...

I need to plug some code that Michael created - check out the DiscussMethods package in the public Store. I used that to mark up the code in this post. It's really easy to use - just select a bunch of code in a browser or workspace, and select the 'copy styled xhtml' menu pick - then post it in your blog (or Wiki page). It's using the same kind of markup scheme as the RBCodeHighlighting package uses in the browser. Very cool.

This is an odd link for me to post, since it's political - take a look at this snippet of MSNBC from yesterday. Never mind the actual points of the argument, or the politics that underlie them, because none of that matters. What's important is the way O'Donnell tries to argue - loudly yelling over his opponent in an angry, spiteful fashion. Now again - never mind the politics - just look at it as marketing. Imagine being at a trade show, and hearing two technologists carrying on like that. Which one are you more likely to pay any attention to afterwards? To paraphrase something Alan has said to me, "the person who remains calm wins the argument". In marketing terms, always remember that perception is reality.

Now let me relate that to marketing in the technology sector. This is a large part of the reason that Sun was never able to out-market Microsoft in the past - McNealy would always come off as angry and unpleasant, while Gates and Ballmer were simply enthusiastic. It seems that Sun has learned from that - Jonathan Schwartz is never angry on his blog, and tries to make his points with humor. Far more effective than yelling or deploying lawyers, I think.

Scoble makes a lot of good points in this article, but the one below really stands out to me is this:

Three, blogs can reduce negatives. Is something bugging your customers? Well, they'll yell about it and yell about it until you listen to them and start having a conversation. Chuq is right on this count. Microsoft has made a corporate decision to change its public face -- I and the other more than 1,000 bloggers at Microsoft are stark evidence of that.

One of the things that Scoble - and the other MS bloggers - provide is a working feedback mechanism. Before Scoble and the rest of them got started, how could you possibly get useful feedback to MS? If you happened to be one of their MVP developers, maybe you could. That's a very narrow slice of their total market though, and the feedback from a group like that isn't going to be all that relevant across the entire company.

Here's an example - I've riffed about things I don't like in Word more than once. Now, I'm just one user, on a corporate Cincom license - if I call support, the kinds of issues I have simply aren't going to get listened to. With blogs though, they at least get noticed. Chris Pratley has responded to Word complaints from a number of people on his blog. Scoble has also been quite good about complaints - follow his blog for a few days and you'll see that he responds to people across the spectrum. What he's doing is nothing less than re-branding MS - slowly but steadily, he's helping make it look less distant and remote. A lot of companies could learn from that.

This is amusing. Via Blogging Roller comes this quote of a Sun response to an HP complaint:

Once again, in certain of the places this is a statement of opinion by Jonathan Schwartz. His opinion is based on his good faith assessment of the current climate of HP. Alternatively, however, Sun will also stand behind this as a statement of fact that is true and accurate based on the above substantiation. As detailed by the above facts, we have seen signs that HP is abandoning HP/UX. Jonathan Schwartz's opinions and even his vigorous debate on this subject as well as Sun's product comparisons and dialog on these commercial matters are inherent in Sun's competition with HP and are part of the free market system in which our companies operate. For our statements of fact, Sun has valid, objective and verifiable evidence. Accordingly, and based on the above, Sun affirmatively stands by its claims regarding HP/UX and will not agree to cease making such truthful and/or subjective claims.

This all comes in response to one of Jonathan Schwartz' bolg posts - this one here in particular. HP's response to this was to have their lawyers demand a retraction. This is a case where Sun and Schwartz have realized that there's a new playing field, and HP hasn't. Schwartz is out there offering up his thoughts - in public. The proper response for HP would be to get one of their people out there blogging as well - instead they've attempted an old school legal response. What's the result of that? Sun and Schwartz look good - and HP looks like a whining set of bullies.

I've had plenty of comments on Schwartz' blog posts, but I give him (and Sun) a lot of credit for playing in the marketplace of ideas. HP doesn't get it.

Freeform Goodness is reporting a new phishing target - Linux sysadmins! Take a look at his post - he's received a clever attack:

Anyway, I thought this was notable -- I've seen phishes like this targeted at Windows users, but this is the first I've seen specifically targeting unix admins. One would assume that they just collected a bunch of webmaster addresses, figuring (probably correctly) that a fair number of those boxes would be running Redhat. The email shows an attention to detail -- the HTML links to Redhat's real logo, linked from a Redhat server, and they even ran their HTML through Tidy!

Words pretty much fail me for this bizarre story.

We had a conversation on edit and continue in the new VS tools that MS is shipping in the IRC channel the other day - Sean Malloy gave his take on things over on his blog. As we discussed the differences in VS and in VisualWorks, we came up with the description Implement and Continue for the way it works in Smalltalk. What does that term mean? It means that you can do a lot more than small scale editing in the debugger in Smalltalk. You can implement new methods (the debugger itself can generate a stub when you hit a Message Not Understood - perfect when doing TDD, and you haven't actually written the code yet.

This all came up in this post from a couple of days ago. Interestingly enough, that item got linked on the MSDN VS 2005 page. If any of the MS oriented developers reading this are interested in having a look at Smalltalk, you can download it here.

I finally got around to watching Farscape: The PeaceKeeper Wars last night. The ReplayTV came to my rescue again, as I'd been watching the Yankees collapse while this was on.

The thing I liked most about this was that they actually ended it - there was no holding out the idea that there might be more - this was well and truly an ending. One of the main characters died (and that character died well) holding off a Scarran assault - I won't say who, because I'm sure there are people who haven't seen this yet. There were a few things that irritated me - the whole baby thing with Rigel was just silly, and - just like in Bond movies - I was always wondering why the bad guys didn't just do the simple thing and shoot the heroes when they had the chance.

Still, it was a worthy ending to the series - I'm glad they were able to wrap it all up.

Michael asked for a small enhancement today - he wanted to have the number of items in a category shown in the category list. That was easy enough, and the server does that now. This was another hotfix executed on the server, and it illustrates one of the strengths of a Smalltalk server, IMHO.

The server has a per-blog cache for each category - until a few minutes ago, that cache was simply a dictionary that looked something like this:

categoryName -> Set(filename1, filename2, ...)

Each entry was a set of filenames - the filenames being files that hold entries in the given category. When a category search comes in, the server merely grabs the set of filenames, gets the appropriate entries from each, and displays them. Well, now we wanted more information - so I created a small Object called CategoryHolder:

Smalltalk.Blog defineClass: #CategoryHolder
	superclass: #{Core.Object}
	indexedType: #none
	private: false
	instanceVariableNames: 'filenames numberOfItems '
	classInstanceVariableNames: ''
	imports: ''
	category: 'Blog'

All I did was move the set of filenames into an object, and included a count of items. That gets incremented each time a new entry falls into a category. Then I made the necessary refactorings in the handful of methods that dealt with this cache. Finally, I had to load the changes into the server. Once those changes were in, I had to script up a conversion method - which consisted of this:

| blogs |
blogs := Blog.BlogSaver allInstances.
blogs do: [:each |
		each categoryFileCache: nil.
		each setupSearchCategory].

Now, it's possible that you got an error browsing the blog if you happened to hit it while I was executing that - but it was all done in less than 30 seconds. The new functionality is there now, and all without cycling the server. That's the power of Smalltalk

The Cincom Smalltalk User's Conference is rapidly approaching - it's taking place this December in Frankfurt, Germany. I don't have a full schedule to link to yet, but I can give you a few details on what to expect:

• Lots of talk from Cincom Smalltalk engineering. A bunch of our technical staff will be there to answer your questions. They will also be giving talks on a variety of subjects:
• GLORP - where this open source O/R framework is headed
• The VisualWorks Web Toolkit and Seaside - both will be discussed
• dotNET connectivity and interoperability
• Opentalk interoperability
• The Product Roadmap for ObjectStudio and VisualWorks

There will be lots of other detailed technical talks as well, and an opportunity to talk to our engineers. Don't miss it!

Bruce Eckel needs a dynamic language. He has a long post going into why this simple exercise doesn't work:

What I was attempting to do in this example is create a sum over a sequence of elements. I'd like to do this in a generic fashion in order to create reusable code. Presumably, the only thing that makes sense to sum is something numerical 13 so the bound or base class should arguably be Number. Generics are called for because I want to work with the broadest number of sequences, so the logical thing to do is only constrain it to be able to produce an Iterable<? extends Number>. It's also important that we get the exact type rather than using wildcards so that the correct operations can be performed.

Many paragraphs later, we discover that it simply can't be done in Java, at least not in a generally reusable way. Here's how you do it in Smalltalk:

| collection |
collection := #(0.75 4 7.27d) asOrderedCollection.
collection add: 3/5.
sum := collection inject: 0 into: [:subTotal :next | subTotal + next].

That's it - and most of that was setting up an example collection with disparate elements. Now compare that to the Java attempts....

Niall Ross has delivered his ESUG 2004 report. It's a large document - 69 pages. My ESUG 2004 posts can be found here

Ed Foster has posted an interesting letter from one of his readers. A fair amount of it reads like the guy's a luddite, but I think there's general cluelessness here as well:

"Essentially -- if anything has to talk to anything else -- I avoid it," the reader wrote. "Because it won't. Or it won't without a lot of coaxing. Or an upgrade. Or a separate service charge. Or the moon being in the right phase, you standing on the left foot, and reciting Shakespeare. In fact, a lot of my 'time saving tools' have been costing so much time spent fixing glitches caused by 'computer errors' on the part of software or the institution that I've dropped:

• "High Speed Internet: The time spent downloading or uploading mid sized files on dial-up is now exceeded by the time spent downloading and installing the latest fixes and patches to keep out viruses I never had to worry about when using dial-up. Cost? Don't go there.
• "PDAs: I loved my PDA -- but problems using it with the Outlooks address book always put me traveling, out in the middle of nowhere, only to find the latest update before I left wrote gibberish all over my travel instruction to where I was going, or over critical contacts. So now my contact file is back in Word, where I have total freedom with the fields and can see what I'm getting. Besides, the PDAs are getting outlawed at many of the facilities I visit, due to camera concerns (mine doesn't have one, but tell that to the sixth grade graduate behind the security desk).
• "High-End Cell phones: No longer do I go for the high end with features, modem capability, and programmability. I get the basic phone. You spend more time trying to learn a new entry system, menu structure, etc every eighteen months or so than the features save you. (Not to mention the cost of the cables, software, etc.) Now I enter the number and punch whatever button I need to call."
• "Online bill payment: The time spent recently correcting an error by the electronic transfer company -- which modified an electronic address on its own initiative, and sent the money elsewhere -- cost me over $300 in fees and interest charges that I never recovered and more time and phone effort than I would have spent if I had written checks for the last two years.

I think this guy complains too much on a bunch of this. Broadband vs.dialup? Please... if you think dialup is competitive, you're either stupid, or haven't actually had broadband. And mind you - dialup is no protection against viruses and worms. Thinking so is a sign of very, very weak thinking. Now, I've never really glommed onto a PDA - but then again, I've never been much for paper organizers either. I think this is just an entire segment that I don't deal with at all. The riff on picture phones is a real problem though - I like my picture phone, but I'm just waiting for the first time I have to leave it behind. His riff on cell phones just sounds clueless. I've upgraded phones a bunch of times, and each time, my vendor has been able to transfer my old phone book over. As to menu systems - please. Each phone I've gotten has been different, and I've spent maybe 10 minutes adapting to each one. The next riff, on online bill payment and fraud - I've had the kinds of problems he's mentioned with paper and checks. Guess what - this is a function of your vendor's back office technology, not of the front end browser interface. If that sucks, you're going to get stuffed whether you point and click or use a pen and checkbook

This guy is clearly an old dog, and the new tricks are apparently way too much for him. And by old dog, I don't mean age - my father in law, who is over 80, is way more clued in on this stuff than this guy. Someone issue this guy a baseball cap and send the golf cart to his place - it's time for the retirement home...

I've been following the Yankees for decades now - I started watching their games on WPIX in New York when I was a kid in the late 60's. Back then they stank - a lot. They improved in the late 70's, and won a pair of World Series - but Steinbrenner drew all the wrong conclusions from those wins. Rather than recognizing the great pitching, he remembered Jackson's home runs. So all through the 80's, he kept looking for just one more lefthander to clinch the deal.

Not coincidentally, the Yankees went nowhere all those years, until 1996. The last 8 years they've been really, really good - and they also had really, really good pitching. Does anyone think that Vazquez and Brown are adequate replacements for Pettite and Clemens? The lineup is still fearsome - but the pitching is anything but.

The only questionable thing for the Sox all night was bringing in Pedro Martinez in the 7th - on short rest, in relief - and at a point whe Lowe had completely handcuffed the Yankees. He gave up 2 runs, but Bellhorn got the Sox back one against Gordon (who's been a punching bag in the ALCS) the next inning. On the Yankees side, I cringed when Gordon came out. The Sox had just man-handled him the entire series; and like clockwork, he gave it up again to make it 10-3.

The Red Sox managed the unthinkable - they came back from a 3-0 deficit and earned a trip to the World Series. Heck, after vanquishing the Yankees they may even win it. Meanwhile, here's hoping the Steinbrenner figures out what the problem is - and here's a tip - it's not hitting...

I guess in Dave Winer's universe, only RSS syndication sources exist:

New feature: If you do a search on the Baltimore Sun website, you'll see a beautiful white-on-orange XML icon providing the search results in RSS. Subscribe to the feed to peform the search every time your aggregator updates. Don't you wish Google did this?

I guess that the Google feed building support in BottomFeeder just doesn't exist then. Huh. That's too bad, since I subscribe to feeds that way.....

Over on the VAST Newsgroup, there's an interesting thread about the future of VAST. This is all the more interesting due to the "transition" talk that IBM hosted at last year's Smalltalk Solutions. If you've been hearing about transitioning from VAST - and if the transition offer is to Java or .NET - you should realize that there's another choice. Any migration to Cincom Smalltalk is going to be a ton easier than moving to Java or C#. We have service partners with experience in this area as well, so if you want to stay in the Smalltalk world, take a serious look at Cincom. Take a look at our product roadmap - we are evolving and advancing Smalltalk rapidly. If you want a committed Smalltalk partner, you know where to find one :)

Panopticon posted on dynamic/static issues awhile back, and I've been pondering the post. This morning, I have a few thoughts on it. The gist of his post is that he likes the immediacy of dynamic languages (which he refers to as scripting languages) - but then caveats it this way:

In the end, I think anything that helps the average programmer be more productive is a good thing. By and large, static typing satisfies this dictum: static typing enables all kinds of programmer productivity features like Intellisense, better error messages at compile time, etc. (One could argue, I suppose, that you could lose the static typing and use type inferencing instead, but I wonder whether it would be possible to build a complete enough type inferencing ruleset that: a) was implementable, b) made some kind of sense, and c) could compete with just stating the damn type of your variables.) Dynamic environments also do this: edit and continue (pace Franz et al.), continuable exceptions, being able to call functions at design time, etc. So I think marrying the two worlds has some facinating possibilities.

Well, most of those things are possible in languages like Smalltalk. Intellisense - we have code completion tools as loadable components in VisualWorks. As it happens, this isn't something most Smalltalkers seem to want... but it's there. Compile time errors - that gets into an interesting definition issue.

In languages like C#, Java (etc), compile time is a "heavier" thing than it is in Smalltalk. Sure, there's incremental compilation support in some of the IDE's out there - but in Smalltalk, there's really no difference between compile time and runtime - we compile methods one at a time as we go, and we can immediately test. Some people use agile methodologies - they write the tests fiest, and then run them. Others do workspace twiddling and then migrate to a browser (this is more in line with how I develop - it's how I learned Smalltalk). Either way, there's no gap between runtime and compile time - there's really no interval that a Smalltalker refers to as compile time.

To some extent, this gets back to what I posted yesterday - in this debate, as with debuggers, Smalltalkers and developers using the mainstream languages tend to talk right past each other - we use the same words, but we don't really mean the same thing at all. From our perspective, you really can have your cake - and eat it as well.

There's been an annoying issue with manual text size changes - up until this morning, if you had adjusted (using the toolbar) the size of the text up or down, the html would render once in the default size, and then again in the size you had selected. I figured out how to address that issue this morning, and posted a fix. Grab all updates and restart - you should see smoother rendering after that.

I have to give the Sox credit - over the last three games, they've had plenty of opportunities to lie down - and they didn't do it. At the same time, the Yankees lead batters have gone to sleep since the third game.

Tonight's game had some oddities - two calls by the umpires were reversed after huddles (both correctly, as it happens, and both against the Yankees). That's a fairly new thing in baseball - I remember watching games as I grew up, and calls were never reversed - once an ump made a call, it stuck - good, bad, or indifferent. The main problem with the huddle approach is that it slows down a game that desperately needs speeding up.

Either way - the Sox have done what no other MLB team has ever done - forced a seventh game. Either the Yankee bats wake up, or the unthinkable will occur.

So the Google Desktop is getting a lot of hype - lots of people are saying "why do we need WinFS if we have this?" - an example is the BitWorking post I linked to the other day. Others are trying to point out that WinFS is supposed to be more than desktop search, like Julia Lerman and Dare Obasanjo. What they are trying to do is an attempt at swimming upstream against the perception tide:

The fundamental premise of the above statements is that the purpose of WinFS is to make local desktop search better or to use a cruder term to create "Google for the Desktop". It may be true that when it first started getting pitched one of the scenarios people described was making search better. However as WinFS progressed the primary scenarios its designers focused on enabling didn't have much to do with search.

That may be the case - I haven't followed WinFS talk well enough to have any real idea. However, the conventional wisdom on WinFS is that it's all about search - and that's why the "Google Desktop obviates WinFS" meme has been spreading. Microsoft is going to need to start talking about the other benefits of WinFS in a compelling way, or this notion of WinFS is all that will exist...

I have to admit - even though I hated the final outcome - the last two games of the ALCS have been great baseball. At this point, the main question is this - just how much more sleep am I going to lose before it ends? And heck, we haven't gotten to the World Series yet!

Here's an example of something that tends to lead to Smalltalkers and non-Smalltalkers talking right past each other. Take a look at Sam Gentile's post on unit testing and debugging:

As a C# MVP, I am obligated to tell you that Edit & Continue is in C# now in this build regardless of my feelings, which are the same as John Robbins here. Program logic should not be tested in the debugger. You are wasting your time and your company's time if you do so. The debugger is for intractable problems. Unit tests with the place to verify, refector and edit as you go; not the debugger. Thats why NUnit-AddIn and VS Team System are so important - that's the way to do it.

Now, this is where you'll often see people like me queuing up the Wow, he doesn't get it music. I've come to the conclusion that it's a development culture difference. In tools like VS, the debugger is a completely separate tool - it's not a code browser that can debug, it's a debugger with some very, very limited code changing capabilities (see this post for an example of that).

In Smalltalk, the debugger is just another browser that knows how to step through code. If we run across an unimplemented method - we can generate it in the debugger, and then step into it. We can then write the code with all the state being available. In particular, we can do this while running the test we wrote - the one we wrote first, before we wrote the code.

Ultimately, the difference is at the environment level - the strengths and weaknesses of the tools affects our view of how to appropriately use them. For people using languages like C#, Java (etc.), most agile developers have come to the conclusion that the debugger is a tool that you use only in extremis. For Smalltalk developers, on the other hand, the debugger is your friend - it's part of your normal development toolset, to be used on a regular basis.

Scoble has a great post up on the value of blogging to a company. To summarize, it puts a face on an otherwise faceless entity. Take Scoble for instance - there's the caricature of Microsoft you see in trade rags, and then there's the voice that Scoble himself provides. heck, he doesn't always agree with his management, which gives him more credibility.

He gives a number of good examples - both of the failures of silence:

Kryptonite. Lately I've been asking audiences I've been speaking to "who knows the Kryptonite story?" 75% do . If you don't know the story, do a Google search for Kryptonite and "Bic Pen". We'll wait. We just watched the destruction of an American brand. 75% know about it. Why? Because of one or two weblogs and the new word-of-mouth network. Yes, Engadget and Gizmodo do have that kind of power. Engadget alone has 250,000 of the most influential readers the world has ever seen. My second question is: "What have you heard from Kryptonite about this issue?"

And also of good examples - Mark Cuban, Channel 9 at MS, a nifty looking Ford blog. Scoble has all the links and explanations on this - go ahead and read his article.

Want to hear about the future of Cincom Smalltalk? Then come to the upcoming Cincom Smalltalk User Conference this December. Here are the details:

When does the conference begin and end?
Tuesday, December 7, 2004, 10:00 AM - 7:30 PM - followed by a buffet dinner.
Wednesday, December 8, 2004, 08:30 AM - 7:30 PM - followed by a buffet dinner.
Thursday, December 9, 2004, 08:30 AM - ~1:30 PM

Where does the conference take place?
Relexa Hotel Frankfurt/Main
Merton-Viertel/Lurgiallee 2
D-60439 Frankfurt/Main
Germany
Phone: +49(0)69/ 957 78-0
E-mail
www.relexa-hotels.de

We'll have a lot of our engineers there ready to answer your questions - see you there!

Well, in the "should have made you think" category, we have this post from Jonathan Schwartz. You might have thought that the Kodak suit and settlement would have led to a new set of thinking on IP and software patents; you would be wrong:

First things first: nothing's changed. In terms of patents, and intellectual property, nothing whatsoever about the Kodak settlement suggests Sun's position on IP has changed. Moreover, the settlement was all about ensuring what Kodak was doing to Sun, they could do to no one else in the Java community.

Translation: "Your house just burned down! Are you going to rethink your objections to smoke alarms? Nah, they could still wake me up from a sound sleep". Mind you, it's not just Schwartz and Sun that need to wake up from this patent slumber; it's the entire industry. Sun can't move until or unless other companies do. This particular suit could have been a useful teaching moment though...

And obtw - there's no capital T in Smalltalk :)

This was just too much irony to see all at once. Sixth inning, Pedro went over 100 pitches. 3 runs were in already on a double by Jeter with the bases loaded and 2 out. He stayed in. He loaded the bases again, but managed to get out of that jam. That's not the ironic part though. As I was watching that, ESPN Classic was showing last year's seventh game. As Jeter came to the plate, the Classic game was showing the eighth inning - right as Little came out and left Martinez in. The rest, as they say, is history - 5 outs from the World Series, a tired Martinez lets the Yankees tie the game.

Flash back to this year - Mussina pitching well after the first inning, and Boston into their completely depleted bullpen. All the Yankees had to do is get to the eighth - and hope that Rivera didn't do what he did last night.

The Yankees threatened in the 8th. Cairo doubled, Jeter moved Cairo to third, and Sheffield walked. Timlin got A-Rod on strikes, but that brought Matsui - who's been Godzilla to the Red Sox' Tokyo this series - to the plate. That was it for Timlin - Francona brought in Foulke. He did the job, getting Matsui to fly out to left. Gordon stayed in for the eighth, which may have been a mistake - Ortiz led off with a shot over the monster to make it a 4-3 game. Typically, Gordon didn't hold up to the pressure, and left with runners on 1st and 3rd - with no outs.

Varitek tied the game with a shallow fly to center - if the 6th looked like Deja-Vu all over again from last year's game 7, then the eighth looked like a replay of last night's late game. Rivera retired three in a row, with the damage being the man Gordon left on third. Fans in New York and Boston had to be glad that this game began at 5 - because another extra innings game could last awhile. Foulke continued to his job in the top of the 9th - 3 pop ups. There was some excitement with a walk to Sierra and ground rule double from Clark - but nothing came in. The Sox really got lucky on that double - had it stayed in the park, Sierra would have scored easily.

Damon's agony continued in the 9th. He finally got a hit - and then got thrown out stealing on a great throw by Posada. The 9th ended quietly, and sent the game into extra innings. Through 13, the closest anyone got to scoring was the Yankees in the top of the 13th - on the power of 3 passed balls. Wakefield pitched out of that mess though, and the game just kept going.

That set up Ortiz to be the hero again in the 14th - he hit a looper into center, sending Damon home for the winning run. The series goes back to New York. Once thing is for sure - it's not going to be an easy win for whoever gets to the World Series

Sam Gentile is missing something fundamental about Google Desktop. The fact that it's the same interface as the internet search People like metaphors that they are already familiar with