Smalltalk Tidbits, Industry Rants

Sean Malloy explains just how lame edit and continue is in the VS world.

If I hadn't thought that the Atom mailing list had gone mostly around the bend already (how many date format emails can you possibly pay attention to?) - then the reaction to this piece on advertising in a feed would have convinced me. There was a suggestion to add an extension to Atom in support of ads. Yeah, that'll work. All clients that would so much as bother to enable that extension raise your hands now.

This is astoundingly simple, and it's already happening. Syndication content is text or HTML. Which means.... you can put ad content in as an item. Apparently that's just way, way too simple for some of the deep thinkers over on the mailing list - if there's not an RFC deploying toxic word fog on the issue, it's not solved yet. Pardon me while I snore...

I have to say that I agree with BitWorking on Google Desktop:

This is just like WinFS.

Except that it is shipping today.
And it just works.
And it doesn't require an upgrade to your operating system.
And it doesn't require you to manually tag all your files with meta-data.
And it works outside your "Documents and Settings" folder.
And it's only a 400K download.
But besides that, it's just like WinFS.

With the latest BottomFeeder updates you should see some nice improvements. Image download happens in the background, and there's reporting on the progress of that. The update that went live this morning added reporting of that progress to the status bar area - you'll need to restart after grabbing it. As always, let me know if there are issues.

Well, the 6th inning is the point that people would have pointed to had the Sox lost tonight - it's where the Sox made the kind of muffs that make Boston fans wince. They were up 3-2, and then Matsui tripled. He came in on a funky chopper that Mueller couldn't handle at third. Then a wild pitch sent Williams over to second. Posada got on. Another wild pitch, but they got lucky throwing Williams out. Then a bobbled ball by Bellhorn at second let Sierra get on, and they walked Cairo to load the bases for Jeter. They got Jeter on a routine ground ball, but by this time the score was 4-3 Yankees. While they only scored two runs, it had all the features of recent post season Yankees/Sox matchups - a questionable pitching change (Lowe out, Timlin in), followed by a bunch of muffed plays in the field. You could almost hear the oxygen being sucked out of Fenway at that point.

It stayed that way though 7, and then the Yankees brought in Rivera - who closed out the 8th easily. Meanwhile, The Red Sox brought in Foulke in the 7th - and this was his first 3 inning appearance in a long, long time. He did the job though; held the lead to one run.

So it came down to the bottom of the ninth, with Rivera on the mound, protecting a one run lead. He couldn't do it; Millar got on with a walk, a pinch runner came in to steal second, and Mueller drove him in. A sacrifice bunt sent Mueller to second, and that brought up Damon - who only had one hit in the entire series. And now it was the Yankees turn to muff plays - a hopper to first was bobbled, leaving mean on 1st and third with 1 out. Rivera struck Cabrera out, bringing up Ramirez - and no rbi's to this point in the series. Walked on a hight 3-2 pitch, leaving the bases loaded for Ortiz. Things just couldn't get more tense than this. Ortiz popped it up, sending the game into extra innings. At that point, the real question was: With the Sox pen so depleted after last night's game, would they be able to hold the Yankee bats down?

The Red Sox sent Embree out - he only lasted 2/3rd of an inning last night. Did his job in the 10th though - got the side out quietly. Gordon came in and did his job in the 10th - it now came down to a battle of bullpen staying power.

To the 11th - Cairo ripped a single, and Jetter bunted him over to second. At this point, Embree was throwing nothing but fastballs. He got lucky on a shot by Rodriguez that Cabrera saved at short. Embree pitched around Sheffield, and Francona brought in Myers to face Matsui - the guy who's done more damage than anyone else during this series. Myers walked Matsui on 4 pitches to load the bases, and off he went - in came Leskanic. He got slapped around last night, and tonight he faced Bernie Williams with the bases loaded and 2 out - he did the job tonight though, getting Williams on a shallow fly to center. One thing is for sure - if the Sox didn't score soon, they were going to run out of pitching choices.

The 12th started with Leskanic still on the mound, and Wakefield warming up. Posada got himself on with a broken bat looper to right. Sierra accidentally sacrificed Posada over - he lined a ball into Leskanic's leg. He stayed in while Wakefield warmed up. Meanwhile, the Yankees got Quantrill ready to face the heart of the Sox's order. Leskanic finished Cairo off, and the Sox came up to try their luck against Quantrill. And that's how things got to game 5 - Ramirez singled, and Ortiz' bat finally woke up with a 2 run shot. The Red Sox get a game 5, in less than 16 hours.

This afternoon my daughter's girl scout troop had an outing to Goddard Space Center. The girls had put together some rocket kits recently - we had a kit from Estes rockets. It turns out that a local rocket club meets at Goddard once a month on Sundays, and helps anyone else who shows up launch a rocket. It's all pretty safe - the rockets are lit of electrically from a launch pad, so there's no fire, or fuses. Here's a shot I took just after my daughter's first rocket went off - I missed the launch itself (kind of hard to catch with a cheap camera phone :) )

That one went pretty well - it had a streamer instead a parachute, but it came down nicely and in one piece - we can slap another engine into it and try again sometime. Here's a shot of the second rocket, which looked like a cruise missile:

Again, I missed the launch itself. This one went up, but the parachute didn't deploy properly - and it came down hard. The nose buried itself in the ground, and the rest of the rocket came apart (it can be re-glued - it didn't actually shatter). Here's what that rocket looks like ( I can't find the first one on the estes site):

All in all, it was great fun - we'll definitely do it again.

After getting a few bug reports, I spoke with Michael. After he pondered the background download code of WithStyle for a bit, he had a patch - not only did it go into his nightly build, but he was kind enough to send it to me immediately. So, there's an update available for the existing build. I decided that this warranted a fresh set of builds, so the downloads on the site have all been replaced - if you want to start over with 3.7, just grab the appropriate base-app*.zip file, and replace the exe or VM.

Gads, as I start writing this, it's only the 4th inning - and the score of the Yankees/Sox game is already 11-6 (Yankees). The pitchers have all been rocked - starters and relievers alike. The Yankees are on their second pitcher - after Brown got rocked, they brought in Vasquez, and he got touched for two - and his control is looking none too good. Meanwhile, Arroyo got shellacked. His replacement, Mendoza, lasted part of an inning before they went to Leskanic - and he just gave up 5 runs. They brought in tomorrow's starter, Wakefield to close the inning down. I guess they'll be handing the ball to Lowe tomorrow night. At this rate, the core could be in double digits for both teams, and it could read like the score of tomorrow's Patriots game.

So here it is the bottom of the fourth - I guess with a 5 run lead, the Yankees feel ok leaving Vasquez in (and he just struck out Bellhorn, so maybe it's ok). This is going to be a game that depletes the pitching staffs if this keeps up. Sheesh, I spoke too soon - a walk to Ramirez after a boatload of pitches, followed by a single to right by Ortiz. Sturtze is getting loose (and let me tell you, that doesn't fill me with confidence) - and there's only one out. Vasquez is not finding the strike zone - and wow, did he get lucky. A line drive right at Olerud, and he was able to tag Ortiz before he got back to the bag. If Vasquez is back out there next inning, I'll be shocked.

I guess this counts as live-blogging - I'll just keep updating the post as stuff of interest happens.

Well, that didn't take long. The top of the fifth was looking quiet with 1 on (Jeter) and one out, and then bam - A-Rod doubled. Now it's 12-6, and there's A-Rod in scoring position. I wonder what the Sox will do if Wakefield starts getting pasted? That bullpen is starting to look very, very thin. And abm - Sheffield pastes another ball for a double, bringing A-Rod home - 13-6. The Sox bench is looking like a wake right now, and I don't think I've ever seen Fenway so quiet - you could just about hear a pin drop... As things have gone tonight, that was a fairly light inning - only 2 more runs scored. It's going to be a long, long night for the Red Sox bullpen at this rate...

Well, Vasquez is back out, and he's being effective. He actually threw a 1-2-3 inning! On to the 6th, and we'll see if the Yankees can get Wakefield torched out.

Maybe things are settling down for awhile. Posada went deep, but to dead center for an out. Huh. Sierra pasted one as well, but to left-center - just about the only place that ball could be caught. Two outs, but Francona can't feel good about those balls - Wakefield dodged two bullets just now. Well, lucky or not, Olerud went quietly - 1-2-3 for Wakefield. The Yankees have a little over 2 innings before they can go to Rivera - I expect to see him out there in the 8th. We'll have to see how the next few innings fall out...

Hmm - Olerud must have hurt his foot on that ground out. He's out, and now Cabrera has just slapped a double to that odd corner in center left. Damon just continued his ineffectuallness in this series - shallow pop out. Fox knows that the game is close to being a laugher - they are out in the stands interviewing Steven King (so this must be a horror movie :) ). Bellhorn just struck out, so we have Ramirez up with 2 gone. And Ramirez goes down on strikes, so it looks like Vasquez has found his pitches, and he's making them. On to the 7th - things are just getting quieter and quieter at Fenway.

And Cairo gets on with an infield hit - The Yankees have their first man on again. Jeter hasn't done much tonight (one walk) - and Damon makes a nice grab on a looper to center. Almost another Red Sox moment as Bellhorn drops a pop in shallow center, but they got Cairo at second - he was assuming an easy out there. Still, 2 outs. Now Fox is looking at stats on teams that have come back from 3-0 in a 7 game series - none in the MLB, none in the NBA - and only 2 in the NHL. In other words, if the Sox lose tonight, the fat lady is warming up. And in anticipation of that, Sheffield just pasted another ball into the green monster for a long single. Wakefield is out, Embree is coming in. Another scoring possibility for the Yankees.

And Matsui just got a single, scoring A-Rod and sending Sheffield to 3rd! 14-6, and no pitcher for the Sox has been up to the task tonight. This game is just washing away from them. And wow - Bernie Williams just put it past Damon, and 2 more runs score, and Williams get sto second. Fenway is now like a crypt - it's filled with the walking dead Sox fans. And Posada finds the wall, scoring Williams, Posada to second for another double. It's now 17-6, and Fox needs to go find Steven King again. Sierra grounds out, and the inning is mercifully over for the Sox. I don't think the Yankees are going to need Rivera at this point - not unless a comeback of epic proportions takes place. At this point, I'm not seeing it. For tonight at least, the Sox are done.

Well, there's some life left in the Sox. Ortiz doubled and Varitek homered - 17-8 now. More bad news for the Sox though - it looks like Ortiz may have hurt himself on his swing somehow. The Yankees have Quantrill warming up, and it looks like Vasquez is just about done. After a shaky start he did enough; he chewed up the middle innings. Nixon got a lot of that ball, but not enough - just a long out. Looks like Quantrill is coming in.

Another hard hit ball by Mueller, but Matsui managed to hold it to a long single. Two outs though, and Quantrill looks like he's got decent stuff tonight. Well, I spoke too soon. Cabrera just lined a single, so Damon comes up with 2 on. Damon flied out harmlessly to left, continuing his inability to hit in this series. No further damage - and Sturtze and Gordon are both warming up. With the 8th inning coing up, I don't expect to see Quantrill go back out. Of course, the open question is - will the night get any longer for Red Sox pitching in the top of the 8th?

Wow, another pitcher for the Red Sox. Mike Myers, and they've also put Mirabelli behind the plate. He's usually the guy who catches Wakefield, and Varitek has been hitting well - I'm not sure I understand that move. Myers got Clark (the replacement 1st baseman) on strikes - his delivery muct be hard to track. He's got a normal overhand delivery, but also a side-arm sinker ball thing. Another strike out - the Sox have finally found a pitcher who isn't off his game tonight. Well, Jeter fought out his at-bat for a single. A close play at second on a good stop by Bellhorn - ends the inning, and Sox finally get another quiet half inning. Two left to play - it's looking like 3-0 advantage Yankees.

Well, Quantrill is still in. I suppose I shouldn't be surprised - why waste another pitcher when you're up by 9 with 2 innings left? This is why Torre gets paid the big bucks and I just comment on thus stuff :) Bellhorn struck out, and Ramirez quietly flew out to center, so it looks like Quantrill has settled in. On Fox, the announcers are talking about the likely starter for game 4 (probably Lowe) - and who's left in the Sox pen for tomorrow. They have used a lot of pitchers tonight. Quantrill gets Ortiz on strikes, so keeping him in looks like the right call - on to the 9th.

Heck, I hit the bathroom and I missed something - looks like Sheffield got on base. Whoa - Matsui just hit a 2 run homer. 19-8. Williams just got a single, so it looks like Myers only had one good inning in him. Wow - the Sox are warming up Foulke - their closer. You never want to bring in your closer when you're down by 11 runs. But heck, what can they do? They've blown through the pen. Posada just got a single, and now there are 2 on, no outs, and Sierra coming up. He struck out, and Clark lined into a double play. The Red Sox survived that inning - but it's just too late.

Well, Gordon is coming in after all. One down, Mirabelli popped out to second. I'm always nervous when Gordon pitches, but I don't think he can blow an 11 run lead. Wow - according to Buck on Fox, we just passed (combined for both teams) 400 pitches. Wow. Now, there's why Gordon makes me nervous - fastball out over the plate, and Trot Nixon slammed a double off the wall in left. Down 11 runs, it's ok. In a typical set up man's spot, it's nerve wracking. Speaking of which, a wild pitch sent Nixon over to third. Millar just struck out, so the Sox have one out left. A soft fly to center, and Gordon did the job. Game over - and the Yankees have their foot on the throats of the Sox.

The ironic thing is, the Red Sox were favored coming into this series. Both teams can hit, but everyone thought that the Sox had the better pitching staff. Of course, the Yankees did win the division, so the Sox pitching couldn't be that much better. Still, in a short series, the thinking was that Schilling and Martinez would mow the Yankees down in the first two games, and then they could just split the next four - just like the Diamondbacks did in 2001. It didn't go that way - Schilling is hurt, and he got shelled. Mussina pitched well for almost 7, and Lieber did very well for 7. Martinez pitched well - but not well enough. Brown got knocked out early, but Vasquez and Quantrill chewed up the middle innings well, and the Red Sox staff served up a steady diet of gopher balls. So at the end of the day, the Sox are faced with having to execute a comeback of the sort that has never been managed by any MLB team in a 7 game series - return from a 3-0 deficit. At this point, I'm thinking sweep.

I've had a few reports of people having problems viewing content in the just released 3.7 - as it happens, the WithStyle guys found a flaw in the html component - and they have a fix. I should have an update out within 2 days (I have to await their build). Once it's out, I'll redo all the builds and upload fresh, so that first time users won't have to grab the update. I'll post on this again when the fix is available.

Update: The patch is available. Grab Both WS updates and restart - things should improve after that

Update: Another tip - if you still see problems, check to see whether your system has hyper-threading support. If so, right click on the BottomFeeder process and set affinity to one CPU - that should address the problem.

Blaine Buxton points to an interesting sounding trading system written in Smalltalk (Dolphin). Of course, no discussion of trading systems is complete without a reference to the Kapital System written in Cincom Smalltalk by JP Morgan.

Congratulations are in order for Eric - he's about to start working for Gemstone. This is great news for Eric, and great news for Gemstone. It will be interesting to see Eric making the transition from the end user side of the fence over to the vendor side. It's great to see things turned around so quickly for him!

So I'm looking at the various sports page reporting of the Yankees/Sox series - the Boston Globe has this piece up. They are still trying to figure out a way to get Schilling out there. This is just the height of irresponsibility - they way the injury has been described, it seems to me that they are toying with Schilling's career in the vain hope of a win. And it's a pretty slim hope, too - look at how he fared with that ankle in game 1. It's time for the Sox to sit the guy down and let the surgeons work. They shouldn't be risking any more damage to him. I realize how much they want to win - but this is just irresponsible...

ESPN documents the latest Sox troubles:

Perhaps it's time for the Red Sox to sue George Steinbrenner over missing child support payments for Pedro Martinez because at the moment, the court system appears to be Boston's only viable way of beating the Yankees. Here's the situation after Wednesday's 3-1 loss, perhaps the most dire Boston has ever faced that didn't involve Ben Affleck

We'll see how tonight goes (if it doesn't rain). The Yankees are in a commanding position; a win tonight will pretty much put it away.

Oh look - VS will have edit and continue next year. Good job making it into the 70's guys :)

I should never express the fact that I like a show; every time I do, this happens. Looks like my friend Mike was right; it's toast...

I've just released BottomFeeder 3.7. This release includes WithStyle instead of Twoflower; the html display is much, much better. Bf now supports CSS, and a number of example stylesheets have been included. Documentation now ships with the download, and is thus accessible offline - something a number of people asked for. If you have an existing installation, you should be able to just grab the appropriate baseapp-*.zip file, and replace the image (or exe), VM. You should delete all the files in the 'app' directory as well. I'd like to thank Michael Lucas-Smith - without his help, I'd have never been able to get NetResources and WithStyle working for me. I'd also like to thank Troy Brumley, Travis Griggs, Mark Derricutt, and Eric Winger for beta testing the in development versions of this release; some of the early releases were pretty raw, and they were all a very helpful. Rich Demers did his usual great job with the docs; any bugs or defects you find are all mine. Enjoy.

The Cincom Smalltalk community comes through again - there's an implementation of Store for Sybase out there now. I ran across this on comp.lang.smalltalk

StoreSybase adds Sybase support to Store. It is tested with VWNC 7.2.1, Linux (FreeBSD 4.8) and Sybase 11.9.2. It uses a CTLib connection to Sybase. It may be downloaded from:

ftp://ftp.cincomsmalltalk.com/pub/goodies/StoreSybase/StoreSybase.tgz

Be sure to read README.syb before installing.

ERRATA:

1. If you want user/group management, install it when asked during database table installation. Installing it later doesn't work.
2. Despite what the User/Group Privileges Tool may lead you to believe, checks for read permission are not implemented. Everyone has read permission on everything.

Enjoy!
Carl Mascott
cmascott_del_t...@att.net
If replying by e-mail please correct my address.

I'm off to Raleigh-Durham this morning, giving the "what's new with Cincom Smalltalk" speech to a customer we haven't visited in awhile. I think they'll be pleasantly surprised at all the work we've been doing - they've been in "heads down" development for quite some time. We'll see how it goes.

Mark Baker points to more evidence of creeping complexity (shades of the OMG spec-a-day approach) from the WS_* crowd. Web Services started out looking lightweight. Lately, it looks like a ton of bricks sitting on top of a sea of lard...

Jonathan Schwartz points to some of the recent successes of OpenOffice and StarOffice - having the EU formally recommend it is good for Open Office, and good for competition - heck, maybe MS will wake up and fix some of the lingering annoyances in Word. The hard part is still ahead though - there are lots of shops that have built a large amount of MS Office specific support into their systems - in particular, replacing things like Excel will be very difficult due to the macro intensive systems that are out there. For people getting started, the choice looks pretty clear - Office is expensive - and to my mind, irritating. Open Office isn't much less irritating, but it's free. Things are dicier in the large user community though...

Lieber, against all expectations, out-pitched Martinez. He went nearly 8 innings and allowed one run (and that happened with Gordon in). Meanwhile, Martinez lasted 6 and gave up 3 - which was a good outing, but not good enough this time. The series now goes to Boston with the Yankees in a great position - and with Martinez hearing "Who's your daddy?" echoing in his ears :)

Via Rob Fahrni I ran across this at mini-microsoft:

"What rarely rarely works is #3: people and product. That company succeeded because of their unique development environment and personalities. Now that environment is gone and they are forced to develop the Microsoft Way. They've been blue bagded. And slowly the greatness that was their product has the life drained out of it and it becomes less and less relevant (e.g., Great Plains). These acquisitions are where we should have been spending money to have great partnerships and not blow a bunch of cash and fritter away a product just to feel we have a market segment covered." -

I don't think this is an MS specific problem. Look at the rapid invisibility of Rational's product line, for instance. When larger entities buy smaller ones, they tend to smother them - the corporate culture of the buyer quite often doesn't 'fit' very well on top of the people who ran the product at the smaller outfit. I recall reading an article in Business Week (a long time ago now) about how many mergers just don't work out - and it's for reasons like this.

Scoble wants to know why people think Microsoft products suck (see the linked article - it is his question :) ). Well. I'm sure everyone remembers just how fond I am of MS Word. Just follow the link - I don't feel like going off on another rant about the wonders of stupidly smart software. I will say this as a general comment - ship these things with the (supposedly helpful) features off. I can enable things like dancing menus by myself if I'm feeling like being disabled, thanks.

The primary issue has to be security though. To figure out just how bad this is, all you need is a PC running any Microsoft client OS and a child who uses that machine regularly. You end up having to clean off various pieces of spyware on a regular basis - whether you have a firewall installed or not. And no, SP2 for XP doesn't help - the so called security we all held our breath for is fundamentally broken. I'd actually disagree with the author of that piece in one important respect - it's actually worse than nothing. With nothing, you might realize that you need a firewall. With SP2's badly implemented firewall, you probably won't.

The basic problem for MS is a nasty legacy issue. Everything they've done over the last 15 years placed individual and group productivity ahead of security - which is why every OS up to XP shipped with a variety of services actively listening. It was obvious around 1998 or so that this was a really bad idea - and yet MS kept plowing straight ahead, shipping completely vulnerable products for years. I've seen the results - you've seen the results. Friends and associates who beg you to help them with their PC problems, attacks that come swarming at you the instant you go online - the poor decisions made from the mid 90's forward will haunt us for years given the slow update cycle that people live on.

Bottom line - we are in the market for a new PC. The top of my list is going to be an Apple product, because I'm tired of living in the insecurity swamp that MS has created. SP2 was a half-assed response to the problem. If they want to be taken seriously, they are going to have to start being serious - which means shutting up about the wonders of Tablet PC's and paying attention to fundamentals. Nifty writing recognition doesn't impress me when so many PC's are 0wned...

Now that's what postseason baseball is supposed to look like. It was like watching two completely different games. The Yankees jumped all over Schilling in the first 3 innings, scoring 6 runs. I was wondering what the announcers were smoking at that point - they kept wondering what was wrong with Schilling - even though we knew about his ankle injury, and even though he repeatedly bent down to retie the shoelace on that foot. He was clearly in pain - you could tell by the way he drove off his back leg, you could tell by the way he kept checking the foot - and you could tell by the lack of velocity and movement on the ball.

The other thing that kept the game interesting at that point was Mussina - for 6 1/3 innings, he had a perfect game going. Then suddenly, Bellhorn got a double. Things went downhill quickly for Mussina - 2 runs scored, and with two men on, out he went, replaced by Sturtze. This is the point where I started to get nervous - and Varitek's 3 run shot didn't help calm my nerves. At least the inning finally ended, with the score 8-5. That was the start of the "second game" - we had 6 innings of laugher with a near perfect game, followed by 3 innings of nail biting.

In the 7th, the Yankees brought in Gordon, the set up guy for Rivera. Now, I'm sure that Torre didn't want to use Rivera - he had just returned from a family tragedy - he only got to the stadium in the 3rd inning. Gordon got an out quickly, but then the trouble started. Two guys got on - this is the point where I was wondering why Gordon was still in - and Ortiz was up. Ortiz has power, but he's not the fastest guy on the planet. He hits a triple, making the game 8-7. This is where I had to run out for Tums.... it just looked bad.

Thankfully, Rivera came in and shut things down. Bottom of the 8th - another Red Sox moment. They have Timlin pitching, the set up guy. One out, and the Yankees get two straight hits - A-Rod and Sheffield get on. Matsui pops out (he had a great night though - 5 RBI's). Bernie Williams comes up. Here's the Red Sox moment - two on, Foulke warming up - and Timlin still in. Williams comes through with a double, and the score goes to 10-7 - which is great, because I can finally breathe again. One more chance for the Sox - against Rivera

The Sox insist on making it interesting. With one out, Varitek and Cabrera get on. Bill Mueller comes up - and this guy is no slouch. The tying run is up, and there's only one out. He hits a shot - foul, thank goodness. Then salvation - a roller back to Rivera, and we have a double play. The game ends, with NY on top. What a roller coaster!

Now, this leaves some interesting things for the next few games. Schilling - his ankle is not good. Based on how he looked, I think the Sox would have been better off starting Pedro, following with Wakefield. Sure, the yanks scored two runs off him tonight before the 5 run Red Sox rally - but last year in the ALCS, he utterly baffled them. That would have rested Schilling until game three, which will be in Boston. If they didn't know his ankle was that bad, they simply weren't paying attention. All good for me though - he threw a lot of pitches, and the Yankees are up 1-0. Bring on Martinez - I like the Yankees chances.

Cincom Smalltalk support - Dave Wallen, specifically - has provided an answer to a fairly sticky problem some CST developers on Windows have - dealing with Windows API functions via the windows.h header file. There are a few problems that the DLLCC parser has with it - here's a solution with some instructions:

I played around with getting windows.h to parse. Only a few #defines were necessary in advance, and the thing now parses. The file is uploaded, but here are its contents anyway. Beware that once the parsing was completed, I was able to generate methods, but it took at least an hour on a PIII 500. I set GrowthReg. Upper Bound to 100MB, and it used it all. Also, I did not select any "macro"s for methods. I'm not sure if it works or not, but it would take forever, so I disabled the macros section. All the other sections did receive methods.

This file works on an installation with MSVC 2003. Not sure about other versions.

/*-------------------------------
 *  CincomWindows.h
 *-------------------------------
   This file was successfully parsed by VW 7.2.1 under the following conditions.
   1. Create a directory, called Include, with two subdirectories, called
      C_include and Win_include. 
   2. Copy this file to Include.
   3. Copy the entire \Microsoft Visual Studio .NET 2003\Vc7\include directory
      to C_include.
   4. Copy the entire \Microsoft Visual Studio .NET 2003\Vc7\PlatformSDK\Include
      to Win_include.
   5. In the Builder, enter these three directories as the include directories,
      C_include, then Win_include, then Include.
   6. In the Builder, enter this file as the only include file (CincomWindows.h).
   7. Click Parse Files, and the files are parsed.
  
   The following files were parsed, and no errors were reported. As an aside,
   this list was generated by printing to Transcript, in the method
   CPreprocessor>>include. It won't compile without AdvancedTools loaded.

   ('basetsd.h' 'cderr.h' 'cguid.h' 'CincomWindows.h'
   'commdlg.h' 'ctype.h' 'dde.h' 'ddeml.h' 'dlgs.h' 'excpt.h' 'guiddef.h' 'imm.h'
   'lzexpand.h' 'mcx.h' 'mmsystem.h' 'msxml.h' 'nb30.h' 'oaidl.h' 'objbase.h'
   'objidl.h' 'ole2.h' 'oleauto.h' 'oleidl.h' 'poppack.h' 'propidl.h' 'prsht.h'
   'pshpack1.h' 'pshpack2.h' 'pshpack4.h' 'pshpack8.h' 'reason.h' 'rpc.h' 'rpcasync.h'
   'rpcdce.h' 'rpcdcep.h' 'rpcndr.h' 'rpcnsi.h' 'rpcnsip.h' 'rpcnterr.h' 'servprov.h'
   'shellapi.h' 'stdarg.h' 'stdlib.h' 'stralign.h' 'string.h' 'tvout.h' 'unknwn.h'
   'urlmon.h' 'winbase.h' 'wincon.h' 'wincrypt.h' 'windef.h' 'windows.h' 'winefs.h'
   'winerror.h' 'wingdi.h' 'winioctl.h' 'winnetwk.h' 'winnls.h' 'winnt.h' 'winperf.h'
   'winreg.h' 'winscard.h' 'winsmcrd.h' 'winsock.h' 'winspool.h' 'winsvc.h'
   'winuser.h' 'winver.h' 'wtypes.h')

   Note: the RPC entry below is there to avoid a parsing error. The version may
         not match your installation, in which case you may need to change it to
         the value specified in the file rpcndr.h.
*/


#define WINVER 0x0500
#define _WIN32
#define NOWINRES
#define _M_IX86
#define __declspec(dllexport)
#define __RPCNDR_H_VERSION__ 475

#include "windows.h" 

So there you go. Your mileage may vary on other versions of windows.h from other revs of the OS, but the above should give you a headstart on how to proceed.

I have to admit - with only 3 hours (as I write this) to go before the start of game 1 of the ALCS between the Yanks and Sox, I'm losing my ability to focus on anything else. Work, politics, other news? Blah, who cares. Bring on game 1! Another funny thing related to my earlier post about this rivalry making it into politics - Kerry mentioned the series again today. The hype for this is getting to be bigger than Super Bowl level...

Looks like the Yankees/Red Sox series transcends even politics:

"We all want the same things. We want our country to be respected in the world, we want good jobs, and we all want to beat the New York Yankees!"

Ironically, this was said by Kerry in Ohio - and the Yankees AAA farm team is based in Columbus. Heh.

The Croquet release has been slashdotted

I've been using Wikis for many years now - the VW Wiki at UIUC has been a valuable resource for Smalltalk developers for a long time, and the Cincom Smalltalk Wiki is a place where we (the Cincom Smalltalk team) put out information on what we have coming down the pipeline. As a team, we use Wikis for internal communication as well. Periodically, I get asked "So what's the difference between a Wiki and a blog? Both allow for user editable web page - why use one over the other?"

It's a good question. As it happens, the two kinds of websites have some rough similarities (both are user editable in some sense) - but they also differ quite a bit. In broad terms, a Wiki has what you might call an emergent voice - it's an agglomeration of the input of the community of users who are interested in the content being managed. Here's an example - the history for one of the pages on the VW Wiki:

See how there's a long list of editing changes across time? That page was originally created in October of 1999 - and the last (real) edit of it was in April of this year. Looking at the list of edits, you can see that the page was edited by more than one person across a long time interval. As the interests of people have waxed and waned, content has been added (or subtracted) from the page. Over time, the direction of a page tends to focus in - at least on this wiki.

Other wikis can come to resemble conversations - have a look at the popular C2 Wiki, originally started by Ward Cunnigham many years ago. A glance at the recent changes page shows many changes over the course of a single day - have a look at this page, for instance - you can see that many people have contributed content over a long period of time.

A popular wiki can quickly spawn off dozens of side points and conversations - a good example of this is the Atom Wiki, created to engender conversation and conclusions on the Atom syndication format (an alternative to RSS). The limitations of a wiki for conversational purposes cropped up rather quickly here, and spawned a mailing list. That list tends to have dozens (sometimes over 100) messages a day, quite frequently on the finer points of things like date formats. This is barely manageable as a mailing list; it completely fell apart as a Wiki

In general, Wikis work well when you have a fairly limited number of people contributing content, and a larger body of readers and occasional contributors. If the amount of content creation spirals too quickly, there's just too much information - it becomes a lot like a USENET group that's been taken over by trolls. Even if most of the content is good, there is such a thing as too much of a good thing.

How do blogs play into this? Blogs allow for a much tighter focus on editorial control. You can disable comments, for instance - and thus ensure that only one voice is being heard. Even with comments enabled, they typically are not shown when a visitor comes to the site - the visitor has to choose to view comments (as opposed to a Wiki, where anyone can add any content they want anywhere on a page).

Have a look at this screenshot of my blog, for instance:

There are individual articles, and the start/end of an article is determined by the author. With a wiki, a reader can re-edit to their heart's content; a blog is much more like an editorial page - and comments are akin to letters to the editor. The author of the blog retains ultimate control; he can choose to allow comments or not, and - like a newspaper - decide whether a given comment (or article) stays or goes.

With a wiki, you don't have full control over the content - and attempting to keep control will lead to "last one to edit wins" types of battles. A wiki is most useful when you want to create a community accessible site, where constructive criticism is encouraged. A blog is far more useful for getting your specific message out, without changes being made by someone else. Sure, other bloggers can make comparison posts linking to yours (I do that all the time) - but the author still retains complete control over the content - and ultimately, over the message.

There's another very large difference as well. A wiki is much more like a "traditional" website than a blog is. Once a page is created on a Wiki, it's just there (unless, it's deleted, just as with any other site). A blog is much more ephemeral. Typically, the main blog page will show the last N items or the last N days worth of items. Sure, everything is accessible via the archives or via search - but it's not just there. The top page of a Wiki, with the items it points to, is a (relative) constant. The top page of a blog changes with every posting. Thus, a Wiki is better for information that needs to be "sticky" - a blog is much more akin to the morning newspaper. Fresh, but changing.

So, getting back to the comparison - under which circumstances do you want a wiki, and under which do you want a blog? If you want to encourage input from a community, a wiki is a good tool for that - however, you need to be aware of the limitations. Most people are uncomfortable writing html, and many people will be uncomfortable using wiki markup (most wikis support a simplified markup scheme). Given that, you are going to get feedback from a limited subset of the total readership. This subset will be even smaller if you aren't targeting a technical audience. If you want to get a personal voice "out there" talking about your products and services, blogs are the way to go. A caveat here though - a blog is only useful if the author(s) post regularly on topics of interest to the target audience. It's very easy to start a blog - it's much more difficult to sustain one.

In many cases, you are going to want to use wikis for some circumstances, and blogs for others. Either way, commitment is very important. Blogs need regular postings - wikis need maintenance and pruning. If you plan to use either or both, you need to walk into the venture knowing full well that a decent sized time commitment is necessary.

Croquet is out - grab the release here. For a "first impressions" kind of post, see Michael's post

The Register reports that a Chinese couple was denied their request to name a child '@' - as it cannot be directly translated into Mandarin. Lucky for that poor kid, who still has to grow up with these morons....