Microsoft's ASP.NET Teams have released an upgrade to Web Matrix, Mini Me to VS.NET.  It looks great, particularly its HTML editor.  I wish they would release the source code for Web Matrix though.  That would save a lot of folks a mountain of time.  Same comment applies to other projects by Microsoft employees like RSS Bandit which is open source only if you join the project. [correction: RSS Bandit source code is now included in the installation.  Cool.]

Loosen up guys!  I swear I never ogled a Linux box.  Well, maybe for just a couple of blinking prompts, but I never gone all the way.

Congratulations to "Dave" for his first year not-smoking anniversary and my thanks for still being here.  Longest I have gone was six month.  Yeah, Dave.  I hear ya.  I will visualize my way out of this damn expensive habit too.

Tim Bray is going Loopy trying to find a way to get around CLR's array bounds checking overhead by trying different C++ expressions.  Short of using unsafe code, there is no reliable way of doing this because no legal sequence of IL bytecodes can ever tell CLR to bypass array bounds-checking.  If there is, then it is a security hole.

However, one can *make it easier* for the JIT implementation to figure out that there is no need for bounds-checking inside a loop by using simpler or more direct expressions like C#'s 'foreach' statement.  Unfortunately, C++ doesn't have an equivalent to foreach.  I use mostly C# in my .NET projects because I found that C# compiler generates more optimized IL bytecodes in general.

Microdocs, aka Dr Elwyn Jenkins, posted a rather long response to my rant about searching for WebCore source using Google.  He wrote:

"Now what exactly do you want Google to fix Don? That there are a whole lot of other pages now talking about WebCore that are not relevant to you?"

"No longer is it adequate to simply search Google using a single word for anything you want. Any word you use now needs to be contextualized. There needs to be other words surrounding the keyword you are after to identify the theory or ideas about that keyword which are important to you. After all, if Google is to deliver a page of relevant search results, how does Google know which theory, or idea, or context, you are thinking that this lone word is going to be in?"

He has completely missed the point I am making.  "WebCore" is a NAME, somewhat uncommon name at that.  While I agree that discussions about WebCore are valid search targets as WebCore source code, it is NOT unreasonable to expect the NAMED OBJECT to be on the FIRST PAGE instead of being buried under discussions ABOUT the object.

This is not about one word key search, but about common sense.  To find out about Safari, I wouldn't just use "Safari" to find what I want.  I would use "Safari Apple Browser" to remove ambiguity.  Finding reviews is easy enough by adding "reviews" keyword.  Now how the hell do I tell Google whether I want to find an object or discussions about that object?  If Google has problems identifying a blog, so will Google users!

Google's business is helping people finding things in return for exposure to advertisements, not forcing people to adapt to changes that Google can shield them from.  Google should work according to common sense so even my wife and son can find things easily most of the time.  Anything short of that belongs in the junkyard.

This afternoon (I have been working late) I had two dreams.  First one was about why meatspace conferences matter.  Most of the details were lost, but a single piece of thought remained: deeper committment.  Virtual interactions are simply not as binding as F2F interactions at all levels.  I got no further than that unfortunately.

Second dream was about Typepad.  Without a clue, I dreamt they must be working on adding moblogging and fotologging to MT.  I dreamt that, if they are smart, they would have some carriers and cellphone companies onboard to offer Typepad services to cellphone users.  I guess I must have been curious about what Typepad guys were working on.  If what I dreamt is right, Typepad will soar ahead of Google/Blogger.

I received this in the e-mail today:

Tue Jun 10, Mountain View

Scott Cook, Doug Carlston, & Trip Hawkins with Stewart Alsop   

"Jurassic Software:  A Look Back at the Beginnings of Consumer Software"

 

7:30 PM Panel Discussion
Computer History Museum
1401 N. Shoreline Boulevard
Mountain View, CA
http://www.computerhistory.org/events/directions/eventmap.jpg
Free. (Suggested donation of $10.00 from non-members.)
Advance reservations required.
Please RSVP by Thursday, June 5, 2003
For more information or to sign up online go to
http://www.computerhistory.org/jurassicsoftware_06102003/
or call 1-650-810-1027

The entrepreneurs who created the first consumer-software companies gather to reminisce about the early days and recall the lessons learned in the founding of a new industry. Scott Cook is co-founder of Intuit; Doug Carlston is co-founder of Broderbund Software; Trip Hawkins is founder of Electronic Arts and 3DO. Stewart Alsop was the publisher of P.C. Letter and founder of the Agenda and Demo conferences. Stewart will moderate an informal discussion of the beginnings of consumer software and the entrepreneurs have promised to bring pictures and products to show and tell for the audience.

Wow, voices from the past.  I got the e-mail too late to attend the meeting though.  It is tonight.  I remember Scott Cook presenting a little program he wrote at SEF (Software Entrepreneur's Forum) ages ago.  He was a SEF member and his program was selling amazingly well, so we were there to hear his experiences.  I think Phillipe Kahn was there also talk about how his Turbo Pascal was doing.  A lot of memories from that little Hof'brau at Palo Alto golf course.

I first met Doug Carlston when the company I worked for had a baseball game with Broderbund.  I remember sitting in the shade talking to guys who wrote Loderunner and a chopper rescue game (I foget the name, but it was good).  Later, I kept running into them because the next company I worked for was into educational games.  I really admired the way Broderbund was doing business back then.  Compared to them, Sierra Online was dorky and EA was too glitzy (I wonder what happened to Bill Budge, the first software artist?).

I never met Trip Hawkins, but I remember hearing about 3DO and how much he was asking from game developers to develop for 3DO.  Yikes.  He had guts.  Too bad, the game console market turned out to be much tougher than it seemed.  He shouldn't feel too bad that 3DO crashed because Microsoft is not doing too much better than him.

If that was the beginning of consumer software, I guess I was there from the beginning.  We were naive back then, but it was exciting times.  Still we weren't naive enough to think anything will sell.  The picture is somewhat different now.  Internet, Microsoft, open source, and Bubble looms in the picture.  I can still feel the excitement though.  It is like wunderlust.  Once you have it, every little road you come across pulls you and a vision of golden cities fills you.