Subject: Jake's BorCon 2002 Report--Wednesday
Date: Thu, 23 May 2002 21:00:43 -0500
From: "John Jacobson"
Newsgroups: borland.public.delphi.non-technical
You may have already heard this, but Chad Hower and by association the Indy
group acheived a bit of notoriety Tuesday night by going to every PC in the
computer lab and installing a little program that popups an ad for Indy,
annoying the hell out of many users of the lab. Rumor has it that he was
thereby banned from the lab for the rest of BorCon. Odd behaviour there from
Chad and friends.
Wednesday morning I went to "Advanced Patterns in Delphi" by Mark Miller.
This was an excellent session on a topic that really deserves more attention
than many people are giving to it. Yet another set of books I must buy. The
wife will love that. Anyway, Mark is a great presenter.
I wasn't particularly bowled over by anything in Chris Thomas' keynote
speech. Rah-rah web services. What requires more computing power and bandwidth
is good for Intel. No big showstoppers there.
Lunch on Wednesday was a box lunch. I stole away to the most isolated table
in the area and quietly ate my lunch, ruminating over the events of BorCon
2002. One of the odd things about BorCon that needs to be fixed is that there
are Birds of a Feather sessions at 12:30, but when lunch is the
sit-down-and-waiters-serve-your-meal type of thing, you would get only the
initial salad appetizer if you left at 12:30. It would be a good idea if they
would have box lunches available for those who want to go to a BoaF session.
For the most part, I thought the food at BorCon 2002 was very good.
Wednesday afternoon I went to Chuck Jazdewski's "Moving Delphi" session. He
is very passionate about what Delphi is, but not always as articulate about
that vision. He basically described the Delphi "feel" as a combination of the
Form Designer, Editor, Compiler and Debugger, while Delphi is defined as the
superset of that "feel" with the Components and Language. He felt it was
necessary to define Delphi before describing how it was ported to Linux and
now to .NET. One of the interesting tidbits to come out of this session was
the fact that one of the biggest obstacles they had to overcome in creating
Kylix was the linker, as Linux itself seems very biased in favor of C
constructs in general. (The new inline assembler was also one of the biggest
pains.) Other obstacles to porting Delphi to Linux was that Linux has no
built-in exception mechanism, a loader bias in favor of C, and no concept of
resources. He stated that Qt was chosen for it's friendliness to the Windows
model that Delphi was based on. He also revealed that Delphi.NET will treat
assemblies just like packages. One point in the session made me feel old, and
that was when Chuck asked how many people had programmed in 16-bit mode and
only about one third of the audience raised their hands (he was discussing a
silly limitation in Linux that unnecessarily limits the number of CPU
registers available to the programmer when this came up).
Eddie Churchill's session on "Delphi Framework Preview for .NET" was next.
Tidbits include:
--Delphi.NET code will run faster on a P4 and better because the .NET JIT
compiler is able to perform optimizations on these chips that D6 does not.
--The redistributables for .NET might be a very large 20MB but if you tally up
all the BPL's that come wth Delphi they come to over 40MB.
--Delphi.NET executables will be very small for the same reason that D6
executables compiled with packages are very small, with the exception that the
.NET redistributables will most likely already be on the user's machine after
a while.
--If you wondered whatever became of Corel after their failed merger with
Borland, just look up the "ROTOR" project that is an Corel/MS collaboration to
port .NET to FreeBSD and XP (IIRC).
--D7 will have an MS SQL Server driver in dbExpress.
--As Eddie Churchill was discussing the fact that the relationship between
Borland and Microsoft has never been better than it is now, thanks to .NET,
the thought occurred to me that instead of disdaining the fact that .NET and
Java seem to have ripped off ideas from Delphi, we should appreciate the fact
that ideas spread through reuse. Therefore if you think an idea is good you
should be happy when it is reused, rather than going on and on about theft and
nefarious business practices.
--Eddie Churchill stated "Object Pascal is still Object Pascal" when
discussing Delphi.NET. It seems there is still a difference of opinion here
between different people at Borland.
The closing keynote was the usual gift giveaway and cheerleading session.
Dale once again shot T-shirts with the T-shirt cannon and then used it to
randomize the drawing of Borland Bucks for prizes. I've attended three BorCons
so far and I have never won anything with my Borland Bucks. That's too bad
too, because someone near me won and so I got a peak at what they had won.
Several excellent books, Topaz for Delphi, etc. My search for a replacement
for Interbase would probably have been aided by my winning one of those bags,
but alas, the fates were not smiling on me this time.
It only takes a few minutes for the place to clear out when the closing
session is over and what was a bustling and festive collection of excited
programmers getting exuberant over the future of programming is dissolved into
the quiet clanking of doors and the muted discussions over who's flight leaves
when. Walls of glass and steel that once housed a sea of humanity now echo the
sounds of silence. The diaspora of Delphians is once again commenced as we all
return alone and in small groups to our small corners of the globe and plot
and plan how we are going to get to the next Borland Conference.
[Fade to black. Roll credits.]
