VOICE Home Page: http://www.os2voice.org
|By Menno Willemse © December 2001|
This year, Warpstock Europe was held in the Belgian university town of Diepenbeek,
close to Hasselt. Since this is only a three-hour drive from my front door, I decided
to go and seek out the Katholieke Hogeschool Limburg. The trip went rather
smoothly and I was able to find the site with little trouble. Just follow the signs
to Diepenbeek and no doubt you will run into a sign to the University Campus. At
one of the university buildings, I found a notice saying "Warpstock Thattaway".
Since I had reserved in advance, I quickly got in. Almost immediately, I ran into
Bob St. John, who apparently wasn't vacationing in the Bahamas with his ill-gotten
gains after all. Even though Bob didn't give any lectures as far as I can tell,
it was nice to put a face to the name I only got from this and other groups. (The
photo got erased in a SmartMedia event but more about that later).
I had decided to roam the compound for a while first. There were several exhibitors.
Innotek was there showing their new Virtual Machine for OS/2. I personally saw a
pre-release version of the thing run Windows XP. The Virtual PC is just that. It
emulates a PC with a given chipset, video card and network interface whatever hardware
you have. You can make it use certain OS/2 directories as shared drives so you can
exchange information between OS/2 and the guest operating system. In a later lecture,
Achim Hasenmüller said that the Virtual PC will run anything except BeOS because
of some bug. Windows 98 was recommended against. It does work, but due to the fact
that '98 disables the cooling cycle of the CPU to make it run faster, your CPU will
be at 100% activity constantly. One nice side effect of running XP under OS/2 is
that no matter how many devices you swap, the virtual PC will still look the same
to XP so no re-licensing is necessary. Let's hope MS doesn't take steps to make
XP not work under VirtualPC. (But let's face it, do we really care?) Innotek
was selling preview versions for fifty euros, but I didn't go for it because a)
my PC is not fast enough to run Windows even without emulation, and b) it only works
till newyear after which the full version will be available.
My first lecture was one by Christian Langanke about HyperText/2, which is a
preprocessor for producing INF and HLP files. He is offering it for free in the
hope that more people will produce decent helpfiles with their applications. I think
this program fulfills a need. The IPF file (sources for INF/HLP files) is not very
complicated. It is a flat ASCII file where you specify things like Bold, Underline
and Italics with tags, much like HTML. However, to produce a good helpfile is a
lot of work. The input file for HyperText/2 looks much cleaner and it will produce
a nice info-file with indexes on the left and pages on the right. If you program
at all, give it a try.
Next was, of course, lunch. This was provided by the University's catering. It
was nice, but they gave you all the food at once. This means by the time you've
finished your starter, your main course will be cold. Experienced cafeteria feeders
eat the warm stuff first.
My next lecture was about the Golden Code trace suite. This is a set of tools
that can really help solve problems. They compared working without them to gunfighting
in the dark. Basically, the tools are a network tracer, a kernel tracer (so you
can see what Netscape is actually trying to do in its more meditative states) and
the Trace Analyzer. The network sniffer is a nicely low-profile piece of software
that hooks into the NDIS driver and dumps everything the network card sees into
a buffer, which can then be saved to a file. The kernel tracer does the same for
kernel activity. These files can then be analyzed with the Trace Analyzer, which
is written in Java. Most network protocols can be displayed. As an example, Greg
showed us a log of a web browsing session. This is the sort of tool that I really
want on my portable for those hard to find problems. I wonder if I put it nicely
to my boss...
I didn't stay for the Kernel trace part of the session, because I wanted to go
to another lecture given by Starfire on Titan. Titan is an administration tool that,
when used properly, can take a lot of load off the shoulders of your system administrators.
Central to the product is a database containing system administrative procedures
for a variety of platforms including Windows, AIX and of course OS/2. (Otherwise
what would they be doing at Warpstock?) All the operations are browser-based and
SSL encrypted for security. Starfire also keeps track of all the modifications to
the system. A very nice feature is that if the target machine happens to be down
at the time the command is issued, the modification will be remembered and once
the machine has recovered from its Blue Screen of Death, it will be performed at
that time. One weakness is that once Starfire is in place, you have to stop using
the original procedures for management or the machine will go out of sync with the
database, which can lead to unforeseen problems. At the end of the lecture, T-shirts
were given to all comers!
The lecture on Virtual PC concluded the Friday event and I was off in search
of my hotel. I was told that Hasselt has the highest bar density in the whole of
Europe, and I believe it. It is impossible to go for a walk in Hasselt and not have
at least one bar in sight. And of course, Belgium is famous for its beers. It was
a good evening.
Day dawned. Quick shower, breakfast and off to Diepenbeek. The first lecture
that day was on all the new stuff that goes into MCP2 and ACP. It seems that most
of the work is going on under the hood. New drivers, new ways of tweaking the kernel
and such. It also showed a time line that shows that by 2006, IBM will have achieved
full platform independence, whatever that means. It didn't sound altogether hopeful,
but then again, IBM is not known for its glowing enthusiasm for OS/2. Not since
David Barnes was last seen. (I wonder what he is doing now).
The next session was done by Rainer Feuerstein and Alain Rykaert. Rainer got
rather technical and showed us lots of REXX code. Learn this, guys! Audiences are
NOT interested in bloody source code unless it's a programming session! Alain
took over and showed us the stuff he used to install 60.000 seats at France Telecom
(single-handedly? wow!). This involved a central database with configuration information
somewhere on the net and a carefully prepared bootable CD. I'm afraid it completely
bypassed all IBM's installation routines. It partitioned (with LVM) and formatted
the hard drive, then unzipped pre-configured images on them. TCP/IP was configured
by careful hacking of the source files using a REXX script called cube that lets
you edit files from the command lines. Even while explaining about it, it took him
only fifteen minutes to install a fully functional Aurora server, which is pretty
fast! What I remember most, however, is Alain's enthusiasm about the product he
was working with.
After lunch, I went to Kim Cheung's lecture on how to call REXX scripts from
a C program. This is very useful if you have an application that benefits from scripting.
For instance, you could write an email client that automatically sends the latest
news to all your friends... Oh no. Been done before. Kim showed how you can fire
off REXX scripts from C programs, how to pass parameters to the REXX script and
how to share variables between C and REXX. It seems that even though REXX is an
interpreted language, that doesn't mean it's slow. Kim showed us an example with
about 3000 threads running at once. Pretty impressive for a notebook.
The last session of the day was about XWorkplace, a free Workplace Shell enhancer,
hot off the compiler. The lecture was given by Ulrich "DJ Mutex" Möller.
I'm pretty impressed with both Mr. Möller and XWorkplace. Even now, when he
says the project is under "heavy development", it looks very polished.
It comes in several languages, it uses the WarpIn installer, and it is extremely
well documented on the website (www.xworkplace.org) as in the program itself. Ulrich
is one of the people who knows the Workplace Shell inside out and still resists
the hackerish trend to produce undocumented oracle utterings.
Today, I also took the opportunity to visit Mensys and Team OS/2 from different
parts of Germany. I scored some software for e-Biz (Apache, MySQL, PHP, Perl) and
some Palm utilities. (Gotta back up that thing one of these days).
Finally, Saturday was the night for the beer bash or Social Event. There was
one band consisting mainly of Mensys and HCC-attached people and one blues-band.
Roderick Klein could be observed installing eCS on his portable even during these
proceedings! Roderick, for Chrissake! it's *party time*! (Mary Sparlock, if you're
reading this, I've got a nice photo of you. Mail me if you want it). I didn't stay
too long. Since I had to drive to the hotel, I was on non-alcoholic drinks for most
of the evening. Finished off with a nice Palm at the hotel. [Editor's note: A
Palm is a beer.] Bed.
The next session covered how USB devices could be used from normal applications.
Markus explained how to set up connections to USB devices and used a USB radio as
The last session went into the development of USB drivers for all kinds of devices.
I'm afraid Markus lost me pretty quickly there. I'd like to know this kind of thing,
if only to implement a Netdrive plugin for USB Mass Storage devices that adhere
to the other half of the standard than IBM does.
The final session (after lunch) was by Kim Cheung again. He showed a diskless
workstation and the tools he developed (or helped develop) at Serenity to enable
remote booting over Netbeui and also TCP/IP. He then proceeded to install several
applications on the diskless station simply by dragging them over to the machine
using WiseManager. (Who would want XWorkplace and Object Desktop though?)
This shows that the WiseMachine thing you get in eCS is only part of a larger product.
After that session ended I went over to the Mensys stand in search of some last-minute
goodies. I saw that they were selling AOpen Crystal Labs Soundcards that are well
supported in OS/2, so I got one of those for my next machine. I then saw Daniela
Engert and got an original copy of her DANIS506 driver off her, which would enable
my thinkpad to read the SmartMedia cards I use in my camera using a PCMCIA adapter.
I am happy to say that it worked like a charm, but you have to be careful when you
use the cards and always use the "EJECT" command on them before
removing them. This is true even if you only read from the card. I think I swapped
the cards too quickly and corrupted the FAT on the card with all my WarpStock photos
on it. The only thing I can think of is that OS/2 was writing the latest access
dates from one card onto the other, which corrupted the FAT. Luckily, I was able
to recover most photos by running a chkdsk on the card (most notably the one I took
After that, I decided to head home. It was a nice conference. I had been able
to meet and pick the brains of some very knowledgable people. I was slightly worried
by the corporate presentations. I suppose it's only good business sense to calculate
in the possibility of OS/2's impending demise, but a more constructive attitude
would have pleased me more. On the whole, the best sessions were those done by OS/2
users. Their enthusiasm for the OS/2 operating system (and its derivates) are good
to see. David Barnes commented on this years ago: It is the users that keep
OS/2 alive and happy. The users would not let the product die.
See you all at Warpstock Europe 2002!
[Previous Page] [Newsletter Index] [Next Page]
VOICE Home Page: http://www.os2voice.org