Re: Where to download free Fujitsu COBOL compiler

On Nov 6, 1:06 pm, "Pete Dashwood"
<dashw...@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
Richard wrote:
On Nov 5, 1:59 pm, "Pete Dashwood"
<dashw...@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
William M. Klein wrote:
The Fujitsu V3 compiler was NOT for commercial use either.

Well, actually, when it was first provided (in 1996 or so) it could
be used for anything, but for the last decade or so it was
explicitly posted with "restrictions" saying that it was not for
commercial use.

I received it (for free...) in 1996, as you noted. I developed a
commercial system with it using PowerCOBOL and it is running to this

(This, as Bill noted, did not violate the license at the time.)

Originally, Fujitsu marketed it into the vacuum created by the
withdrawal of support for Micro Focus Visoc, and I was one of the
Micro Focus customers who changed to Fujitsu at that time.

I found the product to be excellent and took out maintenance and
updated as new versions became available.

By the time version 5 arrived, it was a pretty good development
environment, although the IDE has always been "primitive" compared
to Eclipse or Visual Studio.

It was in version 6 that everything turned to custard. Fujitsu
(USA), later to become Alchemy, decided that piracy was a major
concern (I remain unconvinced to this day; I was talking to a number
of Fujitsu customers and I never met one who was making illegal
copies for supply to someone else; for the most part, their user
base was honest COBOL developers for whom not having a backup put
their businesses at risk...) and implemented an insane system of
registration using Casper on a remote server. Although this was
supposed to be helpful for users, inasmuch as the system would allow
you 30 days to download and use a copy of the software if your main
implementation went down, the process involved was really unwieldy.
For users outside the continental USA where time differences matter
and you can't just dial an 0800 number, it was frightening. What if
the Casper server was down or wouldn't recognise your registration?
Added to that, the procedure for transferring a licence to a
different machine was just plain silly. (It required a floppy long since you saw one of those...? :-))

It is not true that it required a floppy disk, though this was the
usual mechanism, it could be done to any medium accessible to both
machines. eg a USB drive.

The documentation which arrived as printed paper notes with version 6,
stated that a floppy disk should be used. There was no mention of any other
device. I never tried it with a USB drive but I'll take your word for it.

It works with a network share too. I had it on my laptopn with no
floppy drive.


Version 3 was initially for Windows 3.x (I have a CD here) and was
later implemented for Windows 95. Apparently it can be made to install
and run on XP though it can fail to do so.

I can quite understand why Alchemy don't bother with it anymore.

It looks to me like a COBOL company that has no commitment to COBOL
(except maybe .NET COBOL)

This is the "flagship" product and most of the other tools require you
buy it.

# NetCOBOL for .NET

Certainly they have written a bunch of stuff to move 'legacy' to
Windows. Presumably they make more money in that environment.

These are the "second class citizens" and poor relations for which you will
pay annual maintenance and receive very little.

Why do you think that they are '2nd class' ? Is it just because you
have no interest in them ? I bought Cobol for Linux (when it was on
special) and have not paid maintenance yet have received updates to
7.1, 7.2 and 7.3.

SPARC is certainly a focus for Fujitsu because they sell SPARC systems
and even co-operate in chip development (via the UK ex-ICL branch).

There have been new versions of SPARC and HPUX but I don't follow
these closely.

It isn't like they are
implementing the 2002 COBOL standard.

Is anyone ? Is there a .NET of 2002 ?

Since version 6 most of the "new
releases" have been about better documentation, and fixes to obscure bugs
you would be unlikely to encounter anyway. There hasn't been innovation of
new features, as far as I can tell., however, I haven't had a lot to do with
later releases and am going on what others tell me.

# NetCOBOL for Linux
# NetCOBOL for SPARC Architecture
# NetCOBOL for Windows

Each Linux 7.x have had new features. The old Fujitu site listed the
enhancements which was useful.

My own experience is with the Windows product and it has always been
excellent. But I don't think there is a committment to update any of these.
That is what prompted my comment.

(If I'm wrong about this and they are actually working to implement more of
the 2002 standard, I unreservedly apologise to Alchemy.)

Exactly. Free Internet Explorer was to drive Netscape out of business,
Free MSN (the original '95) was to kill the internet. Free C# was to
kill Java. Bung (or something) is to kill Google.

I won't comment on the other points, but free C# was never designed to "kill
Java" (although that was probably a hoped for side effect...). And there is
no past tense; C# is STILL free and has been for 7 years. I believe the idea
was to get people quickly onto .NET and it has largely succeeded in doing
that. So we could expect, having  largely achieved the market objective,
they would pull it. So far, they haven't.

Fortunately Microsoft have been failing at many things because when
their market share is high the free stuff stops and they rake in the

I'm having a little difficulty with "raking in the loot" and "failure"...
perhaps we have different definitions of "success" :-)

Failures have been 'Bob', the original MSN. Zune, Danger/Sidekick,
XBox has never made any money, Playforsure, the original .NET stuff
(prior to what is known as .NET now). MS Mobile smartphones has around
5% market share and falling.

'Microsoft Live Search' failed, buying Yahoo failed (though that
probably saved MS from an expensive failure), Bing only gets 3-4% of
the market now though it was up to 10% and this is even when 'updates'
change the default search to Bing.

IE share is falling and IE6 is still the largest version (and Firefox
is > any IE version).

Vista 'failed' in that >60% of Windows is still XP or 2000.

MS ties up OEMs by bullying them (ie stringent discount dependencies)
and this is where MS drags in most of its revenue.

So their market share for Visual Studio and C# isn't high? I did some
searches... In the last 2 years, growth of C# use has been phenomenal and
now exceeds both Java and VB. However, there is no clear picture I could
find, and different sources have different views. The only conclusion I
could make was that C# and PHP are growing strongly, while Java seems to be
in decline.

Some examples...

That is based on _book_ sales. People who don't know a language buy a
book on it so the 'massive growth' is based on people who _don't_ know
it. Of course Java book sales fall off because it is a mature product.

Anyway JavaScript book sales are growing faster than C# and will soon
overtake it. This is probably due to JavaScript being not only for Web
2 and AJAX but also for Moblin, Android, Palm Pre, and many other

(Tiobe is interesting... it shows COBOL as 23 in the language list, and Java
as supreme ruler, but on a declining trend.)

Java is 18% and C# is 6th with 4.4%, around the same as Python and
PERL. C has twice C#'s growth rate and PHP has triple C#'s growth.

Your comment is just spin. Yes, 4% market share for C# is _not_ high.

Looking at the graph on that page the end of 2006 line for C# shows
about 3.5%, 4 years later it is 4.4% That is hardly "phenomenal".

"I used to write I can do anything."