Serious conformance BUG in lcc-win32 still present

[James Kuyper]

Right, and people are under no moral, ethical, or contractual obligation
to regard his compiler (a compiler for a language that isn't C!) as an
option for their use.

You've got it! That's the whole point.

If a compiler vendor is not prepared to fix his product to conform on this
very simple point, how can anyone expect that vendor to be willing to
ensure his product conforms to the tricky, intricate points of the
Standard?

You can't. You can only count on him to be willing to ensure that his
code conforms to those aspects of the standard that he approves of. If
your list of approved features is not a sufficiently close match to his
list, don't use his compiler - it's just that simple.

Refusing to fix a simple BUG gives the impression of a slipshod vendor who
is happy to ship crap. ...

This is not a BUG. It's not even a bug. It's a deliberate decision not
to fully conform to the standard because he thinks that delivering a
compiler which conforms to this particular requirement of the standard
would be, in your words, "shipping crap". He's entitled to that
opinion, and he's entitled to implement his compiler accordingly; it's
only a problem if he claims full compliance to the C90 standard, and he
doesn't. If someone decides to build a Fortran compiler, are you going
to criticize him for not deciding to conform to the C standard?

 

[Andrew Smallshaw]

It is a vicious circle. I definitely won't appear on a "database of
registered customers" until you take standards conformance seriously.

I think your attitude is very bad. You should act on BUG reports received
from ordinary users as well as registered users. Don't you have any pride
in your work? Why don't you care that it has serious problems? Is money
the only thing that matters, and you don't mind if what you produce is a
load of crap as long as you get paid?

Well, let's see. A bug is an undocumented featrure. On that level
the very presence of the the option you are using is a bug. Don't
use the option that causes the buggy behaviour and it won't affect
you. If you want more support than the bug introduces you have
already been told how you can get hat added. It seems to me that
you have nothing to moan about.

There isn't much software out there that doesn't lack features that
I like (even if that feature is only a smaller footprint). I don't
start accusing the authors of that software of incompetence or
negligence in not providing the features that I want.

 

[Kenny McCormack]

....
There isn't much software out there that doesn't lack features that
I like (even if that feature is only a smaller footprint). I don't
start accusing the authors of that software of incompetence or
negligence in not providing the features that I want.

I'm beginning to think that Tea Pot might be Jacob's sock.

Think about it - the effect of this thread has been to get most of the
CLC rank-and-file (*) defending Jacob and his software.
Who could ask for more?

(*) Not, mind you, the top, top echelon (Heathfield, Thompson, and a few
others). Although, now that I think about it, it is possible that
Thompson has weighed in here (defending Jacob); I'm not sure...

 

[Antoninus Twink]

I've decided that you're not Heathfield (unless he's pulling some
funny shit in this thread, which I wouldn't put past him), and you
don't sound like Falconer, who doesn't seem to have a problem with
Jacob anyway.

It's definitely not CBF - there's no way he has the mental faculties to
express himself coherently in three separate messages.

I agree that it doesn't /sound/ like Heathfield, but we have to remember
that Heathfield is a deceitful toad with a burning hatred of Jacob, and
he's also the only "regular" whom we know for sure has used sock puppets
in the past.

If it is him, then his shtick in this thread (acting all
holier-than-thou in public, while trying to stick the knife in under the
cover of anonymity) is an all-too-believable modus operandi for a
cunning little weasel possessed by a demented hatred for another person.

 

[Guest]

Jocab's compiler is free for non-commercial use

Well, let's see.  A bug is an undocumented featrure.

no. If a leg is a tail how many legs does a dog have?

 On that level
the very presence of the the option you are using is a bug.

since I don't accept your premise this doesn't follow

 

[James Kuyper]

no. If a leg is a tail how many legs does a dog have?

A bug is a failure of a program to do what it's supposed to do. As a
matter of deliberate choice, Jacob's compiler is NOT supposed to fully
conform to the C90 standard. So how does non-conformance count as a bug?
It's just non-conformance.

 

[Richard Bos]

No, the supplier won't change it partly _because_ he won't pay for
product Y. Regardless of whether jacob is right not to (IMO, as long as
he doesn't document conformance to C90, he is), this _is_ a vicious
circle. Either party could break the circle; neither party will.

my favourite version of this:

The inhabitants of a small town requested the railway company to
add their village station to ones stopped at by the morning train
that passed through on the way the London.

The railway company replied that there was no demand for such
a service and they knew this because *no one was waiting for
the morning train at that station*

Oh, Doctor Beeching, what have you done...

Richard

 

[James Kuyper]

my favourite version of this:

The inhabitants of a small town requested the railway company to
add their village station to ones stopped at by the morning train
that passed through on the way the London.

The railway company replied that there was no demand for such
a service and they knew this because *no one was waiting for
the morning train at that station*

My company used to use a code-checking tool named PR:QA, which
determines whether or not a given piece of code meets a highly
configurable set of coding standards. We accept code into our production
system, a lot of which was written by people who are much better at
science than at computer programming. We impose coding standards that
are pretty strict to protect our production system from being damaged by
that code. Having an automated tool for checking conformance to our
standards was pretty useful, and PR:QA was configurable enough to handle
most of our requirements.

One day, the SAs did an OS upgrade, and PR:QA stopped working.
Naturally, people complained, but nothing was ever done about it. A year
or two later, some bureaucrat noticed that we were making no use of our
license for PR:QA (because we couldn't use it), and decided that there
was no point in continuing to pay for that license.

 

[teapot]

Someone claiming to be "Tea Pot" wrote:

[poor forgery attempt snipped]

You're not doing it right! No member of the Pot family would
ever bring up the same issue more than once. No member of the
Pot family would ever make demands.

Even though I don't use lcc-win32 anymore, I'll run my
test suit one last time on the latest version and demonstrate
how to do it right.

 

[Kaz Kylheku]

Someone claiming to be "Tea Pot" wrote:

[poor forgery attempt snipped]

You're not doing it right! No member of the Pot family would
ever bring up the same issue more than once. No member of the
Pot family would ever make demands.

Certainly not papa Pol.

 

[Tea Pot]

Someone claiming to be "Tea Pot" wrote:

[poor forgery attempt snipped]

You're not doing it right! No member of the Pot family would
ever bring up the same issue more than once. No member of the
Pot family would ever make demands.

I hope it was clear to everyone that I am Tea Pot and you are teapot, and
we are not the same person! I apologize if there was any confusion.

 

[Tea Pot]

Tea Pot wrote:
...

No, he has pride in producing a non-conforming compiler, because he
considers the way in which it fails to conform to constitute an
improvement over the behavior of a fully-conforming compiler. It
follows quite naturally that he doesn't consider the non-conformance
to be a serious problem. He minds very much what he produces, and does
not want to produce what he would consider a "load of crap" , which is
precisely how he would feel about producing a fully-conforming
implementation of C90 (though I have no idea whether he'd use that
particular phrase to describe his attitude).

His target customer base doesn't care any more than he does about this
non-conformance. So why does it matter to you?


That attitude clearly marks you as not a member of his target customer
base. So don't use it. You've got lots of alternatives.

The only argument I've every had with him about this issue is about
whether or not he's actually claiming conformance that doesn't exist,
and whether his non-conforming implementation is topical on this
newsgroup. He has every right to produce a non-conforming
implementation if that's what he wants to do.

He has claimed conformance to both C90 and C99 in this newsgroup.

Everyone knows that the C99 claim is bullshit.

But many people don't realize that the claim of C90 conformance is also a
lie.

(By the way, the -ansi89 switch has been documented, described and
discussed at length in this group, so I don't accept that it's
"undocumented".)

To show my good faith, I will volunteer to pay Mr Navia up to 75 GBP of my
own money to make the adjustments needed to make his compiler 100%
C90-compliant. So now he can't use money as an excuse for his contemptuous
attitude to International Standards.

 

[James Kuyper]

He has claimed conformance to both C90 and C99 in this newsgroup.

And also backed away from that claim. He's admitted that it does not
produce the diagnostics required by C90, and has admitted that they are
in fact required.

(By the way, the -ansi89 switch has been documented, described and
discussed at length in this group, so I don't accept that it's
"undocumented".)

This newsgroup's discussions are not part of the documentation of his
compiler (unless that documentation contains a cross-reference to this
newsgroup; I've never seen that documentation, but I rather seriously
doubt that it contains such a cross-reference).

Even if this newsgroup's discussions of that option could be considered
as part of that documentation, those discussions made it abundantly
clear that this option does NOT provide the diagnostics mandated by C90,
and that it was never intended to provide such diagnostics. Therefore,
if they were to be counted as documentation, the option is in fact
behaving exactly as documented, and is therefore not a bug.

To show my good faith, I will volunteer to pay Mr Navia up to 75 GBP of my
own money to make the adjustments needed to make his compiler 100%
C90-compliant. So now he can't use money as an excuse for his contemptuous
attitude to International Standards.

The price you need to pay him is a matter for negotiation between you
and him - you don't get to set the price. From what I've seen of your
approach to this matter, even a person much less prone to taking
personal offense than Jacob is would be unlikely to negotiate a price
that low.

I don't know anything about the design of his compiler, and only he can
tell you how much work would be involved, so I can't be sure whether 75
GBP even comes close to being a reasonable price. However, at current
exchange rates, 75 GBP wouldn't buy very much of my time; I'd be
surprised if the amount of his time that it would buy would be
sufficient to complete this task.

 

[Keith Thompson]

He has claimed conformance to both C90 and C99 in this newsgroup.

Please cite a messsage in which jacob has claimed conformance to C90.
I don't believe he's ever done so.

[...]

(By the way, the -ansi89 switch has been documented, described and
discussed at length in this group, so I don't accept that it's
"undocumented".)

That's not what the word "documented" means. It's been *mentioned*
here. A newsgroup posting is not compiler documentation.

 

[Tea Pot]

I certainly do. I don't particularly care about whether a specific
compiler conforms to a standard *to which it does not claim
conformance*.

lcc-win32 attempts to conform to C99. (Last I heard, it was still
missing a few features, but that's beside the point.) The "-ansi89"
switch, as you have been told repeatedly, is not even documented; why
should anyone care what it does?

It might be nice if lcc-win32 had a mode in which it fully conforms to
the older C90 standard, but jacob is under no moral, ethical, or
contractual obligation to implement such a mode.

Right, and people are under no moral, ethical, or contractual obligation
to regard his compiler (a compiler for a language that isn't C!) as an
option for their use.

If a compiler vendor is not prepared to fix his product to conform on this
very simple point, how can anyone expect that vendor to be willing to
ensure his product conforms to the tricky, intricate points of the
Standard?

Refusing to fix a simple BUG gives the impression of a slipshod vendor who
is happy to ship crap. There isn't even any need to reject // comments:
all that is needed for Standards compliance is to emit a diagnostic. But
even that trivial change is too much for this arrogant vendor with no
interest in Standards compliance.

 

[James Kuyper]

Right, and people are under no moral, ethical, or contractual obligation
to regard his compiler (a compiler for a language that isn't C!) as an
option for their use.

Right. You've got it. Jacob makes certain promises about what his
compiler does. People buy (or download for free, if they qualify) his
compiler in the reasonable expectation that those promises are met, and
those people would have a legitimate complaint if those promises are not
met. People who want something he doesn't promise about his compiler
(such as full C90 conformance), shouldn't use it, and have no legitimate
grounds for complaint if it doesn't provide something he didn't promise
it would provide.

Just because you want vegetables doesn't obligate a butcher shop to
provide them. Why do you feel that lcc-win is obligated to provide full
C90 conformance?

If a compiler vendor is not prepared to fix his product to conform on this

Why would you consider this a "fix"? His compiler isn't intended to
fully conform to C90, so why should it be considered a defect that it
doesn't? Would you consider it a defect if his compiler doesn't conform
to the Fortran standard? I'm pretty sure it doesn't, at least not when
invoked in the mode you're complaining about.