Re: Matt's Maturity
ken mortenson wrote:
>
> Matt Lewis wrote:
> >
> > Unless we're looking at different things, here was how you got a value for
> > rid_dbOpen:
> > }}}
<eucode>
> > atom rid_dbOpen
> > rid_dbOpen = define_c_func(dbSQL, "dbOpen", {C_POINTER}, C_LONG)
> > </eucode>
{{{
> >
> > The 2 was returned by define_c_func, which has no knowledge of a vector
> > table in the PowerBASIC DLL. Your intuition was coincidence.
>
> dbSQL is a PowerBasic DLL. "dbOpen" is the second function defined
> therein. Are you ignorant of how define_c_func works?
>
> My intuition was spot on. Thank you for reinforcing that point.
No, it turns out I'm not ignorant of how define_c_func works. I'll walk
you through what you're seeing:
...from source/execute.h:
struct arg_info {
int (*address)(); // pointer to C function
s1_ptr name; // name of routine (for diagnostics)
s1_ptr arg_size; // s1_ptr of sequence of argument sizes
object return_size; // atom or sequence for return value size
int convention; // calling convention
};
...from source/be_machine.c:
int c_routine_next = 0; /* index of next available element */
struct arg_info *c_routine = NULL; /* array of c_routine structs */
Here is a bit of the back-end implementation of define_c_func (just the
start and the end):
object DefineC(object x)
/* define a C routine: x is {lib, name, arg_sizes, return_type or 0}
alternatively, x is {"", address or {'+', address}, arg_sizes, return_type or
0}
Return -1 on failure. */
{
...
proc_address = (int (*)())GetProcAddress((void *)lib, routine_string);
...
c_routine[c_routine_next].address = proc_address;
...
return c_routine_next++;
}
You can look up the rest of the code yourself, but as you can see, euphoria
uses GetProcAddress along with the name of the exported function. It stores
a pointer to that structure in an array, and returns the index for that
particular arg_info.
> > > Finally, it is cowardly of you to claim me ignorant in any respect without
> > > providing evidence as I've done in the passage above so that I am at a
> > > severe handicap to defend myself.
> >
> > Why is it cowardly?
>
> Because in your original post you made the accusation without support.
> Only in subsequent post did you provide evidence, but not so I could
> defend myself but so you could prove how right you were. As I am
> demonstrating here I am able to defend myself when provided with
> specific allegations rather than indefensible slurs.
>
> To accuse without giving the accused the means of defending themselves
> is a cowardly act. A mature person knows that.
Sorry, I wasn't really accusing you of anything. It's true that I didn't
provide references in the original. I reread my original post:
http://www.openeuphoria.org/EUforum/m21806.html
...and I still think it was a pretty neutral statement of my observations.
> > In the case above, with define_c_func, you proved your
> > ignorance of how that feature of the language works.
>
> As I've already shown above, you are still continuing to show your
> ignorance about the source of the number while my intution has proved
> to be entirely correct. This next part is beautiful...
>
> > According to the docs,
> > which I believe someone responded to your original post with a link:
> >
> > "A small integer, i1, known as a routine id, will be returned."
>
> You read, but without understanding. What is the source of this small
> integer? Magic? No, it's seems it's from the vector table of the DLL.
I don't follow this jump to the vector table in the dll, or what it is.
Are you referring to the export by ordinal, perhaps? Even so, as I
showed above, you were incorrect.
> > It's interesting [to me] that you construed all of this as a personal
> > attack, especially after making such a big deal about your lack of social
> > graces. To wit:
>
> It was a personal attack. Make no mistake. Calling me a troll is also
> a personal attack. I've had the decency of explaining to you why it
> was a cowardly attack. If you were a gentleman you would apologize
> for doing so (rather than continuing with the troll comments and digging
> yourself in deeper by continuing to try to prove what I have repeated
> demonstrated to be wrong.)
I'll grant the troll comment. But as Keith David said in _Men at Work_,
"He was provokin me!"
I don't have a problem admitting it when I'm wrong. I even admitted once
(and a half?) where I was wrong. But you continue to say false things
about define_c_func, displaying your ignorance. Who is digging?
> > > You can believe anything you like about me.
> > >
> > > I believe that you are no gentleman and a coward.
> >
> > And you are welcome to believe whatever you like. I'm done feeding the
> > trolls.
>
> The ball is now in your court. Your actions define who you are.
> Notice I'm not calling you names. Calling you a coward is a specific
> response to your actions and you can prove me wrong with an apology.
You're not calling me names in *this* post (though you still quoted the names
you called me previously). I'll admit that you're not really a troll
(though some of your posts about gotos, structured code, and
other general programming philosophies seem at least borderline), just that
you were correct when you claimed to be a poor communicator.
I often have a similar problem. I've often thought a lot about a particular
issue, and have developed a strong opinion. When the topic comes up, I'm
not hesitant to let others know. It comes across as a knee jerk reaction
and irritates people (esp my wife, who never fails to let me know!),
because my opinions seem unfounded or poorly thought out--even though the
opposite is true.
Something I read every so often, that is a huge help for fora like this:
http://www.catb.org/~esr/faqs/smart-questions.html
You may have seen it before.
Matt
|
Not Categorized, Please Help
|
|
