"Charlie Patterson" <charliep1@excite.com> wrote in message news:cuddto$2o74$1@digitaldaemon.com...
> "Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cubfs7$t5n$1@digitaldaemon.com...
>> >I think you and I have very similar opinions on this matter.
>>>
>>> I think most all of us here agree on what the best outcome is, what we seem to disagree over is what the compiler can best do to achieve it.
>>
>> Absolutely. That's the entire problem. Walter thinks that if the compiler tells the user there's a problem, the most likely outcome is a shut-up because programmers are unprofessional. This hamstrings all diligent engineers. Pessimism vs Optimism/Responsibility. As has been observed, there's no resolution of this difference, so we need to find a compromise.
>
> I don't see the problem as that.
>
> <preach>
> I don't think it is fair to call one side pessimistic.  That's pretty
> rhetorical.

Agreed. As I've just posted on the 'const/readonly string' thread, I think this issue maybe needs to go back to a simple distillation of the problem.

> And if this is a discussion between pessimists and optimists, then I'm not interested because both camps are typically full of non-thinkers. I think the appropriate position would be realist here. Sorry. </preach>
>
> <ot>
> Let me weigh in that most programmers *are* unprofessional.  (-:  You
> don't have to dig through too many books on software engineering to
> find that out.

Yes, alas, I guess I maybe have to admit that a significant proportion of them are. Lord knows - btw, is there an aetheistic/agnostic equivalent to "Lord knows"?? - I've worked with enough of them in years gone by. I suppose my thinking's been coloured these last few years because I've spent most of my time writing, wherein one learns all the myriad ways in which one's assumptions / implementations are wrong/bad, and any that are missed are gleefully pointed out by reviewers. The times I have worked have been consultative projects where I've done most/all the implementation myself. I confess that some years past I've worked with people who should be selling fake jewellery on street corners, rather than working on complex systems.

But if we agree that there are many unprofessional programmers, must we not also agree that there will be a continuum, rather than just, say 10% programmers are highly professional and 90% are utterly unprofessional. Given that, I think a language which actually lays traps for people who are somewhere in between - and I am convinced that D does indeed do that - is a bad thing. Someone who's half-arsed may well put in "return 0;"s in every function by rote, so as to avoid the dreaded "indeterminate return" that their (somewhat more professional) colleague has warned them about (or their team-leader has castigated them about, after they've caused run-time f-ups for the third time).

To me the only sane solution is a middle ground. But this clashes with Walter's strong intent to avoid warnings (and for good reason).

My position is that something's got to give.

> There is a factor of 30 between the productivity of the bad and good coders, for example.

Indeed, but try getting a 3000% pay rise on the strength of that. (btw, IIRC, it's only 29x <g>.)

>  And most people in any "sweatshop" environment, of which there are
> plenty in programming, do the minimum work they can.  But I don't
> think that matters.
> </ot>
>
> I see the problem as a matter of elegance and consistency.  And I think it is more elegant to provide the assert as a run-time check. Doesn't D also have array bounds checks?

Not in release, AFAIK.

<snip reason = "ran out of time; mentally foggy">

"Derek" <derek@psych.ward> wrote in message news:r5usmcalgzby.16ym9b7gewm7p.dlg@40tude.net...
> On Wed, 9 Feb 2005 11:31:46 -0500, Charlie Patterson wrote:
>
> [snip]
>
>>
>> I don't see the problem as that.
>
> [snip]
>
>> I see the problem as a matter of elegance and consistency.  And I
>> think it
>> is more elegant to provide the assert as a run-time check.
>
> I have no problem with this as well. The issue for me has boiled down
> to
> whether or not the compiler tells me that's what its done.
>
> I want to be told whenever the compiler does this sort of thing on my behalf.

Agreed.

> And I'm happy to have to ask for this too, for example via the -v compiler switch.

I think this is a mistake, but if this is as far as Walter can be pushed on this issue, then this'd an invocation of matthew->ShutUp("You get it as a flag on the compiler") would probably not throw an exception at this stage, (though it would result in the printing of a critical missive to stdwhine).

Matthew wrote:
> 
> 
> Nah! It's not arrived yet. It'd be thing 5 kilo old Dell sitting on the desk, with its miserable little broken hinge.
> 
> 
> 

Does that mean you don't want it? I'll take it. Seriously.

_______________________
Carlos Santander Bernal

Matthew wrote:
> of them are. Lord knows - btw, is there an aetheistic/agnostic equivalent to "Lord knows"?? - I've worked with enough of them in years 

No offense intended, but it reminds me of what a teacher of mine said once: "Atheist are weird, especially when they say: 'I swear by God that I'm atheist'" :D

_______________________
Carlos Santander Bernal

"Carlos Santander B." <csantander619@gmail.com> wrote in message news:cuefto$o4k$1@digitaldaemon.com...
> Matthew wrote:
>>
>>
>> Nah! It's not arrived yet. It'd be thing 5 kilo old Dell sitting on the desk, with its miserable little broken hinge.
>>
>>
>>
>
> Does that mean you don't want it? I'll take it. Seriously.

He he. No, sorry. I've ordered a hinge from Dell - a company that seems to know how to provide at least acceptable, if not great, customer service - and it's about to become a Linux machine. (GDC, here I come!)

"Carlos Santander B." <csantander619@gmail.com> wrote in message news:cuegnn$opn$1@digitaldaemon.com...
> Matthew wrote:
>> of them are. Lord knows - btw, is there an aetheistic/agnostic equivalent to "Lord knows"?? - I've worked with enough of them in years
>
> No offense intended, but it reminds me of what a teacher of mine said once: "Atheist are weird, especially when they say: 'I swear by God that I'm atheist'" :D

No offence to me, I'm not an atheist. (Certainly on either side indicates, to me, a decidedly unnerving degree of certitude.)

 I just wanted to know if anyone could suggest an alternative to "Lord
knows, " or "Heaven knows, ", which I find myself saying more often than
I'd like.

:-)

On Thu, 10 Feb 2005 13:49:51 +1100, Matthew <admin@stlsoft.dot.dot.dot.dot.org> wrote:

>
> "Carlos Santander B." <csantander619@gmail.com> wrote in message
> news:cuegnn$opn$1@digitaldaemon.com...
>> Matthew wrote:
>>> of them are. Lord knows - btw, is there an aetheistic/agnostic
>>> equivalent to "Lord knows"?? - I've worked with enough of them in
>>> years
>>
>> No offense intended, but it reminds me of what a teacher of mine said
>> once: "Atheist are weird, especially when they say: 'I swear by God
>> that I'm atheist'" :D
>
> No offence to me, I'm not an atheist. (Certainly on either side
> indicates, to me, a decidedly unnerving degree of certitude.)
>
>  I just wanted to know if anyone could suggest an alternative to "Lord
> knows, " or "Heaven knows, ", which I find myself saying more often than
> I'd like.
>
> :-)
>
>
>

I tend to use 'Bob' as a generic drop-in replacement for deity invocation - "Bob only knows"

warning: Do not use around people called Bob.

-- 
Using Opera's revolutionary e-mail client: http://www.opera.com/m2/

"Alex Stevenson" <ans104@cs.york.ac.uk> wrote in message news:opslywebqz08qma6@mjolnir.spamnet.local...
> On Thu, 10 Feb 2005 13:49:51 +1100, Matthew <admin@stlsoft.dot.dot.dot.dot.org> wrote:
>
>>
>> "Carlos Santander B." <csantander619@gmail.com> wrote in message news:cuegnn$opn$1@digitaldaemon.com...
>>> Matthew wrote:
>>>> of them are. Lord knows - btw, is there an aetheistic/agnostic equivalent to "Lord knows"?? - I've worked with enough of them in years
>>>
>>> No offense intended, but it reminds me of what a teacher of mine
>>> said
>>> once: "Atheist are weird, especially when they say: 'I swear by God
>>> that I'm atheist'" :D
>>
>> No offence to me, I'm not an atheist. (Certainly on either side indicates, to me, a decidedly unnerving degree of certitude.)
>>
>>  I just wanted to know if anyone could suggest an alternative to
>> "Lord
>> knows, " or "Heaven knows, ", which I find myself saying more often
>> than
>> I'd like.
>>
>> :-)
>>
>>
>>
>
> I tend to use 'Bob' as a generic drop-in replacement for deity invocation  - "Bob only knows"
>
> warning: Do not use around people called Bob.

That's great! Thanks

Bob only knows why I didn't think of it before. :-)

In article <cuetqd$14g2$1@digitaldaemon.com>, Matthew says...
>
>
>"Alex Stevenson" <ans104@cs.york.ac.uk> wrote in message news:opslywebqz08qma6@mjolnir.spamnet.local...
>> On Thu, 10 Feb 2005 13:49:51 +1100, Matthew <admin@stlsoft.dot.dot.dot.dot.org> wrote:
>>
>>>
>>> "Carlos Santander B." <csantander619@gmail.com> wrote in message news:cuegnn$opn$1@digitaldaemon.com...
>>>> Matthew wrote:
>>>>> of them are. Lord knows - btw, is there an aetheistic/agnostic equivalent to "Lord knows"?? - I've worked with enough of them in years
>>>>
>>>> No offense intended, but it reminds me of what a teacher of mine
>>>> said
>>>> once: "Atheist are weird, especially when they say: 'I swear by God
>>>> that I'm atheist'" :D
>>>
>>> No offence to me, I'm not an atheist. (Certainly on either side indicates, to me, a decidedly unnerving degree of certitude.)
>>>
>>>  I just wanted to know if anyone could suggest an alternative to
>>> "Lord
>>> knows, " or "Heaven knows, ", which I find myself saying more often
>>> than
>>> I'd like.
>>>
>>> :-)
>>>
>>>
>>>
>>
>> I tend to use 'Bob' as a generic drop-in replacement for deity invocation  - "Bob only knows"
>>
>> warning: Do not use around people called Bob.
>
>That's great! Thanks
>
>Bob only knows why I didn't think of it before. :-)
>
>

What about Void?  Isn't that supposed to cover everything?

Void only knows!

or

cast(void)(Diety) only knows!

Sorry, couldn't help myself!

"Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cueank$jur$1@digitaldaemon.com...
>> I see the problem as a matter of elegance and consistency.  And I think it is more elegant to provide the assert as a run-time check. Doesn't D also have array bounds checks?
>
> Not in release, AFAIK.
>
> <snip reason = "ran out of time; mentally foggy">

When the fog lifts... (-:

So why is it OK to remove array-bounds checks at release but not functional return points?  They seem remarkably similar to me.  The assumption seems to be that if you didn't catch them in debug, you'll be alright.

And why should the user be forced to insert dummy return point (or assertions at return points), but not dummy array-bounds checks or assertions?

I'm also OK with Derek that a compile option such as --sanity would point out the automatically inserted assertions, but how big might this list be if it includes, again, array-bounds checks, etc?  For the record, I hate it when compilers dump warnings about things that aren't a problem.  I guess I'm anal like that, but it frustrates me for the compiler to point out non-errors.  It's like being micro-managed.  Like I'm painting and someone is standing over my shoulder saying, "Hey you missed a spot!  Did you mean to leave it uneven?  Are you going to put something there?"  Plus, I really hate inheriting code that throws warnings.  I dont' know the code base yet so how worried should I be?