 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich <tho### [at] trf de> wrote:
> popov wrote:
> > So you suggest not to turn warnings on?
>
> When did you get that idea? You certainly know that there is more than just
> a warning on/off switch in every compiler out there. So how to conclude
> from saying that not all warnings should be on by default that all warnings
> should be off is a rather strange interpretation...
>
> > But wait, why do I tell this. You have so much experience...
>
> Good question!
>
> Thorsten
What I'm trying to say:
It is better to have some warnings to much, than miss a warning which may
kill you (or the guy which uses your software).
But it is not my intention to convince you.
How could I...
Post a reply to this message
|
|
| |
| |
|
|
From: Thorsten Froehlich
Subject: Re: 3.6.1 won't compile for me on OSX
Date: 3 Feb 2005 14:53:23
Message: <42028133@news.povray.org>
|
|
|
| |
| |
|
|
popov wrote:
> What I'm trying to say:
> It is better to have some warnings to much, than miss a warning
Exactly, and in the flood of bogus warnings you get with all wwarnings
enabled, you cannot see the useful ones.
Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich <tho### [at] trfde> wrote:
> If you have to rely on warnings to do your quality assurance you are in
> really serious trouble!!!
When did you get that idea? You certainly know that there are other QA's.
So how to conclude from saying that turning on all warnings is some kind of
QA, it IS THE QA process I'm using.
'..rely on warnings...' is a rather strange interpretation...
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> And in other cases the
> warning cannot be eliminating without code duplication, which you would have
> noticed when compiling reasonably complex template code and getting tons of
> unused variable or argument warnings...
You probably do not have to fix warnings if you know these are harmless.
However, it is worth to ONCE look at ALL possible warnings the compiler is
able to tell you. Therefore: turn them on.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich <tho### [at] trfde> wrote:
> popov wrote:
> > What I'm trying to say:
> > It is better to have some warnings to much, than miss a warning
>
> Exactly, and in the flood of bogus warnings you get with all wwarnings
> enabled, you cannot see the useful ones.
>
> Thorsten
This way you have a chance to see it. If you force the compiler not to print
the warning, you will miss it FOR SURE.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
popov wrote:
>>And in other cases the
>>warning cannot be eliminating without code duplication, which you would have
>>noticed when compiling reasonably complex template code and getting tons of
>>unused variable or argument warnings...
>
> You probably do not have to fix warnings if you know these are harmless.
> However, it is worth to ONCE look at ALL possible warnings the compiler is
> able to tell you. Therefore: turn them on.
I have a long list of warnings for various compilers I never care about
because they are just bogus as they are for inexperienced programmers. Or
they are portability information that is plain wrong when dealing with
conforming compilers rather than ancient junk compilers.
The best example is POV-Ray: If you were to "fix" the warnings you get with
all warnings enabled, you are going to seriously damage the code. This
actually happened a while back when someone had the idea to "fix all warnings".
Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
popov wrote:
> This way you have a chance to see it. If you force the compiler not to print
> the warning, you will miss it FOR SURE.
No, the only thing for sure is that I will miss important warnings. As
said, there are warnings that are not worthy a message at all, and those
just should not be shown, *ever*.
Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich <tho### [at] trfde> wrote:
> popov wrote:
> > This way you have a chance to see it. If you force the compiler not to print
> > the warning, you will miss it FOR SURE.
>
> No, the only thing for sure is that I will miss important warnings. As
> said, there are warnings that are not worthy a message at all, and those
> just should not be shown, *ever*.
>
> Thorsten
How can you miss the important warnings if you simply don't get them on your
screen. This could happen if you are not going to the highest warning
level.
I mean not routinely, however when compiling a final build or employing
another compiler.
Post a reply to this message
|
|
| |
| |
|
|
From: Thorsten Froehlich
Subject: Re: 3.6.1 won't compile for me on OSX
Date: 3 Feb 2005 19:15:46
Message: <4202beb2@news.povray.org>
|
|
|
| |
| |
|
|
popov wrote:
> How can you miss the important warnings if you simply don't get them on your
> screen. This could happen if you are not going to the highest warning
> level.
Because the important warnings are those that are *on* my default! The
useless ones are *off* by default and need to be explicitly requested by
asking for all warnings. Probably you should read up what enabling all
warnings actually does!
Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich <tho### [at] trfde> wrote:
> popov wrote:
> > How can you miss the important warnings if you simply don't get them on your
> > screen. This could happen if you are not going to the highest warning
> > level.
>
> Because the important warnings are those that are *on* my default! The
> useless ones are *off* by default and need to be explicitly requested by
> asking for all warnings. Probably you should read up what enabling all
> warnings actually does!
>
> Thorsten
Period
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
