template gives long name

[gorgen]

My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
double > > > parList;
Name is truncated to 255 char.
Some suggestions for a solution? Thanks in advance.

[Saar]

You need not worry, it's a debug warning which can be disabled using:
#pragma warning ( disable : 4786) .


"gorgen" <gorgen.givik@era.ericsson.se> wrote:
>
>My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
>double > > > parList;
>Name is truncated to 255 char.
>Some suggestions for a solution? Thanks in advance.
>

[Danny Kalev]

there's no real solution to this except for disabling this warning, as
Saar suggested. This is a bug in the MS debugger; you can safely ignore
it.

Danny

gorgen wrote:
>
> My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
> double > > > parList;
> Name is truncated to 255 char.
> Some suggestions for a solution? Thanks in advance.

[Ralph D. Cole]

Danny Kalev <dannykk@inter.net.il> wrote:
>there's no real solution to this except for disabling this warning, as
>Saar suggested. This is a bug in the MS debugger; you can safely ignore
>it.
>
>Danny
>

Danny,

A 'bug'?

All debuggers have a limitation on SYMBOL size. However, not all debuggers
(including several you are very familar with) will warn you when the symbols
are truncated, since in most cases this is beniqn.

So where is the 'bug'?

Is it in informing you that something benign has happened? Or is it in
not warning you, so if you go looking for that sysmbol in the debug database
(assuming you are a real geek <g>) you will not know the NAME has been truncated?

IMHO: It can be a pain in the <blank>, but I am not sure I would go so
far as to call an over-zealousness to report information a 'bug'.

-ralph

>gorgen wrote:
>>
>> My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
>> double > > > parList;
>> Name is truncated to 255 char.
>> Some suggestions for a solution? Thanks in advance.

[Ralph]

Somehow I knew you would say that. No Miss Cleo needed here. ;-)

<g>
-ralph

Danny Kalev <dannykk@inter.net.il> wrote:
>
>
>"Ralph D. Cole" wrote:
>>
>> Danny Kalev <dannykk@inter.net.il> wrote:
>> >there's no real solution to this except for disabling this warning, as
>> >Saar suggested. This is a bug in the MS debugger; you can safely ignore
>> >it.
>> >
>> >Danny
>> >
>>
>> Danny,
>>
>> A 'bug'?
>>
>> All debuggers have a limitation on SYMBOL size. However, not all debuggers
>> (including several you are very familar with) will warn you when the symbols
>> are truncated, since in most cases this is beniqn.
>>
>> So where is the 'bug'?
>
>the bugs is here: a proper debugger should have a reasonable symbol size
>limit, at least one that can represent normal templates properly. A 2000
>character limit is reasonable, considering VC++ name mangling scheme. MS
>is aware of that but failed to increase its debugger's limit. I believe
>this was fixed in VS 7.0. So, yes, it's an implementation bug just a
>compiler that crashes when encountering valid code is a bug.
>>
>> Is it in informing you that something benign has happened? Or is it
in
>> not warning you, so if you go looking for that sysmbol in the debug database
>> (assuming you are a real geek <g>) you will not know the NAME has been
truncated?
>
>Yes, I'm a geek, and besides, if the symbol of shorter identifiers is
>accessible, the lack thereof in the case of complex template names is
>certainly a limitation of the implementation, or a bug. I wouldn't call
>it 'benign'.
>
>>
>> IMHO: It can be a pain in the <blank>, but I am not sure I would go
so
>> far as to call an over-zealousness to report information a 'bug'.
>
>It's a limitation, one that can certainly become a pain in the <blank>
>when you debug STL-based apps. Besides, why bother innocent programmers
>with this message which is not their fault anyway? Finally, a 255 char
>limit is simply outdated, just as a 640 kb RAM limit is.
>
>Danny
>>
>> -ralph
>>
>> >gorgen wrote:
>> >>
>> >> My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
>> >> double > > > parList;
>> >> Name is truncated to 255 char.
>> >> Some suggestions for a solution? Thanks in advance.

[Ralph]

Now, now. M$ admits to a bug in using "#pragma warning(diable:xxx)". i.e.,
even the #pragma statement will sometimes fail to shut the compiler up.




Danny Kalev <dannykk@inter.net.il> wrote:
>A more compelling evidence: if MS calls it a bug, then it is a bug:
>http://support.microsoft.com/support/kb/articles/Q167/3/55.ASP
>
>Danny
>
>"Ralph D. Cole" wrote:
>>
>> Danny Kalev <dannykk@inter.net.il> wrote:
>> >there's no real solution to this except for disabling this warning, as
>> >Saar suggested. This is a bug in the MS debugger; you can safely ignore
>> >it.
>> >
>> >Danny
>> >
>>
>> Danny,
>>
>> A 'bug'?
>>
>> All debuggers have a limitation on SYMBOL size. However, not all debuggers
>> (including several you are very familar with) will warn you when the symbols
>> are truncated, since in most cases this is beniqn.
>>
>> So where is the 'bug'?
>>
>> Is it in informing you that something benign has happened? Or is it
in
>> not warning you, so if you go looking for that sysmbol in the debug database
>> (assuming you are a real geek <g>) you will not know the NAME has been
truncated?
>>
>> IMHO: It can be a pain in the <blank>, but I am not sure I would go
so
>> far as to call an over-zealousness to report information a 'bug'.
>>
>> -ralph
>>
>> >gorgen wrote:
>> >>
>> >> My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
>> >> double > > > parList;
>> >> Name is truncated to 255 char.
>> >> Some suggestions for a solution? Thanks in advance.

[Danny Kalev]

"Ralph D. Cole" wrote:
>
> Danny Kalev <dannykk@inter.net.il> wrote:
> >there's no real solution to this except for disabling this warning, as
> >Saar suggested. This is a bug in the MS debugger; you can safely ignore
> >it.
> >
> >Danny
> >
>
> Danny,
>
> A 'bug'?
>
> All debuggers have a limitation on SYMBOL size. However, not all debuggers
> (including several you are very familar with) will warn you when the symbols
> are truncated, since in most cases this is beniqn.
>
> So where is the 'bug'?

the bugs is here: a proper debugger should have a reasonable symbol size
limit, at least one that can represent normal templates properly. A 2000
character limit is reasonable, considering VC++ name mangling scheme. MS
is aware of that but failed to increase its debugger's limit. I believe
this was fixed in VS 7.0. So, yes, it's an implementation bug just a
compiler that crashes when encountering valid code is a bug.
>
> Is it in informing you that something benign has happened? Or is it in
> not warning you, so if you go looking for that sysmbol in the debug database
> (assuming you are a real geek <g>) you will not know the NAME has been truncated?

Yes, I'm a geek, and besides, if the symbol of shorter identifiers is
accessible, the lack thereof in the case of complex template names is
certainly a limitation of the implementation, or a bug. I wouldn't call
it 'benign'.

>
> IMHO: It can be a pain in the <blank>, but I am not sure I would go so
> far as to call an over-zealousness to report information a 'bug'.

It's a limitation, one that can certainly become a pain in the <blank>
when you debug STL-based apps. Besides, why bother innocent programmers
with this message which is not their fault anyway? Finally, a 255 char
limit is simply outdated, just as a 640 kb RAM limit is.

Danny
>
> -ralph
>
> >gorgen wrote:
> >>
> >> My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
> >> double > > > parList;
> >> Name is truncated to 255 char.
> >> Some suggestions for a solution? Thanks in advance.

[Danny Kalev]

A more compelling evidence: if MS calls it a bug, then it is a bug:
http://support.microsoft.com/support/kb/articles/Q167/3/55.ASP

Danny

"Ralph D. Cole" wrote:
>
> Danny Kalev <dannykk@inter.net.il> wrote:
> >there's no real solution to this except for disabling this warning, as
> >Saar suggested. This is a bug in the MS debugger; you can safely ignore
> >it.
> >
> >Danny
> >
>
> Danny,
>
> A 'bug'?
>
> All debuggers have a limitation on SYMBOL size. However, not all debuggers
> (including several you are very familar with) will warn you when the symbols
> are truncated, since in most cases this is beniqn.
>
> So where is the 'bug'?
>
> Is it in informing you that something benign has happened? Or is it in
> not warning you, so if you go looking for that sysmbol in the debug database
> (assuming you are a real geek <g>) you will not know the NAME has been truncated?
>
> IMHO: It can be a pain in the <blank>, but I am not sure I would go so
> far as to call an over-zealousness to report information a 'bug'.
>
> -ralph
>
> >gorgen wrote:
> >>
> >> My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
> >> double > > > parList;
> >> Name is truncated to 255 char.
> >> Some suggestions for a solution? Thanks in advance.

[Danny Kalev]

Ralph wrote:
>
> Now, now. M$ admits to a bug in using "#pragma warning(diable:xxx)". i.e.,
> even the #pragma statement will sometimes fail to shut the compiler up.

Clearly, this is a sign that they should fix the debugger rather than
the preprocessor:)
OK, maybe a bug is not the most accurate term but it's a limitation, not
to say a flaw.
Danny
>
> Danny Kalev <dannykk@inter.net.il> wrote:
> >A more compelling evidence: if MS calls it a bug, then it is a bug:
> >http://support.microsoft.com/support/kb/articles/Q167/3/55.ASP
> >
> >Danny
> >
> >"Ralph D. Cole" wrote:
> >>
> >> Danny Kalev <dannykk@inter.net.il> wrote:
> >> >there's no real solution to this except for disabling this warning, as
> >> >Saar suggested. This is a bug in the MS debugger; you can safely ignore
> >> >it.
> >> >
> >> >Danny
> >> >
> >>
> >> Danny,
> >>
> >> A 'bug'?
> >>
> >> All debuggers have a limitation on SYMBOL size. However, not all debuggers
> >> (including several you are very familar with) will warn you when the symbols
> >> are truncated, since in most cases this is beniqn.
> >>
> >> So where is the 'bug'?
> >>
> >> Is it in informing you that something benign has happened? Or is it
> in
> >> not warning you, so if you go looking for that sysmbol in the debug database
> >> (assuming you are a real geek <g>) you will not know the NAME has been
> truncated?
> >>
> >> IMHO: It can be a pain in the <blank>, but I am not sure I would go
> so
> >> far as to call an over-zealousness to report information a 'bug'.
> >>
> >> -ralph
> >>
> >> >gorgen wrote:
> >> >>
> >> >> My compiler VC++ gives a varning for std::vector< std::vector < std::complex<
> >> >> double > > > parList;
> >> >> Name is truncated to 255 char.
> >> >> Some suggestions for a solution? Thanks in advance.

[Ralph]

Danny Kalev <dannykk@inter.net.il> wrote:
>
>
>Ralph wrote:
>>
>> Now, now. M$ admits to a bug in using "#pragma warning(diable:xxx)". i.e.,
>> even the #pragma statement will sometimes fail to shut the compiler up.
>
>Clearly, this is a sign that they should fix the debugger rather than
>the preprocessor:)
>OK, maybe a bug is not the most accurate term but it's a limitation, not
>to say a flaw.
>Danny

I agree heartily. Especially, as you were saying in another posting we need
a FAQ for things like this, since this question seems to appear quite regularly.


Quite regularly for how long now? Three or four years? At least since VC5.
That would make it about 8 or 9 service packs, right? It's not like they
haven't had the opportunity to repair it. Why should DevX waste disk space
and links for an FAQ to report something M$ has had literally years to fix.

There must be something more to it than just symbol length. Or rather I hope
there was more to it than just a change in the SYMBOL table. <g> It is this
seemingly careless attitude on their part over little things that does bother
me about M$.

-ralph


>>
>> Danny Kalev <dannykk@inter.net.il> wrote:
>> >A more compelling evidence: if MS calls it a bug, then it is a bug:
>> >http://support.microsoft.com/support/kb/articles/Q167/3/55.ASP
>> >
>> >Danny
>> >
>> >"Ralph D. Cole" wrote:
>> >>
>> >> Danny Kalev <dannykk@inter.net.il> wrote:
>> >> >there's no real solution to this except for disabling this warning,
as
>> >> >Saar suggested. This is a bug in the MS debugger; you can safely ignore
>> >> >it.
>> >> >
>> >> >Danny
>> >> >
>> >>
>> >> Danny,
>> >>
>> >> A 'bug'?
>> >>
>> >> All debuggers have a limitation on SYMBOL size. However, not all
debuggers
>> >> (including several you are very familar with) will warn you when the
symbols
>> >> are truncated, since in most cases this is beniqn.
>> >>
>> >> So where is the 'bug'?
>> >>
>> >> Is it in informing you that something benign has happened? Or is
it
>> in
>> >> not warning you, so if you go looking for that sysmbol in the debug
database
>> >> (assuming you are a real geek <g>) you will not know the NAME has been
>> truncated?
>> >>
>> >> IMHO: It can be a pain in the <blank>, but I am not sure I would
go
>> so
>> >> far as to call an over-zealousness to report information a 'bug'.
>> >>
>> >> -ralph
>> >>
>> >> >gorgen wrote:
>> >> >>
>> >> >> My compiler VC++ gives a varning for std::vector< std::vector <
std::complex<
>> >> >> double > > > parList;
>> >> >> Name is truncated to 255 char.
>> >> >> Some suggestions for a solution? Thanks in advance.