Re: RQuantLib_0.1.0 in incoming

[Dirk Eddelbuettel]

Ok, I confirmed that on my 'Debain testing' laptop: The -pedantic switch
breaks building of RQuantLib, using the Jan 20 CVS snapshot of QL.

Kurt et al: Is that no-no which will prevent it from CRAN inclusion?

QLers: g++-2.95 breaks with -pedantic, g++-3.0 breaks outright (with
-mieee-fp -fPIC).  Ideas?   Those C++ classes make my head spin.

Dirk

--
Good judgement comes from experience; experience comes from bad judgement.
                                                            -- Fred Brooks

[Luigi Ballabio-4]

At 10:47 AM 2/26/02 -0600, Dirk Eddelbuettel wrote:
>Ok, I confirmed that on my 'Debain testing' laptop: The -pedantic switch
>breaks building of RQuantLib, using the Jan 20 CVS snapshot of QL.

Dirk,
         I didn't try it, but from your previous report, I see that
compilation fails on the instruction

std::vector<Iterator>::iterator it = iteratorVector_.begin();

which is replicated in a few methods.
Does it get better if you replace every occurrence with

typename std::vector<Iterator>::iterator it = iteratorVector_.begin();

?

Bye,
         Luigi

[Dirk Eddelbuettel]

On Tue, Feb 26, 2002 at 06:24:57PM +0000, Luigi Ballabio wrote:

> At 10:47 AM 2/26/02 -0600, Dirk Eddelbuettel wrote:
> >Ok, I confirmed that on my 'Debain testing' laptop: The -pedantic switch
> >breaks building of RQuantLib, using the Jan 20 CVS snapshot of QL.
>
> Dirk,
>         I didn't try it, but from your previous report, I see that
> compilation fails on the instruction
>
> std::vector<Iterator>::iterator it = iteratorVector_.begin();
>
> which is replicated in a few methods.
> Does it get better if you replace every occurrence with
>
> typename std::vector<Iterator>::iterator it = iteratorVector_.begin();
>
> ?

Yes!  Good stuff! I am always afraid of people who actually understand their
own code :)

It now works:
        -Wall -pedantic "clean", no warnings
        -Wall -pendantic -02 some warnings on uninit. enum optionType
        -Wall -pendantic -06 same,some warnings on uninit. enum optionType

Nando: Time for a new point release?

Thanks, Dirk

--
Good judgement comes from experience; experience comes from bad judgement.
                                                            -- Fred Brooks

[Luigi Ballabio-4]

At 11:18 AM 2/26/02 -0600, Dirk Eddelbuettel wrote:
>On Tue, Feb 26, 2002 at 06:24:57PM +0000, Luigi Ballabio wrote:
> > Does it get better if you replace every occurrence with
> >
> > typename std::vector<Iterator>::iterator it = iteratorVector_.begin();
> >
> > ?
>
>Yes!  Good stuff! I am always afraid of people who actually understand their
>own code :)

Oh, but I don't understand. I just prophesy :)

I've checked the fix into CVS.

Bye,
         Luigi

[Ferdinando M. Ametrano-2]

At 11:18 AM 2/26/02 -0600, Dirk Eddelbuettel wrote:
>Nando: Time for a new point release?
OK

>         -Wall -pendantic -06    same,some warnings on uninit. enum optionType
are these worth checking? Should we have -pedantic -06 as default?

ciao -- Nando

[Ferdinando M. Ametrano-2]

Dirk wrote:
>Nando: Time for a new point release?
Is a new 0.3.0a6 branch enough, or do you need a 0.2.2 release?

I would prefer to wait for 0.3.0 final, if possible.
Sad, I don't want to put any kind of pressure on you, but I'm only waiting
for the documentation of your work to finalize 0.3.0
Any news from you?

ciao -- Nando

[Dirk Eddelbuettel]

Ciao Nando

On Fri, Mar 01, 2002 at 04:42:41PM +0100, Ferdinando Ametrano wrote:
> Dirk wrote:
> >Nando: Time for a new point release?
> Is a new 0.3.0a6 branch enough, or do you need a 0.2.2 release?

Doesn't matter to me. We can for now stick with your numbering scheme at
your end (0.3.0a6) which I transform into mine at my end (due to the sorting
requirement you're now fully aware of). I could stick with 0.2.1cvs20020301
or call it 0.2.2.  I just have to make sure it is higher (in dpkg sorting
terms) than the existing one.

> I would prefer to wait for 0.3.0 final, if possible.

Debian might realease soon. 0.3.0 might miss that, so a bug fix release
could help there.

We can still release 0.3.0 after that.

Dirk

--
Good judgement comes from experience; experience comes from bad judgement.
                                                            -- Fred Brooks