Re: [lmi] [lmi-commits] master 8cccd20 5/9: Note a mystery

[Greg Chicares]

On 2018-09-23 13:38, Vadim Zeitlin wrote:
> On Sat, 22 Sep 2018 23:21:47 +0000 Greg Chicares <address@hidden> wrote:
> 
> GC> On 2018-09-19 15:30, Vadim Zeitlin wrote:
> GC> > On Wed, 19 Sep 2018 09:19:33 -0400 (EDT) Greg Chicares <address@hidden> 
> wrote:
> GC> > 
> GC> > GC>     Note a mystery
[...]
>  Sorry for all the confusion which probably cost you a lot of time
Not really--it cost only seconds to commit an inline comment, as this
was the only thing that I just couldn't figure out the first time I
went through all the code in that large TU. Finally figuring it out
after your last reply did take time, in exactly the same way that
studying a complicated example in a textbook takes time; but that's
time well spent.

The mystery that actually took far more time than I'd care to admit
is the one that I decided to resolve myself instead of just noting
in an online comment--viz., that a wx_table_generator cannot be
created in a ctor-initializer in the hierarchy rooted in class
logical_page, because that results in calling a pure function. The
virtual is of course overridden in the hierarchy rooted in class
pdf_illustration (not logical_page), but I was convinced it was
otherwise. IOW, I knew what the cause of the problem had to be,
which is a basic C++ FAQ, but I was sure the FAQ didn't apply here,
because the pdf_illustration object had been fully constructed by
the time its individual pages are instantiated; but chronological
precedence turns out not always to imply an inheritance relationship,
or everything would derive from main(). That's the problem I should
have asked you about instead, because it was simple snowblindness
and could be dispelled quite easily. But all's well that ends well.