32 lines
1.1 KiB
Plaintext
32 lines
1.1 KiB
Plaintext
=== on 10 Dec 2015, 17:32:09 Ann Campbell wrote:
|
|
Targeting to {cpp} because according to [~evgeny.mandrikov],
|
|
|
|
____
|
|
from language specification - there is no limit
|
|
|
|
templates are turing complete
|
|
|
|
so in theory - this is valid
|
|
|
|
____
|
|
|
|
=== on 11 Dec 2015, 09:14:58 Tamas Vajk wrote:
|
|
\[~ann.campbell.2] I changed the description a bit
|
|
|
|
=== on 11 Dec 2015, 13:58:31 Ann Campbell wrote:
|
|
looks good [~tamas.vajk]
|
|
|
|
=== on 27 Sep 2018, 20:11:18 Loïc Joly wrote:
|
|
\[~ann.campbell.2] -I would remove {cpp}, because such inheritance is indeed totally valid, and is even a commonly used idiom see for instance the \https://en.wikipedia.org/wiki/Curiously_recurring_template_pattern link-
|
|
|
|
|
|
After better understanding of this rule, I would still remove it for {cpp}, but for other reasons:
|
|
|
|
* There could be places where that works, because template specialization could break the recursion
|
|
* In places where it does not work, since templates are fully resolved during compilation, the code would not compile, so no need to report anything during analysis.
|
|
|
|
|
|
=== on 27 Sep 2018, 20:48:07 Ann Campbell wrote:
|
|
Feel free, [~loic.joly] :-)
|
|
|