On 7/11/2016 7:23 PM, deadalnix wrote:
On Tuesday, 12 July 2016 at 01:28:31 UTC, Walter Bright wrote:
I don't see anything actionable in your comment.
Defining by which way @safe actually ensure safety would be a good start.
I'm sorry for the frustration, but the "mention a problem, get asked for
an example, provide example, example is debated to death while problem
is ignored" cycle have become the typical interraction pattern around
here and that is VERY frustrating.
The example you gave of .ptr resulting in unsafe code has been in
bugzilla since 2013, and has an open PR on it to fix it.
https://issues.dlang.org/show_bug.cgi?id=11176
You didn't submit it to bugzilla - if you don't post problems to
bugzilla, most likely they will get overlooked, and you will get
frustrated. @safe issues are tagged with the 'safe' keyword in bugzilla.
If you know of other bugs with @safe, and they aren't in the list,
please add them. Saying generically that @safe has holes in it is
useless information since it is not actionable and nobody keeps track of
bugs posted on the n.g. nor are they even findable if you suspect
they're there.
----
If I may rant a bit, lots of posters here posit that with "more
process", everything will go better. Meanwhile, we DO have process for
bug reports. They go to bugzilla. Posting bugs to the n.g. does not
work. More process doesn't work if people are unwilling to adhere to it.