Comments on: Gem #31: Preconditions/postconditions http://www.adacore.com/2008/04/14/gem-31/ AdaCore technology and news Mon, 06 Feb 2012 18:59:19 -0500 http://wordpress.org/?v=2.9 hourly 1 By: Chris Schwering http://www.adacore.com/2008/04/14/gem-31/comment-page-1/#comment-10390 Chris Schwering Fri, 04 Jun 2010 12:11:09 +0000 http://www2.adacore.com/2008/04/14/gem-31/#comment-10390 I've just replaced many assertions with pre-/postconditions in subprogram bodies; it's very handy, especially postconditions. Wouldn't it be nice if we could place use-type-clauses before the pre-/postconditions to make operators usable in the conditions? I’ve just replaced many assertions with pre-/postconditions in subprogram bodies; it’s very handy, especially postconditions.
Wouldn’t it be nice if we could place use-type-clauses before the pre-/postconditions to make operators usable in the conditions?

]]> By: Cyrille Comar http://www.adacore.com/2008/04/14/gem-31/comment-page-1/#comment-1380 Cyrille Comar Wed, 16 Apr 2008 17:17:36 +0000 http://www2.adacore.com/2008/04/14/gem-31/#comment-1380 Thanks for your nice comment Jeff. The issue of static vs dynamic pre/post conditions is indeed very interesting. Static analysis tools exploiting the information provided by the dynamic versions are already on their way... More on this later ;-) The static versions have been provided by tools such as Spark from Praxis for a while already. Finding a way to marry the static & the dynamic versions is an interesting challenge that we are considering seriously! Thanks for your nice comment Jeff. The issue of static vs dynamic pre/post conditions is indeed very interesting. Static analysis tools exploiting the information provided by the dynamic versions are already on their way… More on this later ;-)
The static versions have been provided by tools such as Spark from Praxis for a while already.
Finding a way to marry the static & the dynamic versions is an interesting challenge that we are considering seriously!

]]> By: Jeff Creem http://www.adacore.com/2008/04/14/gem-31/comment-page-1/#comment-1379 Jeff Creem Wed, 16 Apr 2008 01:32:32 +0000 http://www2.adacore.com/2008/04/14/gem-31/#comment-1379 Nice and interesting timing. I was just re-reading some of the old AIs in this area and thinking that I wish this had been implemented in Ada 2005. One concern that had been expressed with the AIs was that in Ada we typically want static compile time checking and this only was seen as a dynamic check but hopefully tools like Polyspace and other static tools can begin to use this information. Even short of that, I like the idea of being able to use this to help document expectations. Nice and interesting timing. I was just re-reading some of the old AIs in this area and thinking that I wish this had been implemented in Ada 2005.

One concern that had been expressed with the AIs was that in Ada we typically want static compile time checking and this only was seen as a dynamic check but hopefully tools like Polyspace and other static tools can begin to use this information. Even short of that, I like the idea of being able to use this to help document expectations.

]]>