Pitching D to academia

[Ola Fosheim Grøstad via Digitalmars-d]

On Wednesday, 9 March 2016 at 19:45:40 UTC, cym13 wrote:
> Let's add that D binaries are usually too bloated for their 
> disassembly to be as readable as their C equivalent (mangling 
> doesn't help) so even for "reverse engineering" assembly it is 
> less than perfect (although perfectly doable of course).

Yes, I think there are many good reasons for sticking to 
languages like C, Java or a narrow toy language. You usually want 
to use the language used in the best books for the course... And 
you care most about catering for the students that are having 
problems with the topic, i.e. make it easy to self study for weak 
students.

A new rich language can easily distract from learning objectives. 
I once tried to push some key XML technologies by having a 
mandatory pipeline that was like this: SQL -> XQuery -> XML -> 
XSLT -> HTML, which worked out ok, except it became too time 
consuming for the non-geek students. Which made them frustrated. 
It becomes challenging for students if they have to struggle with 
learning both the topic (design) and the tools at the same time.

So, either make learning the language the objective, or stick to 
a language they know or use a clean language they don't know, but 
that is perfect for the course. *shrug*