I don't understand why it's so important not to depend on boost_date_time. It

Might not be a big deal,  I just have no problems to compile tu_timer
related stuff, but I do have problems to work with boost_date_time.

The problem with tu_ was initially knowledge. I tought it was best to
use existing *standard* or de-facto standard classes rather then custom
ones, in particular to help a team of developers with a good documentation
and more peer-reviewd implementations of common algorithms.
I haven't seen the timer one, but that's a single function and I don't
have any big problem with it. If boost provides an equivalent function
"for free" (ie: not increasing deps) that's fine too for me.

Again: it's not the names, it's the lack of classes documentation, proper
copy constructors or assignment operators and so on... it's usually not-clean
code in general. And yes, I confirm we never discussed timers in particular
till now.

I'd say tu_timer is clear and simple enough and with enough commentary
there.  "lack of classes documentation" is unfair this time:)

I just give my vote for tu_timer against boost_date_time, it makes my
programming easier at the moment.


_______________________________________________
Gnash-commit mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/gnash-commit

Reply via email to