On Mon, Feb 1, 2016 at 12:13 PM, Roy Stogner <royst...@ices.utexas.edu>
wrote:

>
> We ought to just store order internally as an int, change the existing
> constructors to convert, and add new constructors that take int.
> ---
> Roy
>


I just wanted to check in about this.

The order member in FEType is public, and it's used in many places, so
changing the order member necessitates lots of other changes in the library
(for example, consider all the methods in the FE class that expect
fe_type.order to provide an Order enum).

I'm not sure I see a good way to do this while retaining backwards
compatibility. John's suggestion would be backwards compatible ("you might
have a new FEType constructor that takes an unsigned int parameter while
setting the existing enum order==INVALID_ORDER" ) but then we'd have to
store two order members, an int and an Order, which is probably not what we
want.

Alternatively we could do something backwards incompatible, like this:
- make order private in FEType
- change it from an Order to an int (where negative values are interpreted
as INVALID_ORDER)
- provide getters FEType::get_order_enum() and FEType::get_order(), where
get_order() returns an int and get_order_enum() returns an enum (or an
error if the order is higher than the max value of the Order enum).

Anyway, input on this would be appreciated, since I'm not sure if there's
an appetite for backwards incompatible changes. If we can agree on an
approach, then I can go ahead and make a PR.

David
------------------------------------------------------------------------------
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140
_______________________________________________
Libmesh-devel mailing list
Libmesh-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/libmesh-devel

Reply via email to