Re: [aldor] algebra/matrix: Added some additional tests. (b08727a)

[Ralf Hemmecke]

Hi Peter,

Let me move the discussion to the mailing list.

What exactly are you disliking?

The only thing I am missing would be something like

define Exception: Category == with;

i.e. a root for all the exceptions. But otherwise that exceptions are
domains, is not so bad IMHO.

No, there isn't yet a well-thought-of exception hierarchy. Maybe it
would be a good idea to look into the Java world.

Ralf

On 09/24/2014 10:57 PM, pbroadbery wrote:
> Yes, probably (although I do really dislike the current implementation of
> exceptions).. Is there an already existing exception type to use?
>
> On 24 September 2014 21:18, Ralf Hemmecke <notifi...@github.com> wrote:
>
>> The error function is only very sparsely used in libalgebra. Wouldn't it
>> be better to throw an exception?
>>
>> —
>> Reply to this email directly or view it on GitHub
>> <https://github.com/pippijn/aldor/commit/b08727a0db3cc0665739a11f914a39e2fa9a3b27#commitcomment-7913268>

[Peter Broadbery]

the hierarchy thing is ok - but just abusing types to provide it seems
a little wrong; exceptions are not logically types; moving the
hierarchy down to the object level in some way would be preferable.
There's some debate about declared vs undeclared exns too; undeclared
seems to be the default these days, with declared being used only in
specific circumstances.

> --
> You received this message because you are subscribed to the Google Groups "aldor-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to aldor-devel...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.