| 1 |
On Mon, 24 Aug 2009 15:09:52 -0700 |
| 2 |
Zac Medico <zmedico@g.o> wrote: |
| 3 |
> > I'd like die to respect nonfatal. People using nonfatal should |
| 4 |
> > check beforehand that the functions they're calling won't do |
| 5 |
> > anything stupid if die's are ignored. |
| 6 |
> |
| 7 |
> If you're doing that then it might be wise to add an 'assert' helper |
| 8 |
> that is guaranteed to generate an exception regardless of 'nonfatal' |
| 9 |
> status. |
| 10 |
|
| 11 |
Probably not the best choice of name... 'assert' is already in use for |
| 12 |
checking PIPESTATUS. |
| 13 |
|
| 14 |
The other problem with doing things that way is that people have to |
| 15 |
start writing code that does an EAPI check just to be able to force a |
| 16 |
die. An IGNORE_NONFATAL environment variable, whilst slightly ickier, |
| 17 |
avoids that issue if we do change the definitions of die and nonfatal. |
| 18 |
|
| 19 |
*shrug* I still think the way it was originally worded is the best |
| 20 |
option. Otherwise we'll just replace the "some helpers die, some don't, |
| 21 |
and no-one knows which" situation with "some eclass functions are |
| 22 |
nonfatal aware, some aren't, and no-one knows which", with an added |
| 23 |
side helping of "people need to find out whether any of the callers in |
| 24 |
any overlay anywhere expect this function to work with nonfatal before |
| 25 |
changing it". |
| 26 |
|
| 27 |
-- |
| 28 |
Ciaran McCreesh |
Archives