| <<<Back 1 day (to 2016/08/20) | 20160821 |
sebras | tor8 (also for the logs): hm.. in Array.put(idx, obj) if the idx is out of bounds I guess Java programmers expect IndexOutOfBounds exception, but the only way for me to detect this is to check the size using pdf_array_len() and validating the index before calling pdf_array_put(). do you want to have these kinds of policy decisions in the JNI layer? | 04:13.13 |
| tor8 (also for the logs): or should we skip throwing IndexOutOfBounds exception in these cases? | 04:13.38 |
| tor (for the logs): maybe the JNI interface should be as forgiving as the core interface and not throw exception as frequently as java programmers may expect..? | 05:16.13 |
bodeen | what | 07:56.38 |
sebras | tor8 (for the logs): how do you feel about having from_*() throw exceptions if the argument is NULL and then having each JNI_EXPORT-function just check for exceptions using env->ExecptionCheck()? | 16:28.37 |
| tor8 (for the logs): this lends to shorter code and probably easier to understand. the top patch over at sebras/wip illustrates what I mean. | 16:29.00 |
| tor8 (for the logs): diffstat doesn't show savings yet, but I only converted a handful of JNI-functions so far. | 16:29.47 |
| Forward 1 day (to 2016/08/22)>>> | |