More actions
[19:50:08] <NUXI> jfledvin: my previous job way back when was embedded [19:50:40] <NUXI> and we had rules about how the code should be endian independent (#ifdef BIG_ENDIAN if needed) [19:50:56] <NUXI> well I needed 20 bytes of random data for cryptographic purposes [19:51:05] <NUXI> and the random number function call gave me 4 bytes [19:51:16] <NUXI> so I cast it onto an array [19:51:26] <mjgardes> gotta get those in order [19:51:28] <NUXI> i was told I had an endian issue in my code by a coworker [19:51:41] <NUXI> on casting RANDOM data into another type [19:52:20] <NUXI> it took entirely too long to convince this person that you cannot have an endian issue in that case [19:52:33] <NUXI> checklist programming [19:53:15] <NUXI> well halfway through that I changed my IRC nick to NUXI here and it kind of stuck [19:56:20] <Byan> what would the solution even be >_> [19:56:22] <mjgardes> in other words, you were too lazy to change it back [19:57:02] <NUXI> Byan: i could actually have changed it so that the array always changed the integer the same way [19:57:20] <NUXI> via #ifdef [19:57:26] <NUXI> or ntohl() [19:57:30] <NUXI> or a few other ways [19:57:42] <NUXI> it was just stupid [19:57:56] <Byan> but then you'd have to cast to something else and then do that? [19:57:59] <NUXI> im not even sure if I argued my way out or if I just wrapper it in ntohl() [19:58:05] <Byan> like that doesn't even make sense [20:00:01] <NUXI> i think I put ntohl() on it, I often ended arguments there with mathematical proofs [20:00:35] <NUXI> which they either got or gave up