Timeline for What is the ending-in-y version of 'atomic'?
Current License: CC BY-SA 4.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 22, 2018 at 14:30 | comment | added | Jesse Williams | @HagenvonEitzen - very good point! | |
| Oct 21, 2018 at 9:56 | comment | added | Hagen von Eitzen |
Nitpicking, "as small as it canpossibly be" may not be the right notion either. E.g., in a high level language such as C, you cannot subdivide an instruction such as x = 42 and yet it may not be atomic (depending on the type of x). Even at CPU level, a single instruction such as INC mem is not necessarily atomic. Locking or other technical measures, sometimes special instruction variants (and in particular some that at first glance or by their very name - COMPAREANDSWAP - are not as small as they can possibly be, are put in place, and this gives you the guarantee of atomicity
|
|
| Oct 19, 2018 at 14:26 | comment | added | Jesse Williams | I would hope not, but I've seen things... terrible things. I assume the OP, with the name void.pointer, probably knows a thing or two. But I'd hate for some tech sales guy to wander in here, see this post, think atomicity means guaranteed success, and use it in glossy slides. And yes, I've experienced this lol. | |
| Oct 19, 2018 at 14:23 | comment | added | Dan Bron | I am not totally persuaded that OP, or his technical colleagues, or any one else, really, thinks there can be or is some characteristic or methodology that guarantees success. I think IT people understand atomicity the same way traders do IOC or FOK orders; I don't think there's any confusion on the matter. | |
| Oct 19, 2018 at 14:16 | history | answered | Jesse Williams | CC BY-SA 4.0 |