Talk:Break OO privacy: Difference between revisions
→Context?
Line 28:
:::::::::: No. For example, consider /proc/FOO/mem under bsd or linux (where FOO is 'self' or a process id). That said, you do need the appropriate file access permissions. --[[User:Rdm|Rdm]] 15:10, 25 October 2011 (UTC)
::::::::::: but that is what i mean: if you don't have permission the OS prevents access, and circumventing OS access is beyond the scope of rosettacode.--[[User:EMBee|eMBee]] 01:36, 26 October 2011 (UTC)
:::::::::::: I thought you were talking about "preventing access to private datastructures within the process". The OS permissions are based on ownership, so anything with generic file system access can read /proc/self/mem. --[[User:Rdm|Rdm]] 03:23, 26 October 2011 (UTC)
::::::::: and is it fair to say that circumventing OS restrictions is beyond the scope of the issue we try to highlight?
:::::::::: That would depend on "we" and "the issue", no? If you define the OS as being outside the scope of your issue, then by definition, it's outside the scope -- but that also tends avoid most practical issues. --[[User:Rdm|Rdm]] 15:10, 25 October 2011 (UTC)
Line 35 ⟶ 36:
:::::::: It's possible for a sufficiently constrained sandbox. That said, even java implementations have had security holes. But this is getting outside the scope of rosettacode (and is totally outside the scope of this task -- this task does not even mention sandboxes). --[[User:Rdm|Rdm]] 15:10, 25 October 2011 (UTC)
::::::::: of course, i only mentioned this in relation to the possibility of circumventing language based access restrictions. i agree this is outside of the scope of the task, but what i am trying to do is to identify what exactly is still within the scope. i'd say accessing memory through /proc/FOO/mem is not.--[[User:EMBee|eMBee]] 01:36, 26 October 2011 (UTC)
:::::::::: Why not? We have a variety of os-specific tasks and implementations here. (Most anything that deals with the command line, or with linking and often enough with i/o winds up being OS specific.) Anyways, by "scope" I was just referring to the task definition -- I was not talking about anything more formal than that. --[[User:Rdm|Rdm]] 03:23, 26 October 2011 (UTC)
::::::: to find out which languages offer this possibility, in particular among languages that allow to inject code at runtime, is a very interesting question.--[[User:EMBee|eMBee]] 12:44, 25 October 2011 (UTC)
| |||