tag:blogger.com,1999:blog-753668960778118906.post4750532330549392328..comments2021-03-22T10:54:50.159+01:00Comments on How Can I Explain This?: Coding styles that make me twitch, part 4bakkushanhttp://www.blogger.com/profile/05417272581094785872noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-753668960778118906.post-82393523956590120302009-10-06T22:59:06.214+02:002009-10-06T22:59:06.214+02:00Yes, this is clearly a YMMV situation.
In the cas...Yes, this is clearly a YMMV situation.<br /><br />In the cases where I've found a need for specific OS semantics, it's usually been better to implement the job in a shell script; I prefer something close to POSIX shell.bakkushanhttps://www.blogger.com/profile/05417272581094785872noreply@blogger.comtag:blogger.com,1999:blog-753668960778118906.post-60663548059179017822009-10-06T04:18:45.576+02:002009-10-06T04:18:45.576+02:00I have a couple of toolkitty like things I keep fo...I have a couple of toolkitty like things I keep for doing system maintenance stuff.<br /><br />I actually like the use of operating system mv functions like that sometimes.<br /><br />If the reason to inline it is to reduce the need to fork, the reason you KEEP using them is you want absolutely certainly that the operation will retain the full system semantics, which the Perl equivalent reimplementation might accidentally break.<br /><br />Of course, sometimes you DO want to have the Perl versions and have more control.<br /><br />Just depends on the situation.<br /><br />(The md5sum on the other hand, should probably be Perl)Unknownhttps://www.blogger.com/profile/16880485256236829010noreply@blogger.com