Re: The delay of printing encountered using loop

Probably overkill for a one-liner, but for something longer you can control flushing with the flush/autoflush-related functions/methods from IO::Handle. edit: one nice thing about this approach is that you can specify how different file handles get flushed. For example, auto-flushing log files only gives you up-to-date logs, but doesn't cause the performance penalty of auto-flushing file handles used for big data.

Comment on Re: The delay of printing encountered using loop

Replies are listed 'Best First'.
Re^2: The delay of printing encountered using loop by morgon (Priest) on Jun 25, 2012 at 23:35 UTC
one nice thing about this approach is that you can specify how different file handles get flushed. I agree that using IO::Handle methods is nicer but just the approach with $\| allows for the same granularity. It is not the case that setting $\| to true turns on autoflushing for all file-handles. It only affects the file-handle that had been selected (or STDOUT when none had been explicitely selected). see "perldoc -f select"	[reply]

Replies are listed 'Best First'.

Re^2: The delay of printing encountered using loop
by morgon (Priest) on Jun 25, 2012 at 23:35 UTC

one nice thing about this approach is that you can specify how different file handles get flushed.

It is not the case that setting $| to true turns on autoflushing for all file-handles. It only affects the file-handle that had been selected (or STDOUT when none had been explicitely selected).

see "perldoc -f select"

[reply]