Anonymous user
Talk:File size: Difference between revisions
→task clarification: add comments about lines vs bytes. -- ~~~~
m (→task clarification=: added comments about CMS being modern. -- ~~~~) |
(→task clarification: add comments about lines vs bytes. -- ~~~~) |
||
Line 10:
:: My first interpretation was, in fact, the size of the file in lines (records), and nobody bothered to comment on the REXX solution which reported it thus. (I've since changed it to reflect bytes.) If you ask most (mainframe) programmers what's the size of that so-and-so program, the answer is almost always: xxx lines-of-code (or xxx lines). To most (old school) mainframe programmers, a ''program'' most likely refers to the ''source'' (measured in lines-of-code), the ''module'' would be the executable, and that would be measured in bytes. Most mainframer programmers tend to think in terms of (logical) records for a file instead of a stream of characters (bytes). -- [[User:Gerard Schildberger|Gerard Schildberger]] ([[User talk:Gerard Schildberger|talk]]) 20:50, 24 April 2013 (UTC)
:: In fact, the original version of the REXX example did use the '''lines''' BIF which returns (for most REXXes) the number of lines (records) in the file. The CMS, R4, and ROO REXX interpreters do this by default, Regina can mimic this by the use of the '''option noFast_lines_bif_default''' statement, or via the 2<sup>nd</sup> argument for the '''lines''' BIF: '''lines(xxx,'C')''' --- '''C''' is for '''C'''ount. -- [[User:Gerard Schildberger|Gerard Schildberger]] ([[User talk:Gerard Schildberger|talk]]) 21:21, 24 April 2013 (UTC)
| |||