sdiff output anomaly

[George Eadie]

We are running sdiff against some test output and finding that we appear to be getting "strange" anomalies in the output from sdiff. It appears that sdiff is indicating differences in lines that are not different, and it is not showing lines that do differ. We are running sdiff on HPUX and find this problem. However when I run sdiff on Linux using the same data files we get the output I would expect.

Has anyone seen this before and can you point me in the direction of any fix.

[Jason Ray]

George

Can you post the original files so that I could try it out myself?

Are these ASCII files or UTF8? I recently was running sort on ASCII and UTF8 files and noticed a difference depending on the locale that was set for the user running the command.

[F Verschuren]

It can be that the normal output of this comand is differedn on linux please try the -s option (that normaly removes the lines that are te same.

ps what does a normal diff say?

[George Eadie]

Gentlemen

Thanks for your replies.

The files we are comparing are text files. Running an sdiff -s using the sample files I have does clear the problem somewhat, however it still appears to show lines that are the same, as different. I've attached copies of the files I am comparing. Running a diff against the two files shows similar anomalies to the output I get from sdiff.