Consider removing redis.Nil error from Pipeline.Exec function generated by subcommands？

[LaodaLuo]

For example, it has two strings 'a' and 'b' in redis, when I try to get 'a', 'b' and 'c' in pipeline it ocur an redis.Nil error from pipeline.exec function.

Next, I use for range to process every cmder's result and got error only in get 'c' command.

[LaodaLuo]

what I mean is that error return by Pileline.Exec seems like an error just for pipeline it self, but the error actually generated by subcommands.

So when I wanna use pipeline I should write something like 'if err != nil && err != redis.Nil { return }'

[vmihailenco]

I am not sure the suggested behavior is better - sometimes you want it this way and sometimes the other way. For example, if your pipeline consists of GET commands, it can be handy to just check for pipe.Err() to be sure that all keys were loaded successfully.

Closing since it is too late to make such change without a very good reason. But thanks for the suggestion.