Re: Patches item #1729825 ready to go.

[<pl...@pi...>]

On Fri, 08 Jun 2007 18:04:46 +0200, Daniel J Sebald  
<dan...@ie...> wrote:

> pl...@pi... wrote:
>> On Fri, 08 Jun 2007 00:25:05 +0200, Daniel J Sebald   
>> <dan...@ie...> wrote:
>>
>>> Also, I should point out that the patch removes all duplicate spaces   
>>> from input
>>> lines before placing them in the history buffer.  This makes things  
>>> more  tidy
>>> and also makes things like "plot x" and "plot            x" eqivalent.
>>> Dan
>>   I'd be careful about "tidying" things up. History should show what  
>> was  done , not some arbitary sanitised version. If ppls want "clean"  
>> they  should tap clean commands. I see no justification for this  
>> feature.
>>  Without searching for a specific example, I may want to see why  
>> something  happened differently to I expect (bug of finger trouble).  I  
>> NEED to see  exactly what I typed. Please dont mess with.
>>  Appreciate your efforts on this I think the clean up and new ideas  
>> look  very good in general.
>>  ;)
>
> "history" is a very particular item, I see.  I can leave it out; its  
> effect probably isn't that significant in reducing redundancy when in  
> condensed mode.
>
> Dan
>
Not sure if you understood what I was meaning. I was refering to stripping  
out spaces etc. not whether "history" appears. IMHO history command should  
reproduce what was typed at the command line (unless some option like  
condenced is activated).

As for the question of "history" itself appearing at the last command , I  
would suggest it's probably non-std behaviour of ksh you quoted and  
gnuplot should probably follow bash/csh/zsh etc. and display the current  
command, ie history, in the list.

The latter is minor detail not worth further discussion. I consider the  
first point to be important. Dont try to rewrite history ;)