[RAS] automatic suggestions count

[Christian Zimmermann]

On Fri, 14 Aug 2009, Thomas Krichel wrote:

>  Christian Zimmermann writes
>
>> After getting some complaints from users that could not find the
>> promsed works, I have monitored those that bounce, and I not found a
>> single instance where this number was right. In this particular
>> case, it should 107.
>
>  Are you shure this is ann issue of the perl code, it could
>  be in the presenter
>
>> I could not find a systematic error. Sometimes, it is too high,
>> sometimes too low. But it is never lower than the number of items
>> directly suggested in the message.
>
>  Then why don't we just add "at least" to the wording?

No, because sometimes the actual number is below what is reported.

>
>> I have also added this to my RAS wish list at
>> http://ideas.repec.org/zimm/personal/raswish.html. Note that I have
>> been maintaining this list for years, yet not a single item has been
>> dealt with. I find this very sad, and frustrating, as these are all
>> suggestions coming from the actual use of the product, i.e.,
>> consumer tested. I have even added priorities to the items so that
>> one can figure out what is more urgent.
>
>  Well, we have to find additional people to maintain RAS.
>  It will be an uphill struggle beacuse ACIS is almost
>  completely undocumented.

Note that the vast majority of the suggestions also apply to ACIS and to 
the current implementation you are working on. And many of them should be 
easy to implement for somebody who has some knowledge of the internals 
(you and Ivan).

>
>
>  Cheers,
>
>  Thomas Krichel                    http://openlib.org/home/krichel
>                                RePEc:per:1965-06-05:thomas_krichel
>                                               skype: thomaskrichel
>