Re: [naviserver-devel] Naviserver and OACS (compatibility)?

[Zoran V. <zv...@ar...>]

Am 30.12.2005 um 12:46 schrieb Stephen Deasey:

> On 12/30/05, Zoran Vasiljevic <zv...@ar...> wrote:
>>
>> Am 30.12.2005 um 03:28 schrieb Vlad Seryakov:
>>
>>>  So, now with your patch it works the same way as aolserver? Then
>>> keep it.
>>
>> We will have to maintain a path between compatibility and cleanness.
>> Not for any price but at places where it seems oportune.
>> I agree with Vlad.
>>
>> Zoran
>
>
> Oh.  Looks like Gustav hacked around this already:
>
> http://cvs.openacs.org/cvs/openacs-4/packages/acs-tcl/tcl/request- 
> processor-procs.tcl? 
> r1=1.82.2.2&r2=1.82.2.3&sortby=date&only_with_tag=oacs-5-2
>
> If that's OK, then we could after all swap the args for filter procs.
> Old code wouldn't work though.  What to do?

I would not require people to use such hacks, if possible.
But I've been looking in the AS code and saw that ugly
      GetNumArgs(Tcl_Interp *interp, Proc *procPtr)
which was used to decide wether an arg is expected or not.
Brrrr.... that is pretty ugly.

Why would we need to process that optional argument in the first place?

Lets make the ns_register_filter as:
     ns_register_filter when method urlPattern script

The "script" should be completely specified, for example

    set arg "whatever thing"
    ns_register_filter preauth GET /junk [list myfilter $arg]

During runtime we'd append the WHY so it will be called as

    myfilter "whatever thing" preauth

Wouldn't that be better/cleaner altogether? This will not break
any code using no optional args (we never used this arg thing)
and would be much more logical to understand.

Hm?

Zoran