ScriptLimits (was Re: Rev cgi vs. php)

Terry Judd tsj at unimelb.edu.au
Thu Jan 31 17:30:05 EST 2008


>> The scriptLimits is a real PITA.
> ...
>> At times it feels like I'm using a trial version, purely because of
>> this.
> 
> Back when the engine was owned by MetaCard Corp., Scott Raney's position
> on this was intractable:  he felt there was simply no reason to write
> self-modifying code, and moreover that it was a bad practice anyway.  He
> asserted that in each case where self-modifying code was perceived as
> useful there was probably a better solution available which wouldn't be
> affected by the scriptLimits.
> 
> Since acquiring the engine, Kevin Miller's position has been more
> flexible:  if someone feels they have a need to make standalones which
> rely on having no scriptLimits, he's eager to hear from such customers
> and work out a solution for them on a case-by-case basis.
> 
> Over the years, going back even before RunRev to the MetaCard days,
> there have been more than a hundred complaints similar to this one.

I guess it's worth noting (again) that Director has no equivalent to
scriptlimits and over the years I have taken advantage of this to build a
number of educational tools that create relatively complex chunks of code on
the fly. These tools usually involve simulations that require the
user/student to construct a series of logical statements that will 'control'
the simulation. While it is possible to 'hard code' all the various options
creating the code on the fly is usually easier and more elegant.

Some possible modifications to scriptlimits that might prevent a Digital
Chisel type scenario could include...

(i) making 'long' do statements self-contained - i.e. They can't call any
other functions or handlers
(iii) preventing access to custom properties, globals and script locals
(ii) preventing the creation of interface controls

...sounds a bit like Rev cgi!

Terry...

-- 
Dr Terry Judd
Lecturer in Educational Technology (Design)
Biomedical Multimedia Unit
Faculty of Medicine, Dentistry & Health Sciences
The University of Melbourne
Parkville VIC 3052
AUSTRALIA

61-3 8344 0187




More information about the use-livecode mailing list