Mail Archives: geda-user/2015/02/06/13:54:06
On Fri, 6 Feb 2015, Jason White wrote:
> On Fri, Feb 6, 2015 at 1:07 PM, Edward Hennessy <ehennes AT sbcglobal DOT net> wrote:
>>> On Feb 6, 2015, at 9:07 AM, John Doty <jpd AT noqsi DOT com> wrote:
>>>> On Feb 6, 2015, at 9:29 AM, Jason White <whitewaterssoftwareinfo AT gmail DOT com> wrote:
>>>>
>>>> For those who have not seen, I implemented an example program in C
>>>> using Lua to safely read data defined in Lua files.
>>>
>>> And that is of little use. The schematic and configuration files need to be trivially readable/writable by *any* language
>> <snip>
>> Additionally, it becomes more difficult to create a GUI to edit the data when the data uses a complex file format.
>>
>> I'd prefer simple configuration and data file formats and save the languages for extensions.
>
> What is more simple than name=value or name={value, value, value}? I
> do no see how that is any more difficult to parse than the current
> format.
What is a value? Can it be an expression, even as simple as 1+1?
If yes, users will start depending on that feature and any importer will
need to implement a compatible expression eval. After that, does it
support (local) variables, constants or other shorthands?
If not, then we are walking the way that restricts the user to a very
small subset of the language. If we do that, what's the advantage chosing
a subset of lua for the purpose other than cheaper importer/exporter in
lua (or embedded lua interpreter)? Just asking because the same argument
works for any other data language or programming language that can be
restricted in io.
At the end of the day, I find the "easier to handle in one specific
language" or "easier to handle using a specific library" arguments weaker
than the "easy to parse using a random language" argument.
Regards,
Igor2
- Raw text -