Mail Archives: geda-user/2015/02/09/09:57:16
On Feb 9, 2015, at 2:04 AM, Chris Smith <space DOT dandy AT icloud DOT com> wrote:
>
>> On 8 Feb 2015, at 20:51, John Griessen <john AT ecosensory DOT com> wrote:
>>
>> On 02/08/2015 12:54 PM, Svenn Are Bjerkem wrote:
>>> I can't help myself asking: Why isn't the gnetlist code exposing the needed information to the API? Too difficult to do, or too
>>> boring to get done?
>>>
>>
>> It must be the temperament of those that decide to spend time on the project.
>>
>> a. Too boring to do.
>> b. Requires too much slogging for too little braggable results.
>> c. Does not pay well.
>> d. It's a large enough task that some cooperation would be a great help, but it's hard to find any...
>> e. "Where to start?" *AND* "I don't like writing documentation."
>>
>> Just my guesses.
>
> No, I just think it’s the inherent problem with domain-specific parsers and interpreters. Those things were not implemented simply because the author didn’t envisage a need for them. That’s the advantage of generic, structured file formats like XML or Lua: any parser or interpreter must be complete, because there is no way of predicting what may or may not be useful.
In this case, the parser *is* complete, so you don’t understand what the problem really is, do you?
> The advantage of picking a format which is also native to a language, like Lua, means that scripting and manipulation of that data is also much easier, because there no need for a secondary parser or interpreter.
geda-gaf already has that with configuration files (they are Scheme). It has been a serious problem. The development version has support for configuration via .conf files: common, familiar, trivial to parse. That’s what we need: not a fancy format tuned to somebody’s favorite language, but a format that’s easy to read/write for everybody.
John Doty Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd AT noqsi DOT com
- Raw text -