[gopher] Updated Gopher RFC

[Nuno J. Silva]

On 2012-05-20, Bradley D. Thornton wrote:

> On 05/19/2012 12:28 PM, Nick Matavka wrote:
>> On 19 May 2012 14:11, Nuno J. Silva <nunojsilva at ist.utl.pt> wrote:
>>> On 2012-05-19, Nick Matavka wrote:
>>>
>>>> I have an idea for selectors.  There is a large difference between
>>>> word processing files and page layout languages, so I recommend
>>>> keeping both d and p selectors.  d could be for word processing
>>>> documents such as Word, OpenOffice, and WordPerfect, and p could be
>>>> for page layout and markup languages such as LaTeX, PDF, PostScript,
>>>> and Rich Text Format.
>>>
>>> About LaTeX, I believe either you're serving the code as something for
>>> the user to read (hence type 0 or another type that gets established for
>
> I expect to read a 0 in my client, and download a 9 so I can choose from
> the myriad of apps I choose between to open a particular type of file.
>
> Wouldn't anything else require bloated intelligence on the part of my
> client?

The question here is whether it makes sense to say LaTeX source is
something you can see using a "document viewer". You can write a script
that processes the LaTeX and shows the result (in PostScript, PDF, DVI,
...), but other than that, the only way a LaTeX (or plain TeX) document
can be read without further processing is if it's simple and short
enough to be served as type 0. 

(LaTeX code is readable, but you may want to skip the preamble, and some
documents may be too long (although that's not much of an issue))

For already processed documents (PDF, PostScript, djvu), the only
intelligence involved would be launching a document viewer to show the
content. 

The client would either offer the option to open in a document viewer,
or would be coded to treat these selectors the same way it handles a 9.

-- 
Nuno J. Silva (aka njsg)
gopher://sdf-eu.org/1/users/njsg
http://njsg.sdf-eu.org/