• Is there a "modern" Gopher stack?

    From Doc O'Leary ,@droleary.usenet@2023.impossiblystupid.com to comp.infosystems.gopher on Mon Jul 20 23:55:32 2026
    From Newsgroup: comp.infosystems.www.auth

    I recently played around with setting up a Gopher server, and as much as I like the old-school-simplicity of the protocol, what discourages me from maintaining public-facing spaces is that I canrCOt seem to find any sorts of authoring tools that integrate well with my non-1990s work flow.

    For reference, I do have a couple bare bones web sites. They use Apache `autoindex`, which sets up navigation rCLmenusrCY for me much like Gopher. My new content is mainly Markdown these days, which gets converted to HTML
    for the browser. A bit of CSS gets it looking (I think) a little bit
    better than the default render. Done. I do also have some much more
    advanced dynamic sites, but for publishing basic content, IrCOd really like
    to not have to deal with *any* of the everything-and-the-kitchen-sink web.

    The rCLproblemrCY is that I canrCOt just point a Gopher server at that same generic content and call it done. IrCOm willing to find some middle ground and adapt my web site generation, but that has me wondering what other
    people out there might be using (if anything!) in 2026 for their own
    setups.

    In particular, Gopher menus/maps may have somehow seemed like a good
    format choice at the time, but I find them to be *terribly* dated. At
    least conceptually, I could see an easy lateral move to actual TSV
    support, allowing Gopher to maybe be seen as rCLthe web, but using spreadsheetsrCY.

    One other thing I tried that worked well was using YAML as the base
    format, with the document being a list of text nodes mixed in with the
    links as arrays. The raw text file was itself *very* readable/editable,
    and it was almost trivial to convert it to a gophermap file that the
    server was happy with. And with YAML being a superset of JSON, I could
    see that method making integration with a lot of existing web pipelines
    very doable.

    So back to the question: does anything remotely *like* those approaches already exist? Any thoughts/pointers would be appreciated. Thanks!

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From sean@sean@conman.org to comp.infosystems.gopher on Tue Jul 21 05:05:28 2026
    From Newsgroup: comp.infosystems.www.auth

    It was thus said that the Great Doc O'Leary , <droleary.usenet@2023.impossiblystupid.com> once stated:
    The rCLproblemrCY is that I canrCOt just point a Gopher server at that same generic content and call it done.

    Perhaps get a better gopher server?

    IrCOm willing to find some middle ground
    and adapt my web site generation, but that has me wondering what other people out there might be using (if anything!) in 2026 for their own
    setups.

    I wrote my own gopher server [1], and if there doesn't exist a file named "index.gopher" or "index.port70" [2] in a directory, then a gopher map will
    be generated on the fly. The "index.gopher" file is a file that is already
    a valid gopher map file (with tabs and all that), but the "index.port70" is
    an easier format to author that my gopher server will translate on the fly. So, given a ".port70" file like:

    +------[ top of file ]--------
    |Welcome to the server! This is my server!
    |
    | dir selector1 Text to display for user
    | dir selector2 More text
    | file selector3 Blah blah blah
    +------------------------------------------

    will be tanslated to

    +-------[ top of gopher map ]-------------
    |iWelcome to the server! This is my server! example.net 70
    |i example.net 70
    |1Text to display for user selector1 example.net 70
    |1More text selector2 example.net 70
    |0Blah blah blah selector3 example.net 70
    +------------------------------------------

    I also wrote a module (similar to an Apache module) that will convert my
    blog entries [3] from HTML to plain text (wrapping with a margin of 80 characters) [4].

    So are you wanting to just poing a gopher server to an existing web site
    and have it "just work?" Or just to an existing directory and have it "just work?" Because the approach for one isn't the same as the other.

    So back to the question: does anything remotely *like* those approaches already exist? Any thoughts/pointers would be appreciated. Thanks!

    Yes. And I personally avoid YAML.

    -spc

    [1] https://github.com/spc476/port70

    [2] Default names, but can be changed by the configuration file.

    [3] https://boston.conman.org/

    [4] gopher://gopher.conman.org/1Phlog:
    --- Synchronet 3.22a-Linux NewsLink 1.2