Re: [tlug] TLUG Site with Hakyll Update

[Curt Sampson <cjs@example.com>]

On 2019-03-19 13:04 +0800 (Tue), Raymond Wan wrote:

> On Mon, Mar 18, 2019 at 11:54 PM Curt Sampson <cjs@example.com> wrote:
> > Let's keep this moving forward, folks: we're now only six weeks away
> > from losing our server.
>
> Sorry, I thought I had been following this discussion / thread, but I
> guess not closely enough.  What's this about losing the tlug server?

I was wrong about the timing; as Edward said, we lose the server at
the end of May, so we have about ten weeks. That's a reasonably
comfortable amount of time, but still, best not to dawdle.

As far as the current Hakyll on Netlify thing goes, it seems to me to
be working ok so far, with no major complaints from anybody else who's
put any effort into it. We have at least one person (me) willing to
put significant effort into the site compiler side of things (the bit
that takes our markup "source code" for pages and turns it into HTML
with menus and links and whatnot), which is the technically harder
part. (I've been intending to use Hakyll for my own sites for quite
some time now and merely waiting on a sufficient supply of round tuits).

But this is far from fixed in stone. Folks here have used other static
site generators (e.g., Jim has had good success with Gatsby, which as
far as I can tell is an excellent system) and a great deal (probably
most) of the work we're doing for the Hakyll version would carry over
directly to almost any other static site generator. If we decide to go
with the static generated site, we can easily release with something
other than Hakyll, or easily change after release.

A change to another site compiler would mainly involve someone else
(preferably several someone elses) stepping up to do the heavy lifting
on building and maintaining the site compiler. And probably
documentation and support code should be taken into account: have a
look at <http://github.com/tlug/tlug.jp> from the README on down, and
the usefulness of the top-level `./Test` script, and consider how
useful or not it is (particularly if I get hit by a bus, or just bored
of it all) and whether we'd want that level of documentation for
something else we do.

Experimenting with this is even easier than you might think. It would
be no problem to add a second script to the repo that would compile a
second copy of the site from the current "sources" (right now, "copy
four files as is") and develop it in parallel with the Hakyll one. I
am happy to help out anybody who wants to do this, even if they're not
too serious about it. I'll walk you through everything you need in
terms of Git branches, support scripts, merging with the master line,
and so on.

As far as me doing actual development on another site compiler: I'd
have some have in that if it were a Python thing, little if it were
JS, and basically none if it were Ruby. (I've written several times as
much Ruby as those other two languages put together; take my feelings
on Ruby for what you will. :-))

There's also the option of finding hosting and duplicating the
existing system as best we can, or setting up another MediaWiki, or
even Wordpress or what-have-you. I don't think that's a particularly
good idea, but if someone's stepping up to do a serious amount of work
and committing for the long run to the much higher maintenance load, I
have no problem at all going along with that.

Points for TLDR:
1. "Ask not what your TLUG can do for you, but what you can do for
   your TLUG." 
2. Got an idea? Step up and we'll help you implement a
   proof-of-concept.
3. But don't bikeshed, except in designated bikeshedding areas.*

* All CSS is a designated bikeshedding area.

cjs
-- 
Curt J. Sampson      <cjs@example.com>      +81 90 7737 2974

To iterate is human, to recurse divine.
    - L Peter Deutsch