It is said, and by now well-known as a matter of fact, that making roasted cabbage out of one's priorities, words and assumptions inevitably leads to pain, which, when improperly digested1, itself leads to a feedback loop of further pain, spilled guts and rivers of blood. This is how, back in November, after months of rollercoaster-ing, I've ended up in a similar situation; and how, after further rollercoaster-ing and making the conscious decision of settling down into something that's more than a rat race, I've arrived at this particular point in time and space.
It occurs to me2, after circa three years of following the unfolding history of The Most Serene Republic, and about a year and a half of just beginning to participate, that this is by far the most interesting, intellectually challenging (and demanding), fun, promising, etc. project one could think of, or in any case, the only (as far as I know) serious thing happening while the world's busy derping about. So since it's finally occured to me, it's high time I set a (realistic, I hope) schedule for my short-term work, also taking into account saecular matters. Furthermore, I will have a glance into other future work and invite members of the Lordship to comment on priorities.
My schedule for the following two months or so revolves mainly around making the feedbot code publicly available. I have set the following timeline for myself:
- Feedbot maintenance, debugging and development workflow improvement: this remains ongoing work throughout the following period, as small issues continue to appear;
- Reading and publishing XML parser code: 16th of February;
- Reading and publishing RSS parser code: 2nd of March;
- Refactoring and polishing the core functionality, gnarl removal, exposing a clean API for the feedbot programmer: 24th of March;
- Full feedbot V patch: 6th of April.
On the medium and long term, my queue contains other tasks that are either more ample and require further planning, or they are hard to estimate because of (potentially unknown) unknowns. Currently on the list:
- Ircbot sometimes (often? always?) fails to reconnect when Freenode weather goes bad, resulting in deedbot and feedbot outages. This bug remains somewhat elusive, as it's not immediately clear what causes it -- judging by the code, it should reconnect automatically.
- The current incarnation of the Common Lisp curl library known as Drakma (used by, among others, feedbot) is a mixed ball of usefuls and shit, and so it could use a V genesis-ization and a TRB-style cleanup, e.g. removing SSL from it.
- Adalisp exists, but it's not immediately useful, i.e. there is basic functionality missing that prevents it from being a (or the) proper Republican scripting language. This item is especially relevant in the light of the emergence of the Republican Linux distribution.
- Speaking of which: I myself haven't installed the current incarnation of Cuntoo yet. Furthermore, it's likely that vpatch grabbing from within Cuntoo will be a bot-driven service that can stand upon the current botworks tree.
- Catching up with FFA, testing the Keccak TRB regrind, and many others.
Comments are, as usually, more than welcome.
Update, February 4, 2019: comments in the forum have led to the following items being added to my queue:
- The Tar Pit blogotron has various shortcomings that require attention, some of which have been known ever since I started working on the new and improved blog scaffolding, but which have been shelved due to other tasks taking priority. More urgent remains the need for a V genesis of the thing. Although I'm not sure if my Lisp-based blog scaffolding is going to be a replacement for MP-WP in the foreseeable future, it remains a good exercise in building a Common Lisp-based blog from first principles, adding as much functionality as possible from MP-WP over time -- e.g. comments, pingbacks, anti-spam and many others. As a first step, I will have to break everything into a set of well-specified tasks, followed by an actual plan.
- The so-called Project Gutenberg is not long for this world, and thus I and/or another hero will have to salvage it while it's still online. Fortunately, it seems a rsync away, so I should be able to give it a shot by the end of the current week.
-
The pain, although we could probably say the same about the cabbage. ↩
-
To be fair, I was hit by this way back when, (re)reading Trilema and the writings predating it (meanwhile no longer online), discovering Loper OS and then spending countless sleepless nights going through the rabbit holes that are the logs and the Lordship's blogs, I repeatedly went "well, what in the blazing bowels of Beelzebub am I doing with my life?"
It took quite a while to properly put it into words. ↩
[...] proceeding, I read the "Mirroring How-To" guide, which pointed me to a place called ibiblio.org, which [...]
[...] post is part of a series of published artifacts that (will) represent components for The Republic's RSS bot, Feedbot. The [...]
[...] is the second part of a series on building blocks for [...]
[...] of Common Lisp WWW tools, since these seem to be sorely needed by Republicans; they include e.g. Drakma (also used by e.g. Feedbot), but also a HTTPtron (which Lobbes will also look into) and a CL-WHO; [...]