TMSR work: plan for 2019 M8
For teh posterilulz: this post was published one day late.
July started with a committment to a plan and continued with work on CL WWWism: a CL-WHO demo, an overview of Hunchentoot's architecture, a schema and a demo of the same and a review of one of its main components. Not that bad1, is it?
August started with: comments on the lack of a comment box on this here Tar Pit; comments on the quality of Hunchentoot code; and an emergency halt to discussions on #trilema, caused by btcbase logbot downtime and the lack of an immediate replacement.
Another observation would be that I've been completely neglecting documenting other (more or less important) things in favour of my technical work; and as that string of events I've experienced and places I've seen goes further into the past, my mind keeps distorting and forgetting. Which means that if I don't do this ASAP, I might as well screw it... which I won't, because I'm stubborn like that.
All these given, the plan for the next two months (or so) will have to suffer a few minor adjustments in priorities. In particular August is set to look along the lines of:
- Week 32: I will document2 my July travels and, if time permits, I will wipe the dust off some of my other draft posts.
- Week 33: I will review Hunchentoot taskmasters.
- Week 34: I will review Hunchentoot requests and replies.
- Week 35: I will review and demo some (hopefully really cool) Hunchentoot request dispatchers.
This covers about all of Hunchentoot, with the exception of: headers.lisp, compat.lisp, specials.lisp, util.lisp, set-timeouts.lisp and mime-types.lisp; some of these I've already covered as appendixes of the main text, some might need special treatment, I'm not sure yet. Ideally each of these reviews should also come with a demo, but as acceptors demonstrate, sometimes the triteness is inherent in this type of documentation.
Before wrapping this up, let's take a look at the medium term:
- The next Hunchentoot demo will have to materialize into a minimally-working WWW-facing logotron. As I see it, writing a piece of software that turns database entries into HTML shouldn't be that big of a deal, and since I'm going to maintain Lisp code, I might as well put it to some useful work. Said logotron will run 24/7 logging the contents of #spyked and #trilema; although to be honest, I'm doing this with some reluctance, since I'm already going through some of the inherent pains of operating an IRC bot with Feedbot.
- I need to add a commenttron to The Tar Pit, which faces me with the prospect of undertaking a large project, which would require implementing e.g. a proper anti-spam. On the other hand I can cut this commenttron into smaller pieces and then implement -- and specify/document, which in my opinion makes the exercise immensely useful -- an item that provides the full functionality (or very close to that) on the client's side, while automating some of my current manual comment-adding work. Eventually this other CL-WWW demo will materialize into a genesis for The Tar Pit, what else.
As for the long term:
- Since Hunchentoot will eventually have to encompass all of its dependencies, this puts into light the problem of upstream requirements, firstly and foremostly a Common Lisp implementation. Since this is new territory to me, it's hard to say how long it's going to take, but either way, the "no CL without working CLtron" point sticks.
- Ave1 remarks the lack of Common Lisp learning material. Jurov recommends CLTL2, while Stan mentions one of Graham's older books. IMHO the problem's not Common Lisp per se, but the fact that there's very little text out there explaining Lisp (in general) from first principles, which makes me seriously think that I should take the first Lisp texts and transcribe and annotate them. This would probably take at least a month, but I don't think it's a waste of my time to take McCarthy's Lisp 1.5 Programmer's Manual3 and arrange it in blog format.
- That pastebin project might turn out to be an important piece of Republican infrastructure, especially in light of recent downtimes. At the very least I should replicate Ben's paste service and genesize the blueprints if he isn't available to do this himself.
And then there's that huge list that I'll have to get back to later. For now discussion goes into #spyked and will be posted at 4.
Just for the record, writing one of those large posts takes me anywhere between one and five days of full-time work. I shudder thinking at how this planning's going to hold when saeculum decides to hit me out of nowhere. Yes, when, not if.
We'll cross that bridge when we get to it, I guess.↩
Sifting through the photos should take about an hour; processing them another one to three, depending on the photos -- working on DSLR raws is not a mechanical process, mkay? Then writing the whole thing up and proofreading will take me another circa two hours, which brings this to 4-6 hours total.
This then should give me some time at the end of the week to reload some Hunchentoot in my head and carry on with what comes next.↩
No, I won't link the fucking pdf, it wholly beats the purpose of what I'm discussing.↩
No comments here yet.↩