Last month started with my getting my MP-WP infrastructure ready for production; plus that debugging guide and the v.pl patch. Then I slowed my pace as the tumultous week 44 came, having me reviewing logs and doing only a (tsk!) halfway job of getting that logger up. Then, in what's now becoming a tradition, I went back to reevaluating things.
So let's see what's going on there. First, the whole thing, bot, database, WWW viewer and all, works as far as I've tested it, which is to say I have a basic item configured and up which listens on/connects to the proper places and grabs stuff from/into the logger database. This is all nice and well, except my ol' spykedbot still has a bunch of functionality there to which I'd like to add the logging bits, or if I do it the other way around I'd like to at least have a good reason for it.
Currently there exist two logger-bot V trees: Stan's1 logotron on one side, which comes with a WWW viewer, an IRC logger and some other useful functionality, e.g. the last-seen command(s); and on the other, Trinque's barebones logbot, emerging from the CL Botworks tree, which tree also hosts Feedbot, spykedbot, Auctionbot and from my knowledge Deedbot. Now, upon examination: on one hand the WWW viewer from the former would be immensely useful, because it's there and it works, as other people using it have shown; but on the other, I'm not sure porting spykedbot (as it is) from CL to Python would be a very good use of my time... and there's no need for it either, as the logbot can with some minor modifications be plugged into the same database as the Alf-logotron WWW.
These technical details aside, it occurs to me that a. there's at the moment three loggers2 running and standing up a fourth isn't a burning matter, though the exercise above will have been useful for whenever it is I'll need to get it done; and b. I'm already maintaining that one bot, coincidentally the one that speaks in the channel without being asked, and the prospect of having to manage another one, coming with its own peculiarities, doesn't make me extraordinarily happy. In any case, I'll keep this on the table, but I won't make it a priority, at least not for this month.
And well, as if the tree of references didn't get big enough, I've been talking with Trinque regarding getting Ircbot primed for an update to Deedbot wallet and the new voice model, which would also help with what I've initially had planned for Feedbot for this week. So then the plan for the current month would be to:
- Week 45: publish a Feedbot spec that I owe to the bot's users; hopefully I'll get this done before the end of the week, which should leave me some extra time for the next item.
- Week 46: judging from previous experience, I expect this to flow over into W47, though I don't expect for it to take as much as previous code-study pieces, nor do I want it to; the idea would be to a. integrate the cl-irc dependency into the Ircbot V tree, followed by b. in the form of an article, an exhaustive assessment of cl-irc code that is useless as far as Ircbot is concerned; add to this an estimation of the effort it would take to clean it up.
- Week 47: in addition to cl-irc work, there's some minor cleanup in Ircbot itself that Trinque and I have done independently, and I'm to publish my V patch for review.
- Week 48: following previous work, by the end of W48 I should have at least a partial, if not a full cl-irc cleanup V patch published.
I expect December to also be spent on IRC botworks, though I don't know yet what precisely that will consist of, most likely Feedbot and Deedbot related work.
-
"Wartime", he says, though I will say, the code is excellently written and documented. The only aspect that nags me is that I had to go through the "pip install" nonsense in order to get it working, and no, it wasn't straightforward, because there's a "pip" for every Python version and gotta choose the proper one, and because the remote end of this PIP thing requires HTTPS to work, and well, add one or two more items to this pile of bring-up work and I easily get close to an hour wasted just on tweaking, downloading and getting the pieces together. Or maybe I'm just too slow. ↩
-
Stan's, Diana's and Eric's, though with Stan's departure, it's not entirely certain to me whether his logger is still going to log #trilema and/or other castles. For what it's worth, I'm still using it for the time being. ↩
>there's a "pip" for every Python version and gotta choose the proper one.
I loathe python myself but have found using `python -m pip install ` to save me a lot of headache, as it will explicitly use the python interpreter I have chosen using eselect, so I don't have to fuss about with remembering to use "pip" or "pip3.x" or whatever.
Re 1: It's indeed a wartime kludge, and for the reasons others already described, i.e. demands quite heavy tower of ??? deps. Certainly wasn't written as a "logger for all time."
Re 2: I will continue to log all chans that I find of interest, but only where the chan operator permits the bot to echo. (If anyone concretely does not want my bot in theirs, please let me know.)
@shinohai: Figuring that out was the easiest thing. The confusing part came when for some reason pip couldn't access the repository, so I assumed that it went down (as they usually do, which is why I don't like relying on these arbitrary "repositories") and started searching for alternatives. To my surprise, the link was online, only the server now listens only on 443.
@Stanislav Datskovskiy: I still think grabbing all the so-called libraries and adding them ad-litteram to the V tree is the right thing. While I agree that that'll result in a lot of bloat, it will to some degree that the thing will work on existing Unix+Python systems.
An update on this month's schedule: following discussions in the forum (a, b, c), I'm suspending bot works for the time being, in favour of Cuntoo work. For what remains of this week and at least the next one, I will be studying and running Trinque's Cuntoo bootstrapper, with a view towards understanding the way it works, its current limitations and what needs to be done in order to get to a proper TMSR OS. More on this soon.
[...] with it; and I made some minor modifications to ye olde Perl Vtron. Then as November began, stuff happened and... well, let's make a short parenthesis [...]
[...] Last month's (initial) plan got halted and then changed by the waves of the tumult spreading further, followed by a shift in perspective and immediate objectives within TMSR. In other words: that other work I was going to do was deemed "not that much of a priority", so the only major item1 in my M11 was the Cuntoo bootstrapper notepad -- not sure that's going anywhere either, since meanwhile TMSR-OS management changed and there seems to be an alternative out there, different in some respects, and currently in need of further examination2. [...]