spyked: I'm reserving another bot for the rss functionality. but I want to use spykedbot to make a small trilemabot example for implementing prefixed commands. spoiler: I have some code that looks up definitions in asciilifeform's ro-en dictionary ( http://www.loper-os.org/?p=2211 ), it's the smallest application that I could find that also does something useful. ☟︎

spyked: in other work: does anyone object to my taking the next available prefix in http://lobbesblog.com/static/tmsr_bot_directory.html for spykedbot? !S fits it quite well.

spyked: plus, I tried at some point to bring in server-side dynamic content ( http://thetarpit.org/posts/y03/05c-development-log-i.html ), but the cost of all those dependencies is still (from my pov too) big.

spyked: scaffolding), a xml-rpc handling mechanism, and many many others that the wp user may find useful; it has no dedicated db, the db is the filesystem; no web-based editor, etc.

spyked: http://btcbase.org/log/2018-06-04#1820568 <-- I also suspect that that's much of the cost of "works". leaving comments aside, thetarpit blog scaffolding is a few kLoC of CL, but lacks *any* editor-side interface (an item such as e.g. http://btcbase.org/log/2018-04-15#1799520 would have to be done the old-fashioned way, moving the files manually or through ssh and maybe script this somehow, but it would be a *separate* piece of ☝︎☝︎

spyked: http://btcbase.org/log/2018-06-04#1820598 <-- indeed it does, and precisely because of http://btcbase.org/log/2018-06-04#1820603 . I don't know how to design a proper comment mechanism, not necessarily because of spam (I believe mp-wp antispam works), but because I'm not convinced that everyone coming on the blog should be able to leave comments, and if they do, then what format (text-only?) etc. ☝︎☝︎

spyked: aaaand rssbot code currently weighs circa 400LoC, so I'ma probably also have to break that into testable pieces. not sure yet.

spyked: step 5 will probably need to be broken into multiple sub-steps, since I will have to also publish a rss parser first, then a rss bot design document, and only then the code.

spyked: my next steps will be, in order: 1. add prefixed commands to trilemabot (and publish); 2. add (and publish) command and parsing for !!gettrust queries (required for rss bot access control); 3. rebase current rss bot on trilemabot and add access control; 4. bring rss bot in #trilema; 5. publish rss bot code

spyked: ^ comments, suggestions, tomatoes and eggs are more than welcome.