Giugno

TL;DR - today will be about getting ducks lined up for FOAF Retrospective

Note to self : remember the TL;DR on posts, even if they are only notes to self

Yesterday I thought I'd got #Fuseki up and running (not configured). Today a 502 Bad Gateway. Hey ho.

I also resurrected most of the hyperdata-static material on https://hyperdata.it - at this very moment a section that weirdly wasn't actually in that main tree is being git-pushed to github. 5,400 files...

Right now #FOAFRetrospective is my priority for webby world activities. Like it has been for 3? months... I did slip into some serious #YakShaving with #Transmissions (my pipeliney thing), but I believe I've got that under control now.

My motivation for getting Fuseki and hyperdata.it bits live again is to make use of some of the little tools I've got there for FOAF Retrospective. Various bits geared towards information capture and exploitation.

Grr. The git push just stalled. I haven't checked, but I don't think I've got any huge files in there that GitHub would balk at, probably just my connection being a bit flaky (as usual). Size-wise, altogether the tree isn't massive, but there is quite a high number of small files. Running from VSCode (where I'm typing this), it was silent, I've just set it going again, adding 'Show Output'.

Yeah, I made a strategic error with FOAF Retrospective (my working title). Because there was really significant overlap with the things I was working on already, it made sense to fold it in. Fool Danny, I (yet again) forgot I'm really bad at estimating time for projects. While there isn't a hard deadline, before now would be the soft deadline. I'm kicking myself a little because I'm well aware of my own failings in a context like this, and I do have strategies for working around them. The one here is straight out of the Agile playbook, to go iterative, incremental, and evolutionary. I should have got something quick & dirty up first (corresponding to running code in that playbook), then looked at how to improve it. Instead I went somewhat waterfall.

But not to worry. I can flip over to the better strategy now, in a position where I have the tooling the waterfall produced. (That tooling itself isn't exactly finished, but is at the point where I can incrementally improve it as needed).

Oh FFS, I just got a message saying payment was declined for Linode - my server hosting. Annoying, I thought I'd enough left on my card to cover that. There will be next week some time, do I hope they have a reasonable grace period.

Hmm, server says:

New release '24.04 LTS' available.
Run 'do-release-upgrade' to upgrade to it.

It's probably a good time to get that out of the way, once I've got the missing hyperdata-static bits back in place.

> git push origin main:main
error: RPC failed; HTTP 400 curl 22 The requested URL returned error: 400
send-pack: unexpected disconnect while reading sideband packet
fatal: the remote end hung up unexpectedly
Everything up-to-date

Grr. I'd better check for big files. From a couple of days ago:

find ./ -type f -printf '%s %p\n' | sort -nr | head -30 | awk '{print "rm " $2}' > bigfiles.sh

Tweak (thanks ChatGPT!):

find ./ -type f -exec du -h {} + | sort -rh | head -30

A big .log file, some .jars, then images and videos. Another tweak:

find ./ -type f -name '*.log' -printf '%s %p\n' | sort -nr | head -30 | awk '{print "rm " $2}'

The jars aren't needed (may well be unsafe when used), they and the logs can go.

Trying a git push from the command line...

That's taking some time (to fail?). I'm ready for a break. might as well set do-release-upgrade going.

Ah wait - backup first.

Plan

  • reorg FOAF Retrospective
  • hyperdata-static bits
  • Get Fuseki running properly, any config necessary for hyperdata tooling

Giugno