=> back to Techrights (Main Index)
00:26 techrights[sec]; ack
00:26 techrights[sec]; otr: repeat: https://digital-strategy.ec.europa.eu/en/library/cyber-resilience-act
=> ↺ https://digital-strategy.ec.europa.eu/en/library/cyber-resilience-act
00:26 -TechrightsBN/#boycottnovell-digital-strategy.ec.europa.eu | Cyber Resilience Act | Shaping Europes digital future
01:43 schestowitz-pi2; Layoffs at DailyFail https://pressgazette.co.uk/news/mail-mail-on-sunday-closer-together-redundancies/
=> ↺ https://pressgazette.co.uk/news/mail-mail-on-sunday-closer-together-redundancies/
01:43 -TechrightsBN/#boycottnovell-pressgazette.co.uk | Mail and Mail on Sunday to be much closer together; redundancies proposed
02:17 schestowitz-pi2; db Sharp rise in psychiatric diagnoses among young people after the first wave of the COVID-19 pandemic https://www.helsinkitimes.fi/themes/themes/health-a-wellbeing/23058-sharp-rise-in-psychiatric-diagnoses-among-young-people-after-the-first-wave-of-the-covid-19-pandemic.html
02:17 -TechrightsBN/#boycottnovell-Empty reply from server ( status 0 @ https://www.helsinkitimes.fi/themes/themes/health-a-wellbeing/23058-sharp-rise-in-psychiatric-diagnoses-among-young-people-after-the-first-wave-of-the-covid-19-pandemic.html )
02:52 DaemonFC; schestowitz[TR]: Cantrell did seem less abnormal in 2016, but there were many warning signs already. He would have been 18 at the time.
02:53 DaemonFC; Hadn't done so many drugs yet, before he got COVID.
02:53 schestowitz-pi2; i see
02:53 schestowitz-pi2; 2019?
02:53 DaemonFC; 2016.
02:53 schestowitz-pi2; maybe the mind went "off"
02:53 schestowitz-pi2; i read about it todsy
02:53 schestowitz-pi2; or yesterday rather
02:53 schestowitz-pi2; brain damage
02:53 DaemonFC; schestowitz-pi2: I'd say he went from borderline personality disorder to paranoid schizophrenia.
02:54 DaemonFC; From his overall style of writing and the subject matter.
02:58 DaemonFC; It's common for bipolar disorder and schizophrenia to emerge in early 20s after many other diagnoses.
02:59 DaemonFC; In childhood he says they gave him Ritalin for ADHD.
02:59 DaemonFC; He keeps insisting he's straight and that he likes women, but every time he pays to have sex with one, he can't finish.
02:59 DaemonFC; And he keeps making this odd statements about "gays".
03:00 schestowitz-pi2; www.usnews.com/news/us/articles/2023-03-02/fatal-high-school-stabbing-stuns-california-winery-community
03:00 DaemonFC; I think he was trying not to be gay by forcing himself to have sex with women and actively deny it.
03:04 schestowitz-pi2; there can be simpler explanations
03:04 schestowitz-pi2; unless you have evidence men turn him on
03:04 DaemonFC; I doubt there's anything in there to that effect.
03:06 schestowitz-pi2; ok then..
03:08 DaemonFC; He had a verbal spat with his neighbor a few years ago. Said he can't stand gay people.
03:08 DaemonFC; His neighbor was apparently gay.
03:08 DaemonFC; Had a complaint.
03:08 schestowitz-pi2; so your theory is weakened
03:09 DaemonFC; Ever see the ending of American Beauty?
03:10 DaemonFC; The homophobe next door turned out to be gay and shot Kevin Spacey's character in the head in the end over rejection.
03:11 DaemonFC; https://www.scientificamerican.com/article/homophobes-might-be-hidden-homosexuals/
=> ↺ https://www.scientificamerican.com/article/homophobes-might-be-hidden-homosexuals/
03:11 -TechrightsBN/#boycottnovell-www.scientificamerican.com | Homophobes Might Be Hidden Homosexuals - Scientific American
03:38 schestowitz-pi2; "might"
03:38 schestowitz-pi2; I watched American Beauty as a teenager
03:38 schestowitz-pi2; and don't remember the ending
03:39 schestowitz-pi2; anyway, movies are not facts
03:57 schestowitz-pi2; workflow with one full refresh per day is... refreshing!
03:57 schestowitz-pi2; less for me to go over a second time around
06:30 schestowitz-pi2; DANIEL STENBERG needs to bring these CLOSER TOGETHER to be productive. Sliding on a chair from one computer to another is not efficient. https://daniel.haxx.se/blog/2023/03/03/my-hacker-station/
=> ↺ https://daniel.haxx.se/blog/2023/03/03/my-hacker-station/
06:30 -TechrightsBN/#boycottnovell-My hacker station | daniel.haxx.se
06:35 schestowitz-pi2; just counted: on 9 screens i have 20 terminals visible, 9 IRC windows,2 kate, 1 browser... actually 4 kate
06:35 schestowitz-pi2; same session with 3 windows. the rest is dictionary (rarely used), audacious, large clock, 4 or 5 panels
06:35 schestowitz-pi2; but terminals are predominant
07:34 *Jack3l (~jack@freenode-o3t.n7q.01gjo5.IP) has joined #boycottnovell
07:34 *Jack3l (~jack@freenode-o3t.n7q.01gjo5.IP) has left #boycottnovell
08:17 *DaemonFC has quit (Quit: ChatZilla 0.15 [SeaMonkey 2.53.15/20230121223026])
08:30 *DaemonFC (~daemonfc@3khny76cn5jaq.irc) has joined #boycottnovell
09:08 schestowitz-pi2; in local:/tmp
09:08 schestowitz-pi2; I've just put an example of the file I work with, aside from 09.xxx
09:08 schestowitz-pi2; i think there may be room for code reuse, let me explain:
09:08 schestowitz-pi2; in merged you do a link-wise comparison between current day and prior day
09:08 schestowitz-pi2; then it culls thee duplicates 'forward'
09:08 schestowitz-pi2; in the sense that the newer occurence gets removed
09:08 schestowitz-pi2; in my case, i set cutoff to 48 hours
09:08 schestowitz-pi2; if I refresh once a day
09:08 schestowitz-pi2; then it'll be good to be able to do the same with feedlist.html and feedlist2.html
09:08 schestowitz-pi2; but the formats for these files are less consistent
09:08 schestowitz-pi2; that would save time because, atm, i have to manually check where the threshold is
09:08 schestowitz-pi2; for stuff I saw the prior day
09:08 schestowitz-pi2; if done properly the file can be halved in size, assuming current day and prior day are about the same
09:08 schestowitz-pi2; in total volume
09:08 schestowitz-pi2; in firefox i get some help from the URLs turning purple if I opened them before (or in falkon)
09:31 techrights[sec]; Yes, the current mechanism is to rely on the links being purple by the browser
09:31 techrights[sec]; Only the browser can track if the links have been visited or not.
09:31 techrights[sec]; The -merged files are over on tr-new anyway and not really practical to
09:31 techrights[sec]; keep scanning
09:31 techrights[sec]; from a distance
09:31 techrights[sec]; On the other hand, the -d option could be used to shorten the time span to
09:31 techrights[sec]; one day from two.
09:31 techrights[sec]; ./rrrrr.py -d $(date -d '1 day ago' +'%F') -o x.html
09:31 techrights[sec]; Currently there is a defualt action which is to scan and then print to stdout
09:31 techrights[sec]; Should the default to be to show the help text instead and the -o option become
09:31 techrights[sec]; required for output?
09:34 schestowitz-pi2; Overpopulation tackled. "he current birth rate is 1.34 per a woman, which is below the 2.07 that's necessary to keep the population stable." No, a country remaining OVER-populated is not stability ("stable"). Equilibrium is vastly lower than present. https://www.ibtimes.co.uk/japans-population-crisis-worsens-women-choose-career-over-marriage-while-many-opt-not-have-kids-1713623 | Source: International Business Times
09:34 -TechrightsBN/#boycottnovell-www.ibtimes.co.uk | Japan's population crisis worsens as women choose career over marriage, while many opt not to have kids
09:35 schestowitz-pi2; ombg,sky is falling, we might not reach 10 billion humans. we are doomed!
09:35 schestowitz-pi2; the html xslt is fine as it is
09:35 schestowitz-pi2; i am trying to think how to speed up, by means of reducing repetition in the vast files
09:35 schestowitz-pi2; like the one I added to /tmp
09:35 schestowitz-pi2; sort and uniq don't work well for this type of file
09:35 schestowitz-pi2; still thinking...
09:39 schestowitz-pi2; ok, i've mentally reduce the scale of the task to this:
09:39 schestowitz-pi2; for file x and y, in file y remove all lines identical to something in x
09:39 schestowitz-pi2; not sure is o(n) algo exists for that
09:39 schestowitz-pi2; those files are quite big
09:39 schestowitz-pi2; diff makes assumptions about order of lines
09:39 schestowitz-pi2; ans sortruins the order of the original
09:39 schestowitz-pi2; *and sort() ruins
09:39 techrights[sec]; The carrying capacity of the planet is exceeded by about double, and to make
09:39 techrights[sec]; matters worse the way the global population is acting that carrying capacity
09:39 techrights[sec]; is /decreasing/ while at the same time the population grows.
09:39 techrights[sec]; xmlstarlet can help in some simple cases
09:42 schestowitz-pi2; should i check rss_since.pl for some example of how it's done there? (the dupe reover)
09:44 techrights[sec]; links-de-duplicate.pl or links-find-duplicates.pl
09:44 techrights[sec]; if the hash were made into a tied hash, the values would persist over time
09:44 techrights[sec]; across runs, but I'm not sure how far that can scale.
09:48 schestowitz-pi2; so I'm thinking...
09:48 schestowitz-pi2; one day I make file y
09:48 schestowitz-pi2; another x
09:48 schestowitz-pi2; alternating
09:48 schestowitz-pi2; and then
09:48 schestowitz-pi2; make a persistent copy of each
09:48 schestowitz-pi2; cull what's in the prior
09:48 schestowitz-pi2; or mark it somehow
09:48 schestowitz-pi2; i used to wonder, for years already, how I would program a clustering of related stories in the large pool
09:48 schestowitz-pi2; e.g. give me a list of all the stories about the new Fedora release
09:48 schestowitz-pi2; while not adding too much complexity, which raises cost of maintaining the "logic"
09:49 schestowitz-pi2; filters in quiterss were limited/limiting
09:49 schestowitz-pi2; the sooner I "solve" this problem, the less time I will waste doing it manually with human errors
09:49 techrights[sec]; I can take a look some time over the coming days
09:49 techrights[sec]; Using the -d option with rrrrr.py will reduce the amount of redundancy.
09:49 techrights[sec]; The difficulty would be assigning relevant key words for indexing.
09:57 schestowitz-pi2; the code you wrote is super-useful, I use it all the time
09:57 schestowitz-pi2; my own version of it serves another purpose
09:57 schestowitz-pi2; making the full html exposed, organised by feed, and ready to add with one mouse click (which triggers
09:57 schestowitz-pi2; a chain of actions)
09:57 schestowitz-pi2; my barriers atm (that take time):
09:57 schestowitz-pi2; trimming/editing the quoted text
09:57 schestowitz-pi2; trimming trash from the URL sometimes
09:57 schestowitz-pi2; trimming and other unwanted bits
09:57 schestowitz-pi2; changing headings to inside blockquotes
09:57 schestowitz-pi2; cluster related stories
09:57 schestowitz-pi2; avoid seeing (again) stuff I saw in prior days (usually up to a day earlier, as cutoff is 48 hrs)
09:57 schestowitz-pi2; those things can make a big difference to focus and productivity and can evolve over time (they already do)
10:02 techrights[sec]; ./rrrrr.py -d $(date -d '1 day ago' +'%F') -o x.html
10:02 techrights[sec]; Clustering is harder without using some massive LLM or something
10:02 techrights[sec]; That is because articles can be about something or very relevant to something
10:02 techrights[sec]; without actually naming that thing even once.
10:18 *DaemonFC has quit (Quit: Leaving)
10:46 schestowitz-pi2; http://news.tuxmachines.org/n/2023/03/04/Linux_Magazine_s_New_Issue.shtml
=> ↺ http://news.tuxmachines.org/n/2023/03/04/Linux_Magazine_s_New_Issue.shtml
10:46 -TechrightsBN/#boycottnovell-news.tuxmachines.org | Tux Machines Linux Magazine's New Issue
10:46 schestowitz-pi2; this was genertaed by the script
10:47 schestowitz-pi2; grep "date" 09.html | sort | sed -e "s/a hre/li><a hre/" -e "s/</a>/</a></h5></li>/"
10:48 schestowitz-pi2; chedking..
10:53 schestowitz-pi2; i see now
10:53 schestowitz-pi2; it would be good to use yours as the authrative list, wiuth culling
10:53 schestowitz-pi2; then the other file is just a cheat to get into full body
10:53 schestowitz-pi2; by scanning the same url
10:56 schestowitz-pi2; chedking..
10:56 schestowitz-pi2; i see now
10:56 schestowitz-pi2; it would be good to use yours as the authrative list, wiuth culling
10:56 schestowitz-pi2; then the other file is just a cheat to get into full body
10:56 techrights[sec]; It might be doable in the CSS.
10:56 techrights[sec]; Please have a try and then I can tweak it if necessary and after a few rounds
10:56 techrights[sec]; it should be reasonable on all screens concerned.
10:56 schestowitz-pi2; by scanning the same url
10:56 schestowitz-pi2; datestamps can be a plus
10:56 schestowitz-pi2; in Daily Links
10:56 schestowitz-pi2; if not distracting
10:56 schestowitz-pi2; can add style to , e.g. smaller or dimmer
10:56 schestowitz-pi2; span.date
10:56 schestowitz-pi2; let me have a go at this, biab
10:56 schestowitz-pi2; yes, the css
10:56 schestowitz-pi2; wanna add it or will i, via git?
10:56 schestowitz-pi2; kk
11:21 techrights[sec]; Seems a little complicated, visually. Perhaps it can be streamlined more?
11:21 techrights[sec]; The ovals XOR the bullets?
11:21 schestowitz-pi2; http://news.tuxmachines.org/n/2023/03/04/Linux_Magazine_s_New_Issue.shtml
=> ↺ http://news.tuxmachines.org/n/2023/03/04/Linux_Magazine_s_New_Issue.shtml
11:21 -TechrightsBN/#boycottnovell-news.tuxmachines.org | Tux Machines Linux Magazine's New Issue
11:21 schestowitz-pi2; froce-refresh/hard-refresh+crtrl+f5 typically
11:21 schestowitz-pi2; TR can inherit some similar styling, as those propagate onto there, too
11:22 schestowitz-pi2; many of the links won't have a date, only some will
11:22 schestowitz-pi2; I'm open to any changes you see fit
11:22 schestowitz-pi2; the selector is good enough for firefox esr and for falkon
11:22 schestowitz-pi2; did not check other browsers
11:22 *psydroid2 (~psydroid@cbcfptirpkfqa.irc) has joined #boycottnovell
11:26 schestowitz-pi2; rianne just saw this mockup
11:26 schestowitz-pi2; and said ok
11:26 schestowitz-pi2; but her links will likely have no date field
11:26 schestowitz-pi2; because her workflow is very different
11:26 schestowitz-pi2; she also uses some scripts i made
11:26 schestowitz-pi2; my current thinking is something crude and all-over-the-place, but
11:26 schestowitz-pi2; check the outline, put in tandem with full html file
11:26 schestowitz-pi2; so there is an option for each given link to get its 'expanded' version
11:26 schestowitz-pi2; this would solve several existing pitfalls in one fell swoop
11:27 techrights[sec]; The H5 within the LI really seems to mess up the layout
11:27 schestowitz-pi2; h5 in ls very old legacy of daily links
11:27 schestowitz-pi2; going liek 15 years back
11:27 schestowitz-pi2; *in LI
11:27 techrights[sec]; Can it be dropped?
11:27 techrights[sec]; ... at least in the new site(s) ?
11:29 schestowitz-pi2; changing it might imperil some scripts that currently scan for that
11:29 schestowitz-pi2; so there would be debt
11:29 techrights[sec]; Not necessarilly a loss in the tm-new. It would be useful to cull that
11:29 techrights[sec]; going forward, even if it is left for a while longer in tr-old
11:31 schestowitz-pi2; it might complicate the TM->TR workflow
11:31 schestowitz-pi2; so I'd say, let's do one step at a time
11:31 schestowitz-pi2; btw, cnet has many layoffs
11:31 schestowitz-pi2; the terrain for online news changes very fast
11:31 schestowitz-pi2; now with new "journalism" or "save journalism" laws/bills
11:31 schestowitz-pi2; and rianne and I spoke this morning about how awful it is to open "news sites"
11:31 schestowitz-pi2; having to deal with browser level and page-level popups, paywalls, cookies, JS (blockers being blocked etc.)
11:31 schestowitz-pi2; so to make the web readable a lot of "processing" is required
11:32 techrights[sec]; The more cruft which carries in from the one site to the other, the greater
11:32 techrights[sec]; the messs to clean up later ... technical debt as it is called. Yes, usability
11:32 techrights[sec]; is ignore on those sites, I'm not sure what their intentions are at all.
11:32 techrights[sec]; That makes any efforts towards usability on tm-new stand out that much more.
11:36 schestowitz-pi2; roy@vonick:~/rss-tools$ grep "date" 09.html | sort | sed -e "s/a hre/li><a hre/" -e "s/</a>/</a></h5></li>/" | head
11:36 schestowitz-pi2; 2023-03-02 Adam Young: Hello World in rust on ARM64 includes a lot of code
=>
11:36 -TechrightsBN/#boycottnovell-adam.younglogic.com | Hello World in rust on ARM64 includes a lot of code | Adam Young's Web Log
11:36 schestowitz-pi2; 2023-03-02 [Guest post] The NFT Hermés Case: Mainly Relevant for Large Collection of NFTs
=>
11:36 -TechrightsBN/#boycottnovell-ipkitten.blogspot.com | [Guest post] The NFT Herms Case: Mainly Relevant for Large Collection of NFTs - The IPKat
11:36 schestowitz-pi2; aye.
11:36 schestowitz-pi2; thinking or 'brain'storming
11:36 schestowitz-pi2; mostly the workflow
11:36 schestowitz-pi2; merged uses h5, so I don't want to try to solve too many problems at the same time
11:36 schestowitz-pi2; or my head won't work well
11:41 techrights[sec]; It's easy enough to cull the H5 elements at any stage but they should not
11:41 techrights[sec]; find their way into tm-new; search and replace works well at that level
11:41 techrights[sec]; | sort -u | sed -e 's|...|...|;' | head
11:43 schestowitz-pi2; atm on tm-new, for lists h5 is the 'default' style
11:43 schestowitz-pi2; sorry if i am quiet for a few mins, trying to work on the new workflow a bit
11:43 schestowitz-pi2; i can detect dupes against a db of all prior links in TR
11:43 schestowitz-pi2; there is a db like this on vonick
11:43 schestowitz-pi2; biab
12:07 schestowitz-pi2; lol! the CEO of Sirius has left
12:07 schestowitz-pi2; the company is in ruins
12:08 schestowitz-pi2; someone just told me
12:09 techrights[sec]; When will they file for bankruptcy?
12:25 schestowitz-pi2; 4 things happening all at once atm
12:25 schestowitz-pi2; dad on mumble
12:25 schestowitz-pi2; colleague contacting us (lots of info)
12:25 schestowitz-pi2; match across the road starting (loud)
12:25 schestowitz-pi2; and i try to work out the best way to digest news/links
12:25 schestowitz-pi2; rianne is lol'ing a lot (because of the news)
12:25 schestowitz-pi2; maybe i will stay up till 7pm to get ahead
12:30 techrights[sec]; ok
13:26 schestowitz-pi2; When social control media crashes and burns people will be back to speaking their native tongue, not "Algospeak" and "Hashtagish". Writing for fucking algorithms results in awful output.
13:28 schestowitz-pi2; pushed to TR git
13:28 schestowitz-pi2; as a sort of placeholder
13:28 schestowitz-pi2; i realise nano on tr-old is not good at unicode
13:28 schestowitz-pi2; it is saved ok but displayed poorly
13:29 schestowitz-pi2; now back to the links
13:29 schestowitz-pi2; am trying to sort of the pipeline
13:29 schestowitz-pi2; before wasting another week looking at two similar things in tandem :-)
13:34 schestowitz-pi2; i have a 'lazyweb' type of question: from what I recall the --keep option
13:34 schestowitz-pi2; peserves stuff seen before or some such, based on a database, though i think
13:34 schestowitz-pi2; the db trackswhat feeds need refreshing
13:34 schestowitz-pi2; with your latest commit, is is going to keep a "prior" version and then ensure
13:34 schestowitz-pi2; only "new since" are presented?
13:50 techrights[sec]; --keep is for when you load a "new" OPML file, any old feeds that were in the
13:50 techrights[sec]; old OPML file and are also in the new OPML file will have their state retained
13:50 techrights[sec]; So that way you don't end up re-checking feeds when updating the OPML file.
13:50 techrights[sec]; Say you rearrange the feeds and make a new hierarchy but have the exact same
13:50 techrights[sec]; feeds. They'd just be in a different order. Then use the --keep option
13:50 techrights[sec]; to avoid force reloading them all as you load the new structure via the OPML
13:50 techrights[sec]; file.
13:50 techrights[sec]; I'm not sure I'm explaining it well.
13:50 techrights[sec]; Normally when you pass an OPML file as an argumen, the states are overwritten.
13:50 techrights[sec]; With --keep the OPML file will be read and any feeds which were visited
13:50 techrights[sec]; before via the previous OPML file will keep their state but new feeds will
13:50 techrights[sec]; be initialized as new feeds.
13:54 schestowitz-pi2; this makes perfect sense
13:54 schestowitz-pi2; and was clear from your first paragraph
13:54 schestowitz-pi2; so it does not have a concept of "read"
13:54 schestowitz-pi2; the circles at top of nested branches took me a while to figure out
13:54 schestowitz-pi2; inc. their colour
13:54 schestowitz-pi2; and state does not refer to read/unread
13:54 schestowitz-pi2; one thing I can do in the processing is,
13:54 schestowitz-pi2; i can grep for date
13:54 schestowitz-pi2; e.g. show for anything past midnight (today)
13:54 schestowitz-pi2; i'll adapt to changes over time, but it's going in a very very good direction
13:54 schestowitz-pi2; i try to think where to place which windowto maximise utilisation
13:55 techrights[sec]; Right. Only the browser tracks read or unread status. The script can
13:55 techrights[sec]; only track fetched or not fetched.
13:55 techrights[sec]; I wonder if -d can be set to "today"
13:55 techrights[sec]; If so, then grep would not be needed.
13:59 techrights[sec]; ./rrrrr.py -d $(date +'%F') -o x.html
13:59 techrights[sec]; That should show only entries since midnight and nothing earlier
14:19 techrights[sec]; It just occurred to me that it might be possible to use the sqlite datbase
14:19 techrights[sec]; maintained by Firefox, but that'd be a whole other can of worms.
14:19 techrights[sec]; It'd be a project in and of itself to try that but I'm not sure of the utility
14:19 techrights[sec]; However, showing only entries since midnight would miss any entries the next
14:19 techrights[sec]; day which happened to show up between the time you check and the end of the day.\
14:19 techrights[sec]; So it would be best to use -1 day or leave the default, -2 days.
14:22 schestowitz-pi2; yes, i came to a similar conclusion
14:22 schestowitz-pi2; for dupe-checking i use a windows on the raspi
14:22 schestowitz-pi2; it doesn't take much effort
14:22 schestowitz-pi2; maybe I'll even widen the search of days and add colour/similar to indicate "recency"
14:27 schestowitz-pi2; (i am actually quite excited about this and am very grateful; it was effort well spent!)
15:01 *psydruid has quit (Ping timeout: 2m30s)
15:02 *psydruid (~psydruid@jevhxkzmtrbww.irc) has joined #boycottnovell
21:21 *psydroid2 has quit (connection closed)
21:21 *psydroid3 (~psydroid@cbcfptirpkfqa.irc) has joined #boycottnovell
21:29 *DaemonFC (~daemonfc@amrs5w4z676gy.irc) has joined #boycottnovell
22:29 *psydroid3 has quit (connection closed)
23:15 *{{THE_BORG}} (~{{__THE@freenode-oks.rp9.09la3j.IP) has joined #boycottnovell
23:26 *{{THE_BORG}} has quit (Connection closed)
23:28 *{{THE_BORG}} (~{{__THE@freenode-oks.rp9.09la3j.IP) has joined #boycottnovell
23:54 *{{THE_BORG}} has quit (Connection closed)
23:57 *{{THE_BORG}} (~{{__THE@freenode-oks.rp9.09la3j.IP) has joined #boycottnovell
=> back to Techrights (Main Index) This content has been proxied by September (ba2dc).Proxy Information
text/gemini;lang=en-GB