| 00:02:10 | | etnguyen03 quits [Quit: Konversation terminated!] |
| 00:13:37 | <nicolas17> | wtf |
| 00:13:53 | <nicolas17> | https://ntc.party/t/12845 kde.org is blocked in russia? |
| 00:36:42 | <steering> | So I'm looking at Njalla's website and see this: "We hide your real IP address behind one of ours. This means your online activity can’t instantly be seen by service providers, governments or others." |
| 00:36:50 | <steering> | >instantly |
| 00:36:53 | <steering> | at least they're honest :P |
| 00:41:30 | <steering> | also their pricing is ... kinda poopy |
| 00:44:11 | <pabs> | nicolas17: I think .ru just blocked Hetzner |
| 00:45:08 | <nicolas17> | they once blocked large chunks of digitalocean in an attempt to block telegram, and it affected KDE as a side effect, and KDE did some CDN workaround |
| 00:45:28 | <nicolas17> | I hadn't heard that KDE was down again due to Russia blocking Hetzner |
| 01:06:26 | | etnguyen03 (etnguyen03) joins |
| 01:06:39 | | nulldata quits [Read error: Connection reset by peer] |
| 01:07:39 | | nulldata (nulldata) joins |
| 01:32:13 | <nicolas17> | holy fuck |
| 01:32:18 | <nicolas17> | digitaloceanspaces.com is down due to a domain registrar issue |
| 01:33:08 | <nicolas17> | it's in clientHold status |
| 01:35:02 | | etnguyen03 quits [Client Quit] |
| 01:35:36 | <nicolas17> | https://status.digitalocean.com/incidents/jm44h02t22ck |
| 01:38:35 | <nicolas17> | did they forget to pay for renewal or what |
| 01:44:09 | <TheTechRobo> | is it just subdomains that are broken? |
| 01:44:19 | <TheTechRobo> | It correctly redirects to digitalocean.com for me. |
| 01:44:43 | <nicolas17> | that could be thanks to heavy caching |
| 01:45:22 | <nicolas17> | I believe the now-unavailable DNS record has a TTL of 24 hours |
| 01:46:51 | <nicolas17> | "dig digitaloceanspaces.com @a.gtld-servers.net." says NXDOMAIN |
| 01:54:00 | <@JAA> | Hmm, expiration date is 2028-02-23T21:30:06Z. |
| 01:55:25 | <nicolas17> | this seems like a situation that may need phone calls to resolve |
| 01:58:25 | <nicolas17> | I suspect it's something like https://www.backblaze.com/blog/recent-outages-why-we-accelerated-registry-changes/ |
| 01:59:51 | <@JAA> | Yeah |
| 02:07:41 | <nicolas17> | in other news |
| 02:08:03 | <nicolas17> | https://web.archive.org/web/20241121014848/https://www.backblaze.com/blog/rate-limiting-policy/ |
| 02:08:18 | <nicolas17> | looks like this angered many users |
| 02:08:41 | <nicolas17> | so they backpedaled |
| 02:08:42 | <nicolas17> | https://web.archive.org/web/20241129020734/https://www.backblaze.com/blog/rate-limiting-policy/ |
| 02:11:08 | | etnguyen03 (etnguyen03) joins |
| 03:01:39 | | etnguyen03 quits [Remote host closed the connection] |
| 03:05:52 | | linuxgemini (linuxgemini) joins |
| 03:29:25 | <nicolas17> | JAA: looks like this is the third networksolutions fuckup this year |
| 03:29:27 | <nicolas17> | https://cdn.discordapp.com/attachments/286612533757083648/1311896044916834304/IMG_1715.png?ex=674a85e6&is=67493466&hm=943533f08188e334d8df3256e6bf24f5b1c2522e0f78950fae4ade3e51ac49e2& |
| 03:30:53 | <@JAA> | Oh right |
| 03:45:37 | <nicolas17> | https://x.com/henet/status/1808953880404787288 x_x |
| 03:45:38 | <eggdrop> | nitter: https://nitter.lucabased.xyz/henet/status/1808953880404787288 |
| 03:51:21 | <nicolas17> | DigitalOcean CEO: https://x.com/paddix/status/1862338275253301700 |
| 03:51:21 | <eggdrop> | nitter: https://nitter.lucabased.xyz/paddix/status/1862338275253301700 |
| 03:54:41 | <pabs> | from the HN Hetzner threads, an unmetered provider: https://microtronixdc.com/ |
| 03:57:17 | <pabs> | Ottoville, Ohio |
| 04:17:39 | <imer> | > E5-2680v3 |
| 04:17:39 | <imer> | ah yes, ewaste |
| 04:24:30 | <datechnoman> | Xeon E5-2699v4? |
| 04:24:44 | <datechnoman> | Pretty old still I guess |
| 04:27:47 | <nicolas17> | seems digitaloceanspaces is fixed |
| 04:28:20 | <imer> | bit confused by these providers that are just seemingly stuck on ~10 year old hardware, like I get it, it was cheap and cheerful like 5-7 years ago, but why is there nothing remotely current |
| 04:28:55 | <datechnoman> | Milking every last dollar out of them before they go in the bin |
| 04:29:12 | <datechnoman> | Buying new hardware costs $ where keeping old hardware until it dies is free :P |
| 04:29:35 | <nicolas17> | AWS doesn't even reduce prices on their ewaste servers |
| 04:29:58 | <imer> | but you can get new hardware at least |
| 04:30:52 | <imer> | v5 is the newest they have which launched in 2015 (q4.. but still) |
| 04:31:06 | <imer> | iunno, just odd to me |
| 04:33:00 | <imer> | I know of a place that followed that strategy for a good while, but they gave up like two? years back since it just doesnt perform (+ high electricity prices over in europe made it less attractive to run I guess) |
| 04:33:29 | <imer> | and by gave up I mean they added newer hardware, think you can still buy the old crap if you really want to |
| 04:33:48 | <imer> | (which is fine.. tons of memory for cheap) |
| 04:34:01 | <imer> | </rant> |
| 04:38:01 | | Exorcism quits [Quit: Ping timeout (120 seconds)] |
| 04:38:21 | | Exorcism (exorcism) joins |
| 04:54:57 | <@JAA> | Not a single dedi with ECC or more than one disk either? lol |
| 04:55:27 | <datechnoman> | There is a reason its so cheap :o |
| 04:55:49 | <@JAA> | And it isn't even very cheap. :-D |
| 04:56:11 | <@JAA> | Well, I guess for the US, it is. |
| 04:57:44 | <@JAA> | But even Kimsufi servers (except for the one ridiculously cheap one) have two or more disks and ECC RAM... |
| 04:58:11 | <@JAA> | And the hardware on those is similarly ancient. |
| 04:58:39 | <nicolas17> | >more than one disk |
| 04:58:41 | <nicolas17> | what |
| 04:59:01 | <@JAA> | Dual SSD for a RAID1 so it doesn't fall over as easily? |
| 04:59:33 | <nicolas17> | are they seriously renting servers with a single physical disk |
| 05:00:08 | <@JAA> | Basically all their offers are '1x 1TB HDD or 120GB SSD'. |
| 05:00:23 | <nicolas17> | x_x |
| 05:00:26 | <@JAA> | One has a 512 GB SSD instead. |
| 05:00:47 | <@JAA> | This is a theme I've seen at many cheapish US dedi providers and can't make sense of. |
| 05:01:10 | <nicolas17> | a friend told me he was once contacted on linkedin by a company that wanted to sell "cloud" hosting and were interested in his OpenStack experience |
| 05:01:18 | <nicolas17> | so he asked what hardware they had |
| 05:01:40 | <nicolas17> | and they described the specs of their server |
| 05:01:47 | <TheTechRobo> | JAA: If you want redundancy, rent a second server, duh. :P |
| 05:02:21 | <nicolas17> | his reply was something like "I'm not sure how to say this politely but I'll try. There is nothing serious about trying to sell cloud services having one(1) server. I wish you the best of luck." |
| 05:02:56 | <@JAA> | lol |
| 05:03:06 | <@JAA> | Literally xkcd 908, huh? |
| 05:05:10 | <nicolas17> | afaik a lot of cloud management software like openstack and kubernetes requires a minimum of 3 servers *for the control plane* |
| 05:18:23 | <katia> | kubernet |
| 06:07:12 | <nukke> | if your S3 storage hasn't been hacked at least once, are you really using The Cloud correctly?? 🤔 |
| 06:07:22 | <nukke> | s/storage/bucket |
| 06:40:30 | <monoxane> | nicholas17 kube just needs an odd number, i run a single node and its fine |
| 07:14:40 | | Barto quits [Ping timeout: 260 seconds] |
| 07:52:01 | | pixel (pixel) joins |
| 08:09:37 | | flotwig_ joins |
| 08:10:40 | | flotwig quits [Ping timeout: 260 seconds] |
| 08:10:40 | | flotwig_ is now known as flotwig |
| 08:34:34 | <katia> | monoxane: With etcd? |
| 08:35:20 | <monoxane> | yea |
| 08:35:29 | <monoxane> | you can run single node etcd no problems |
| 08:35:46 | <monoxane> | it literally only cares that you have a non-zero odd number |
| 08:35:57 | <monoxane> | and that you're within like 50ms latency between all control plane nodes |
| 08:57:20 | | M60_ quits [Ping timeout: 260 seconds] |
| 08:57:59 | | M60_ joins |
| 09:06:05 | | ThreeHM quits [Ping timeout: 260 seconds] |
| 09:08:02 | | ThreeHM (ThreeHeadedMonkey) joins |
| 09:08:45 | | i_have_n0_idea (i_have_n0_idea) joins |
| 09:11:47 | <szczot3k> | 'requires' not always mean 'will not run without', but a lot of times means 'we don't recommend, if things break, you're fucked' |
| 09:12:37 | <szczot3k> | The same way you can do proxmox 'cluster' with 2 nodes, but without a third quorum node, it won't be a HA cluster, just two servers that can migrate vms between each other |
| 09:15:18 | | nulldata quits [Quit: So long and thanks for all the fish!] |
| 09:16:11 | | nulldata (nulldata) joins |
| 09:32:43 | | Barto (Barto) joins |
| 09:33:35 | | Froxcey quits [Remote host closed the connection] |
| 09:46:51 | | Froxcey (Froxcey) joins |
| 09:48:40 | | mgrytbak quits [Ping timeout: 260 seconds] |
| 09:50:12 | | Gadelhas56283 joins |
| 09:51:01 | | Froxcey quits [Remote host closed the connection] |
| 09:51:07 | | Froxcey (Froxcey) joins |
| 09:53:55 | | Gadelhas5628 quits [Ping timeout: 260 seconds] |
| 09:53:55 | | Gadelhas56283 is now known as Gadelhas5628 |
| 09:56:19 | | mgrytbak joins |
| 10:00:18 | | driib quits [Quit: The Lounge - https://thelounge.chat] |
| 10:00:49 | | driib (driib) joins |
| 10:22:05 | | Froxcey quits [Remote host closed the connection] |
| 10:23:00 | | Froxcey (Froxcey) joins |
| 10:27:45 | | Froxcey quits [Ping timeout: 260 seconds] |
| 10:33:41 | <monoxane> | it all depends on your risk profile |
| 10:34:40 | <monoxane> | but it is semi-officially supported, the docs for kube say "One or more machines running a deb/rpm-compatible Linux OS; for example: Ubuntu or CentOS." |
| 10:40:33 | <szczot3k> | Mostly thinking of enterprise stuff, where 'required' = 'even though you pay for our support, if you do that, we won't help you' |
| 10:41:10 | <monoxane> | yea I get ya |
| 10:49:55 | <monoxane> | I think the only real valid use case for single control plane kube in production is those weird 2+1 hardened edge chassis dell and supermicro make, with 2 beefy nodes and 1 meh witness node |
| 10:50:15 | <monoxane> | but even saying that you can just run 3x cp with the 2 beefy ones also acting as workers |
| 10:51:51 | | Froxcey (Froxcey) joins |
| 10:56:20 | | Froxcey quits [Ping timeout: 260 seconds] |
| 11:06:44 | | Froxcey (Froxcey) joins |
| 11:11:30 | | Froxcey quits [Ping timeout: 260 seconds] |
| 11:21:53 | | Froxcey (Froxcey) joins |
| 11:26:40 | | Froxcey quits [Ping timeout: 260 seconds] |
| 11:36:36 | | Froxcey (Froxcey) joins |
| 11:41:15 | | Froxcey quits [Ping timeout: 260 seconds] |
| 11:54:09 | | Froxcey (Froxcey) joins |
| 11:58:45 | | Froxcey quits [Ping timeout: 260 seconds] |
| 12:00:04 | | Bleo182600722719623 quits [Quit: The Lounge - https://thelounge.chat] |
| 12:01:27 | | Froxcey (Froxcey) joins |
| 12:02:54 | | Bleo182600722719623 joins |
| 12:06:20 | | Froxcey quits [Ping timeout: 260 seconds] |
| 12:11:08 | | Froxcey (Froxcey) joins |
| 12:43:20 | | Froxcey quits [Remote host closed the connection] |
| 12:43:26 | | Froxcey (Froxcey) joins |
| 13:19:34 | | sralracer (sralracer) joins |
| 14:16:30 | <that_lurker> | -feed- Finland, Sweden complete repairs on Baltic Sea cables https://yle.fi/a/74-20128140 https://news.ycombinator.com/item?id=42273288 |
| 14:33:57 | | Froxcey quits [Remote host closed the connection] |
| 14:45:55 | | Froxcey (Froxcey) joins |
| 14:48:50 | <immibis> | nicolas17: what if they want to expand to more servers as they get more customers? and remembering the word cloud doesn't mean anything to begin with... |
| 14:48:55 | | Froxcey quits [Read error: Connection reset by peer] |
| 14:48:59 | | Froxcey (Froxcey) joins |
| 14:52:26 | <Fijxu|m> | Huh |
| 14:52:31 | <Fijxu|m> | How do you repair a sea cable |
| 14:52:36 | <Fijxu|m> | If it's fiber |
| 14:52:43 | <Fijxu|m> | I wonder how |
| 14:53:23 | <immibis> | Somehow attach a rope onto both ends, pull them up and splice all the fibers and other bits? It's not simple and probably takes them many days to do the splice even knowing exactly what the cable is made of and having all the repair materials on hand. |
| 14:53:45 | | Froxcey quits [Ping timeout: 260 seconds] |
| 14:54:26 | <immibis> | haven't got the slightest idea how they attach onto a cable end to pull it up from the sea, either. Underwater robots? |
| 14:54:47 | <immibis> | human diving is only practical down to a certain depth. maybe it's above that depth or not. |
| 14:56:05 | <immibis> | even finding the cable in the vast expanse of water seems difficult |
| 15:01:42 | | Froxcey (Froxcey) joins |
| 15:51:38 | <that_lurker> | there are some good youtube vides on how it's done |
| 15:52:40 | <that_lurker> | Theres is also an image of the remote-controlled tracked vehicle that was used on the repair https://yle.fi/a/74-20127099 |
| 15:54:28 | | Froxcey quits [Remote host closed the connection] |
| 15:55:06 | <nicolas17> | Fijxu|m: https://www.theverge.com/c/24070570/internet-cables-undersea-deep-repair-ships |
| 16:03:59 | | Froxcey (Froxcey) joins |
| 16:18:26 | <kiska> | !reminders |
| 16:20:35 | <kiska> | !remindme 72h downsample influxdb |
| 16:20:35 | <eggdrop> | [remind] ok, i'll remind you at 2024-12-02T16:20:35Z |
| 16:22:23 | <kiska> | !remindme 1mo downsample influxdb |
| 16:22:23 | <eggdrop> | [remind] ok, i'll remind you at 2024-12-29T16:22:23Z |
| 16:22:26 | <kiska> | !remindme 2mo downsample influxdb |
| 16:22:27 | <eggdrop> | [remind] ok, i'll remind you at 2025-01-29T16:22:27Z |
| 16:22:29 | <kiska> | !remindme 3mo downsample influxdb |
| 16:22:30 | <eggdrop> | [remind] ok, i'll remind you at 2025-02-28T16:22:29Z |
| 16:22:32 | <kiska> | !remindme 4mo downsample influxdb |
| 16:22:32 | <eggdrop> | [remind] ok, i'll remind you at 2025-03-29T16:22:32Z |
| 16:22:34 | <kiska> | !remindme 5mo downsample influxdb |
| 16:22:35 | <eggdrop> | [remind] ok, i'll remind you at 2025-04-29T16:22:35Z |
| 16:22:37 | <kiska> | !remindme 6mo downsample influxdb |
| 16:22:38 | <eggdrop> | [remind] ok, i'll remind you at 2025-05-29T16:22:37Z |
| 16:22:40 | <kiska> | !remindme 7mo downsample influxdb |
| 16:22:40 | <eggdrop> | [remind] ok, i'll remind you at 2025-06-29T16:22:40Z |
| 16:22:42 | <kiska> | !remindme 8mo downsample influxdb |
| 16:22:43 | <eggdrop> | [remind] ok, i'll remind you at 2025-07-29T16:22:42Z |
| 16:22:46 | <kiska> | !remindme 9mo downsample influxdb |
| 16:22:46 | <eggdrop> | [remind] ok, i'll remind you at 2025-08-29T16:22:46Z |
| 16:22:48 | <kiska> | !remindme 20mo downsample influxdb |
| 16:22:50 | <eggdrop> | [remind] ok, i'll remind you at 2026-07-29T16:22:48Z |
| 16:22:52 | <kiska> | !remindme 10mo downsample influxdb |
| 16:22:52 | <eggdrop> | [remind] ok, i'll remind you at 2025-09-29T16:22:52Z |
| 16:22:55 | <kiska> | !remindme 11mo downsample influxdb |
| 16:22:55 | <eggdrop> | [remind] ok, i'll remind you at 2025-10-29T16:22:55Z |
| 16:22:57 | <kiska> | !remindme 12mo downsample influxdb |
| 16:22:58 | <eggdrop> | [remind] ok, i'll remind you at 2025-11-29T16:22:57Z |
| 16:26:27 | | th3z0l4 joins |
| 16:30:35 | | th3z0l4_ quits [Ping timeout: 260 seconds] |
| 16:34:26 | <nulldata> | !remindme 13mo kiska upsample influxdb |
| 16:34:28 | <eggdrop> | [remind] ok, i'll remind you at 2025-12-29T16:34:26Z |
| 16:35:13 | <kiska> | lol |
| 16:35:14 | <kiska> | :D |
| 16:41:56 | | Froxcey quits [Remote host closed the connection] |
| 16:49:28 | | nulldata quits [Quit: Ping timeout (120 seconds)] |
| 16:50:26 | | nulldata (nulldata) joins |
| 16:55:23 | | Froxcey (Froxcey) joins |
| 16:57:13 | | Froxcey quits [Remote host closed the connection] |
| 16:57:14 | | Froxcey (Froxcey) joins |
| 17:26:14 | <szczot3k> | !remindme 72h don't do anything with influxdb |
| 17:26:15 | <eggdrop> | [remind] ok, i'll remind you at 2024-12-02T17:26:15Z |
| 17:28:02 | <kiska> | Oh? |
| 17:28:08 | <kiska> | szczot3k: And why is that? |
| 17:28:34 | <szczot3k> | kiska: mostly because I don't run influx anywhere in my homelab |
| 17:29:11 | <kiska> | I run it for https://grafana3.kiska.pw/d/000000/archiveteam-tracker-stats?orgId=1 |
| 17:29:40 | <szczot3k> | And I'm just joking with the reminder :D |
| 17:29:55 | <kiska> | Lets just say that without downsampling, I would be looking at about 1T of samples |
| 17:30:02 | <szczot3k> | Todo ETA: ∞ years |
| 17:30:03 | <szczot3k> | great |
| 17:30:21 | <kiska> | I think it defaults to the urls project |
| 17:30:33 | <kiska> | Yes it does :D |
| 17:35:38 | <szczot3k> | askfm also has infinite eta |
| 17:36:03 | <kiska> | Yeah :D |
| 17:48:50 | | Notrealname1234 (Notrealname1234) joins |
| 17:50:22 | <szczot3k> | Been wondering - what are the icon some people have on the tracker? |
| 17:51:18 | <nicolas17> | means they're running the warrior VM |
| 17:51:47 | <szczot3k> | Got it |
| 17:53:31 | | Notrealname1234 quits [Client Quit] |
| 17:53:41 | | Froxcey quits [Remote host closed the connection] |
| 17:59:53 | | AleX-1337 joins |
| 18:02:32 | <AleX-1337> | What are those restricted items on the Internet Archive, like this (https://archive.org/details/archiveteam_urls_20210709202713_6e2f5350) or these (https://archive.org/details/archiveteam_googleplus) ? |
| 18:05:27 | <@arkiver> | AleX-1337: the web page saved in them are only accessible through the Wayback Machine |
| 18:05:51 | <AleX-1337> | So those are the webpages "internal" files? |
| 18:06:19 | <nicolas17> | you can't download the original WARC files |
| 18:06:45 | <nicolas17> | but you can usually see the contents via web.archive.org |
| 18:07:33 | <AleX-1337> | Is there any way to see which websites each restricted WARC item is? |
| 18:08:00 | <nicolas17> | the .cdx file maybe |
| 18:08:07 | <AleX-1337> | thank you |
| 18:08:16 | | AleX-1337 quits [Client Quit] |
| 18:08:18 | <nicolas17> | I see there's two cdx files and one of them is restricted |
| 18:14:14 | | Froxcey (Froxcey) joins |
| 18:15:25 | | Froxcey_ joins |
| 18:15:25 | | Froxcey quits [Read error: Connection reset by peer] |
| 18:17:45 | | Froxcey (Froxcey) joins |
| 18:17:45 | | Froxcey_ quits [Read error: Connection reset by peer] |
| 18:22:35 | | Froxcey quits [Ping timeout: 260 seconds] |
| 18:45:09 | | Froxcey (Froxcey) joins |
| 18:46:00 | | Froxcey quits [Remote host closed the connection] |
| 19:02:32 | | VerifiedJ9 (VerifiedJ) joins |
| 19:14:10 | | Froxcey (Froxcey) joins |
| 19:18:35 | | Froxcey quits [Ping timeout: 260 seconds] |
| 20:05:41 | <@JAA> | I see too much on-topic chat in the off-topic channel again. |
| 20:05:47 | | klaffty quits [Quit: klaffty] |
| 20:11:24 | <@OrIdow6> | And too little off topic chat in the on topic channel, let's liven it up!!! |
| 20:56:04 | <immibis> | makes sense when people keep getting complaints in the on topic channel for being too off topic if it's not official archive team business |
| 21:03:24 | | Froxcey (Froxcey) joins |
| 21:20:01 | | Froxcey quits [Remote host closed the connection] |
| 21:27:45 | | Froxcey (Froxcey) joins |
| 21:28:24 | | Froxcey quits [Remote host closed the connection] |
| 21:28:30 | | Froxcey (Froxcey) joins |
| 21:40:21 | <szczot3k> | I just love that AT has three 'main' channels |
| 21:41:13 | <szczot3k> | announcements that everyone can write on, bs for official, but not announcement, and ot, for also talking about AT |
| 21:42:29 | <@OrIdow6> | Apropos of editing the wiki I am curious what the history of the channels is |
| 21:43:01 | <@OrIdow6> | I remember seeing godane say somewhere that originally it was just #at and then #bs was created because he was talking too much |
| 21:43:12 | <@OrIdow6> | Then people taleked too much in #bs so #ot was craetde |
| 21:43:44 | <@OrIdow6> | Is it nonobvious szczot3k? I'm pretty used to it after 5 years or however much it's been but IDK how it feels to newcomers |
| 21:43:55 | <@OrIdow6> | The tripartite divison and what to say where |
| 21:44:14 | <szczot3k> | Getting into #archiveteam, and getting yelled GO TO -BS isn't fun |
| 21:44:32 | <nicolas17> | yeah one would expect #archiveteam to be the general chat and something else for announcements |
| 21:45:07 | <szczot3k> | If we really want an 'announcment' channel, it should have +m |
| 21:45:52 | <szczot3k> | One could also set a welcome notice + /topic that yells that it's not the place you ask newbie questions |
| 21:46:02 | <szczot3k> | But yeah, one would expect #at is the main channel |
| 21:46:34 | <szczot3k> | I didn't care that someone asked me to move to -bs, mostly because it's not my first time getting yelled to do so at #at, joined couple of years back |
| 21:47:35 | <joepie91> | it's been a topic of confusion and complaint for as long as I have been in the IRC channels, which is a very long time |
| 21:47:44 | <szczot3k> | #(name) as the main channel is kinda the norm, examples from my chanlist include - #gentoo, #hackint, #dn42, #tor |
| 21:48:48 | <@OrIdow6> | The main reason I can think of to have the "announcements" channel also the "newcomers" channel is that if it's someone saying that a site is shutting down, if they put a single line in an discussion channel it might just get lost |
| 21:48:58 | <@OrIdow6> | joepie91: Do you have any old wisdom to share with us |
| 21:49:24 | <szczot3k> | #dn42 - the 'main one', where everything at least remotely on topic (if it's about a computer, or network it's probably ok) is ok, and #dn42-social where anything offtopic is fine (rarely used to be fair) |
| 21:49:25 | <@OrIdow6> | szczot3k: You mean, the opposite of what we're doing is the norm? |
| 21:49:36 | <@OrIdow6> | Ahh, 1 second too late |
| 21:49:52 | <szczot3k> | There's also #dn42-antisocial where you're banned from talking, because it's moderated, and nobody has +v there |
| 21:50:19 | <@OrIdow6> | Our equivalent of that is #archivebot where you get drowned out by bots talking to each otpher |
| 21:50:38 | <szczot3k> | Sure, but #ab is specifically for bot talk |
| 21:51:02 | <joepie91> | OrIdow6: the "making sure things do not get lost" is the only plausible reason I have ever heard but it does not really explain the -bs vs -ot thing, and personally I would not be surprised if we're losing potential contributors over the often rather harsh reception in main |
| 21:51:30 | <joepie91> | don't really have any wisdom besides "I don't think this is the optimal configuration and I wish we would work on finding a better one" |
| 21:51:35 | <szczot3k> | You don't go to #ab to talk to people. But if you're not remotely familiar with "archiveteam IRC culture", you'll go to #at, and get yelled at after you type 5 messages |
| 21:51:56 | <joepie91> | but it's also not something that I can afford to spend very many braincycles on arguing about, so :) |
| 21:53:00 | <szczot3k> | You might go to #ab to ask about getting some AB god to archive some site you like with AB, but it'd be fair if it's drowned in the bot spam |
| 21:53:27 | <@OrIdow6> | joepie91: The fact that it's the only plausible reason you've heard does make me think it's the only plausible reason period |
| 21:54:43 | <@OrIdow6> | And I don't doubt it, go rest :) |
| 21:54:52 | <szczot3k> | Imma be honest. I tried to get into archiveteam 'community' a year or two before, but after getting yelled at on #at, I backed down, and just remained in my familiar #dn42-* |
| 21:55:21 | <@OrIdow6> | szczot3k: I'm sorry for that, very plausible |
| 21:55:43 | <@OrIdow6> | FWIW (time for me to get banned) I only joined after finding out JS was no longer active here |
| 21:56:08 | <@OrIdow6> | I also rememeber putting super put off by this https://wiki.archiveteam.org/index.php/Archiveteam:IRC#Special_Archive_Team_IRC_rules which, uh, does accurately represent the cutlure |
| 21:56:47 | <@OrIdow6> | And yes I realize I'm a +o and very much a part of this culture now |
| 21:57:26 | <joepie91> | I can see why those special rules exist given the history of archiveteam, I'm not convinced they are still as relevant |
| 21:57:45 | <joepie91> | archiveteam and the internet archive are not as unfamiliar a concept to people as they used to be |
| 21:57:56 | <szczot3k> | Some of them sound bad, but really isn't - like the 'search it yourself', controversial but... fair |
| 21:58:20 | <@OrIdow6> | Maybe if it was better phrased it'd be better, it feels like a list of orders |
| 21:58:58 | <joepie91> | it used to be a big problem with folks totally new to IA stuff coming in at a high rate and Demanding Answers and that's probably where that comes from |
| 21:59:06 | <szczot3k> | Some of them are also an artifact from the past - Don't let your IRC client flood the channels with join/leave notifications due to your unstable connection. |
| 21:59:31 | <steering> | uhhh... not really |
| 21:59:46 | <steering> | still lots of that around irc |
| 21:59:57 | <szczot3k> | Configure your smartfilter in a good way, or hide join/leave messages |
| 22:00:59 | <joepie91> | that's a bit oversimplified, szczot3k; archiveteam has always involved a lot of short-term volunteers who may not be familiar with IRC or its tools at all, and who come here just because they read about an archival project in the press |
| 22:01:01 | | BlueMaxima joins |
| 22:01:26 | <joepie91> | from that perspective, "just" hide join/leave messages can be a significant barrier |
| 22:02:05 | <joepie91> | it's kind of complicated in that archiveteam involves both IRC die-hards and people who have no knowledge of or interest in IRC and that somehow needs to be made to work together |
| 22:02:36 | <joepie91> | in what has at least historically often been high-pressure situations |
| 22:03:19 | <szczot3k> | As with many rules - a double edged sword. Why should someone with a shitty wireless connection at home not be allowed to participate? Many rural areas just can't do better, and you're using a wireless connection, which drops. But at the same time you might have a whole cluster in a remote part of the word, ready to archive. |
| 22:05:06 | <szczot3k> | I don't think I've ever seen a rule such as this one on any irc community. Sure, seen some temporary +b/kicks for really bad cases, but if it's not a case of badly misconfigured bouncer, and your connection just does that... I'm not sure |
| 22:06:19 | <joepie91> | I'm not necessarily in support of these rules, to be clear :) |
| 22:06:42 | <joepie91> | just saying that the 'easy answers' for IRC aren't necessarily the right ones here either |
| 22:07:40 | <@OrIdow6> | BTW I like how the section in Archiveteam:IRC "I can't wait; I need immediate attention. Who's in charge?" is "answered" by a link to this useless page https://wiki.archiveteam.org/index.php/Who_We_Are |
| 22:09:26 | <@OrIdow6> | Think I'll try to edit down the "Special Archive Team IRC rules" section, if anyone has any suggestions be my guest/edit my edits but it sounds like I was in a minority inbeing put off by this specifically |
| 22:09:35 | <steering> | most of the big channels i'm in literally have bots auto-banning anyone who joins/parts too many times. |
| 22:10:16 | <szczot3k> | the only 'big' channel I'm really active is #dn42-* |
| 22:10:25 | <szczot3k> | So I might not be the best example |
| 22:11:31 | <steering> | -!- mode/#debian [+b *!*@user/aaii$##fix_your_connection] by debchange |
| 22:11:51 | <szczot3k> | in AT there's also weird stuff like... nobody really knowing (at least publically) who runs one of the more important pieces of software, which is the tracker, and it being closed sourced, and 'mostly broken' (as per wiki) |
| 22:11:52 | <steering> | (although yes, I agree, if it bothers you just hide join/part/quit) |
| 22:12:29 | <joepie91> | OrIdow6: selection bias and all that :) I don't think we can usefully conclude how many people have bounced off on this since most will logically simply not be here |
| 22:12:55 | <joepie91> | I would personally be in favour of revising those rules but that's just my opinion |
| 22:13:29 | <@OrIdow6> | joepie91: Thanks for the reassurance :) And really good point |
| 22:13:48 | <@OrIdow6> | Does make me sympathetic to those "why are you cancelling your subscription?" questions |
| 22:14:14 | <joepie91> | yeah |
| 22:15:32 | <steering> | OrIdow6: I mean, there's a reason I didn't stick around or talk much before fireonlive passed, I doubt you're in the minority :P |
| 22:15:36 | <szczot3k> | On the tracker topic - were there any tries on making it actually distributed, so one of the main projects can't just die randomly, and it could be run by several people? Don't hurt me too much though, don't know why it's that way |
| 22:16:07 | <joepie91> | having worked on distributed stuff, my first response is: oof |
| 22:16:21 | <joepie91> | I don't know if anyone tried but I imagine that it is going to be extremely challenging |
| 22:16:22 | <steering> | I think overall they all make sense as rules, but are very off-putting in the way they're worded |
| 22:16:45 | <joepie91> | managing consistent state is basically the problem in distributed programming, and the tracker is nothing but state :) |
| 22:17:22 | <szczot3k> | joepie91 I'm not saying tor consensus level of distributed, but better than 'someone runs the service, on some hetzner servers, the service has some admin panel, which is half broken' |
| 22:18:07 | <joepie91> | yeah I'm just talking about distributed, not decentralized, ie. just the idea of splitting it across multiple services - that already introduces the state management issues |
| 22:18:10 | <joepie91> | decentralized would be even harder |
| 22:18:50 | <joepie91> | if you ask me, the first thing that should be taught about building distributed systems is "don't, unless you have no other options" |
| 22:18:57 | <joepie91> | and I would try to solve this as a governance problem first |
| 22:19:59 | <joepie91> | for example: the problem could probably be significantly reduced by having multiple people run trackers, with a project being hosted on the tracker of whoever has time to manage it at that time, instead of having a single instance for all projects. but each project would then still live on one specific tracker, to sidestep distributed state issues |
| 22:20:18 | <szczot3k> | Sure, that's a step forward |
| 22:20:18 | <joepie91> | that would require zero technical change on the tracker side, only a small change in the warrior really |
| 22:20:53 | <szczot3k> | Having the tracker be a black box (closed source code) is also a challenge |
| 22:20:54 | <joepie91> | in this case though, I suspect the maintenance of the codebase may be a bigger problem |
| 22:21:24 | <@OrIdow6> | Yeah this is not really a technical problem |
| 22:22:11 | <joepie91> | huh. I could swear that was open-source before |
| 22:22:16 | <szczot3k> | It was |
| 22:22:26 | <szczot3k> | (as per wiki) |
| 22:22:49 | <szczot3k> | Sometime in the late 2010s the open-source tracker was gradually replaced with the proprietary one. Then or in the early 2020s backfeed and multi-item support was added. As of 2023 most of the admin functionality is broken, as far as I know; everything but setting the minimum version and the rate limit is done with non-public methods by the tracker |
| 22:22:49 | <szczot3k> | admins. |
| 22:23:03 | <szczot3k> | Reading that wiki page yesterday I was like... It's a mess |
| 22:23:09 | <joepie91> | that's weird. why? |
| 22:23:21 | <@OrIdow6> | (Full disclaimar, I wrote that section, but it is 100% accurate AFAIK) |
| 22:24:06 | <szczot3k> | The wiki page tl;dr is "I don't know who runs it, what it runs, how it's run, why does it work, but it somewhat does, mostly" |
| 22:24:08 | <@OrIdow6> | I think the page says somewhere else - from what I've been told, I don't have access to it, the production tracker *is* the open source one, just with tons and tons of wrapper scripts around it |
| 22:24:31 | <@OrIdow6> | That effectively change it to a different piece of software even if the core is still there |
| 22:24:41 | <joepie91> | the 'why' is mostly about that change, like, when did the "everything is open-source so it can be replicated" thing go away and why? |
| 22:24:54 | <joepie91> | it's not consistent with past rationales |
| 22:25:54 | <szczot3k> | When I heard about the askfm's problems because of the tracker I was like "huh, wonder if I can help the cause" and with every words on this wiki I was like '...huh' |
| 22:26:28 | <imer> | My uneducated guess would be quick fixes piling up since there is never enough time to sort out long term issues and now its infeasable to do so |
| 22:27:05 | <imer> | But I know nothing, so could be something else entirely, I doubt there's any "malice" involved though |
| 22:27:19 | <@arkiver> | problems can happen |
| 22:27:22 | <@arkiver> | sometimes IA is down |
| 22:27:39 | <@arkiver> | very big companies sometimes have problems, though shorter usually |
| 22:27:59 | <@arkiver> | ... not trying to say "it is fine this happens", but "it happens sometimes..." |
| 22:28:08 | <joepie91> | I mean, I can totally see why it would happen, I'm just wondering why that apparently didn't get caught and corrected as "hey this isn't how it should be" |
| 22:28:20 | <joepie91> | as this used to be a very strong point in AT |
| 22:28:42 | <joepie91> | (this is process debugging brain activating, basically) |
| 22:28:51 | <@arkiver> | err i should read more backlog here, i thought we were talking about askfm project being somewhat down the last day |
| 22:29:15 | <szczot3k> | Nope, talking about IRC policy, and tracker being closed source |
| 22:29:19 | <@OrIdow6> | No we're plotting a revolution |
| 22:29:24 | <@arkiver> | right |
| 22:29:25 | <joepie91> | arkiver: no, context is the new tracker codebase apparently not being open-source anymore |
| 22:29:31 | <@OrIdow6> | :P |
| 22:29:36 | <joepie91> | and this basically blocking the idea of 'reducing central points of failure by running more' |
| 22:29:45 | <@OrIdow6> | Gonna hack into the tracker and release hte source to wikileaks |
| 22:30:06 | <szczot3k> | If we ever get to the revolution I wanna be known as one of the instigators |
| 22:30:26 | <szczot3k> | OrIdow6 nah, just #archivebot it then |
| 22:30:36 | <joepie91> | not one of those revolutions where things burn down please, that would not be consistent with archiveteam objectives |
| 22:30:37 | <joepie91> | :p |
| 22:30:50 | <@arkiver> | burn it all down! |
| 22:30:52 | <szczot3k> | And if the revolution fails, I wasn't there |
| 22:31:04 | <szczot3k> | Don't put my name in the history books then. |
| 22:32:48 | <szczot3k> | From what I see, every other piece of code is written in a way that supports the idea of having people run their own instances to help, except the tracker |
| 22:33:00 | <nicolas17> | Stripe Black Friday stats https://bfcm.stripe.com/ |
| 22:33:06 | <nicolas17> | "Perhaps this is not obvious: there is an actual video being broadcast live of a real physical machine, rather than simply creative use of CSS." |
| 22:34:27 | <@OrIdow6> | szczot3k: You can see part of the cause also in how the warrior has turned from "a python script you can run!" to "here's a docker container with a million undocumented things that all have to be the same, no one may modify any of it" |
| 22:34:53 | <@OrIdow6> | And the fact that AB runs on some very old Python version |
| 22:34:55 | <@OrIdow6> | Etc |
| 22:35:05 | <steering> | huh, nearly $100/transaction average, that's kinda surprising |
| 22:35:09 | <joepie91> | I mean, speaking honestly, the warrior has never been a python script you can run, exactly |
| 22:35:22 | <joepie91> | it's always been a bit of a headache to get it to run on different systems for different projects |
| 22:35:30 | <szczot3k> | So I'd say at this point we don't really need more warriors or wikipeople, we need programmers |
| 22:36:14 | <joepie91> | historically there were the configuration scripts for specific projects that could, in principle be ran stand-alone headlessly if you had enough patience with your package manager, and the warrior which packaged up auto-download and auto-configuration of those scripts with a nice UI into a virtualbox image |
| 22:36:48 | <joepie91> | a headless "run whatever we need right now" was missing at the time |
| 22:37:00 | <joepie91> | I imagine that's what the Docker container is for |
| 22:37:38 | <steering> | I do feel that AT could use both more time spent on programming and more time spent on infra (see also: log site dead) |
| 22:37:56 | <joepie91> | oh sure, that's always been true though |
| 22:37:58 | <szczot3k> | Sure, but look at boinc for example - (most of the) projects don't need you to install virtualbox. You run an .exe, sign in, and boom. Your computer silently does the work when you're off it. |
| 22:38:04 | <steering> | who's gonna do it though? |
| 22:38:09 | <szczot3k> | And see how good boinc is doing |
| 22:38:10 | <joepie91> | doesn't mean it's not worth fixing but that's not a new problem at least |
| 22:38:22 | <joepie91> | steering: well. that's where meta-issues like "rules putting people off" come in :) |
| 22:38:41 | <steering> | sure, but someone's who's new can't do it anyway |
| 22:38:45 | <joepie91> | szczot3k: boinc also has actual funding behind it though, not an entirely fair comparison |
| 22:38:54 | <steering> | I've looked at some of the docs and some of the code and just... noped right out of there |
| 22:38:55 | <joepie91> | things could be better, sure, but someone needs to do the work |
| 22:38:56 | <nicolas17> | the BOINC codebase is as much of a mess as ours |
| 22:38:56 | <szczot3k> | joepie91 that's the next thing I wanted to say |
| 22:39:18 | <joepie91> | steering: I disagree, most of the work does not need archiveteam-specific skills |
| 22:39:27 | <szczot3k> | In the life of every projects of this scale, comes a point where it should be actually managed be some full time pople (a non-profit most of the time) |
| 22:39:30 | <nicolas17> | and since we're talking about the tracker being open source or messy etc. running a boinc project is no fun either :D |
| 22:39:48 | <joepie91> | a sufficiently experienced developer could, with some guidance from existing archiveteam folks on conventions/goals/etc., absolutely do a lot of the work |
| 22:39:52 | <steering> | joepie91: at the very least, it needs someone with AT specific skills to figure out how everything is connected and manage such a project. |
| 22:40:05 | <nicolas17> | also this makes me feel old: https://github.com/BOINC/boinc/blob/master/checkin_notes_2006#L10313 |
| 22:40:11 | <steering> | right, which implies those existing AT folks spending their time on it instead of other things :) |
| 22:40:21 | <joepie91> | steering: sort of. they are only really needed in an advisory role which is a much smaller ask |
| 22:40:40 | <joepie91> | szczot3k: there are some very good reasons why AT is not a formal organization and definitely not a non-profit :) |
| 22:40:40 | <szczot3k> | nicolas17 I was 3 and a half back then ;) |
| 22:41:11 | <szczot3k> | joepie91 but at the same time there are some very good reasons why it should be |
| 22:41:25 | <nicolas17> | I was 15, and the BDFL of BOINC didn't find that out until a year later |
| 22:41:32 | <steering> | nicolas17: lmao, `printf(some_input)`? |
| 22:41:41 | <nicolas17> | steering: pretty much |
| 22:41:47 | <joepie91> | steering: IMO this all ultimately boils down to just making it easier and more interesting for people to get involved, and on the one hand I think that means being more welcoming (see: rules etc.), and on the other hand I think it means there needs to be a clear way to *get* involved, know who to talk to, have someone available for guidance, and so on |
| 22:42:11 | <joepie91> | szczot3k: AT literally could not function as a non-profit, all the stuff that requires that better fits in the internet archive instead |
| 22:42:19 | <nicolas17> | I think my BOINC project's app printed %s to its stderr and that made the BOINC client crash when collecting the logs? something like that |
| 22:42:36 | <joepie91> | the problem with being a non-profit is that it means you legally are Something and that comes with a lot of new risks and responsibilities that just do not fit with the way AT works |
| 22:42:48 | <joepie91> | it not being formally organized is what makes it work |
| 22:43:10 | <joepie91> | this is the same problem that ~every activist org that formally incorporates runs into |
| 22:43:34 | <joepie91> | they always, without fail, become highly institutional and unable to navigate uncertainties and risk |
| 22:44:02 | <joepie91> | and so we will need to figure out ways to keep things going without that |
| 22:44:11 | <szczot3k> | joepie91 sure, I understand that, but at the same time, as you can see, someone should probably be managing AT full time, including at least some programmers. Volunteers work are great, but managing codebases, community, distributed works takes time. |
| 22:44:38 | <joepie91> | sure. but that doesn't require formal structures |
| 22:44:49 | <szczot3k> | But at least requires money :) |
| 22:44:58 | <joepie91> | also not |
| 22:45:02 | <steering> | I feel like even if you took every AB pipeline, stopped spending money on them, and spent that money on paying someone instead - it still wouldn't even begin to cover an FT salary |
| 22:45:11 | <steering> | except maybe for nicolas17, I dunno what COL is like there :P |
| 22:45:45 | <szczot3k> | If you think about it TOR also probably shouldn't be an organization, I'd say it has the same problems as AT |
| 22:45:46 | <joepie91> | szczot3k: the actual, basic requirement here is an environment that allows people to do this work with minimal friction and hindrances, and an effective way to communicate about it. that's it. everything else can be improvised, found, borrowed, traded, created from whole cloth, etc. |
| 22:47:00 | <joepie91> | and this is a really important point when organizing shoestring volunteer work like this, because as soon as you start seeing things like money or incorporation as things you need (rather than as the societally easy options they are), you will run out of runway faster than you can say "downtime" |
| 22:47:29 | <joepie91> | there is simply not the money to do this in an institutional way and even if there were, it would mostly just make it a very juicy target for anyone with a legal bone to pick |
| 22:48:17 | <joepie91> | so yes, there are real problems to fix here, but no, the standard corporate tools are not the solution :) |
| 22:48:31 | <szczot3k> | Volunteer work only scales so much. Love the idea, I'm doing IT for some nonprofits (including the largest polish education NGO), but I can't spend more than a few hours a week on this. This also affects my co-volunteers. Most of the stuff is cobbled together, just beacuse noone has the time to actually go through it, verify it, make it better. |
| 22:48:31 | <szczot3k> | Most codebases are worked on by only one person, who found 2 hours over the weekend to automate something, and the basic requirement is "it should work" |
| 22:48:58 | <joepie91> | okay, but please consider that your personal experience is not automatically representative of all that could exist |
| 22:49:14 | <joepie91> | I know of tons of volunteer-driven things that don't survive |
| 22:49:29 | <joepie91> | but I also know of volunteer-driven things that do fine, and have been doing fine for a long time |
| 22:50:02 | <joepie91> | the most useful thing is to understand where that difference is, why some projects fail and others succeed, and what the full spectrum of possibilities is |
| 22:50:36 | <steering> | szczot3k: yeah, that's exactly my point with who's gonna do it |
| 22:51:31 | <szczot3k> | I get it, I'm mostly trying to say it all depends on the situation. The current AT solutions got us a barely maintainable code, with tons of technical debt. If volunteer work could feed me, I'd probably be doing IT for NGOs/community projects full time, rather than doing jira workflows, for an enterprise |
| 22:51:44 | <steering> | on top of that you also have to consider timezones |
| 22:52:02 | <joepie91> | I think we all agree that the current situation with AT is not great; it's certainly not the worst it could be, but there's also a lot of room for improvement |
| 22:52:12 | <joepie91> | but that's why I've been trying to bring up some of the factors that I think play into that :p |
| 22:52:15 | <steering> | just because you have a few hours doesn't mean those hours overlap with when other people are available that you might need guidance or help from, so then nothing gets done half of the time that you have available anyway |
| 22:52:19 | <@OrIdow6> | joepie91: words of wisdom |
| 22:52:35 | <@OrIdow6> | (Not being sarcastic) |
| 22:52:47 | <joepie91> | my frustration is that "we need money" instantly terminates that line of conversation and leaves no room to talk about solutions by essentially declaring money The One True Solution upfront |
| 22:53:03 | <nicolas17> | steering: as if my ADHD (plus what recently appears to be covid-induced CFS) would let me do anything fulltime consistently |
| 22:53:08 | <joepie91> | likewise for "we need to incorporate" |
| 22:53:11 | <steering> | nicolas17: :D same |
| 22:54:49 | <szczot3k> | If we can't even agree to try to bump python to semi modern version, it's hard to do anything more than to 'try to maintain it' |
| 22:55:00 | <joepie91> | szczot3k: also, for some additional background: I do volunteer stuff almost full-time, and I get paid for none of it, because I work a limited amount of hours for paid customers and that pays enough to subsidize everything else. this is obviously not magically replicable but it's an example of how time management for volunteer work is not as simple as "people need to get paid". it's nice if you can make that happen but it is not a |
| 22:55:00 | <joepie91> | *requirement* |
| 22:56:23 | <szczot3k> | Sure, everyone has different experiences |
| 22:56:59 | <@arkiver> | szczot3k: are you talking about seesaw? |
| 22:57:25 | <szczot3k> | arkiver: I'm lost, not sure what you're mentioning |
| 22:57:39 | <@arkiver> | the "bump python to semi modern version" |
| 22:57:40 | <joepie91> | (I would prefer to return to more pragmatic discussions about how to improve participation, organization, etc. get to some kind of concrete idea of what the fundamental problems are) |
| 22:58:15 | <szczot3k> | arkiver I was bumping what another person said about the python version |
| 22:59:07 | <joepie91> | arkiver: do you know what happened with the tracker stuff, and why it's apparently no longer open-source? to start with that |
| 22:59:22 | <szczot3k> | joepie91 agreed |
| 22:59:55 | <@arkiver> | joepie91: tons of improvements were made on the server side, but the stuff that needs to stay private is currently not split out well from the stuff that should be made public |
| 22:59:59 | <nicolas17> | as for money here's our budget https://opencollective.com/archiveteam |
| 23:00:59 | <steering> | $5000 in it but only $1250 ever spent, seems right :D |
| 23:01:03 | <szczot3k> | And what's hidden after 'stuff that needs to stay private'? |
| 23:02:01 | <szczot3k> | s/after/behind |
| 23:02:17 | <szczot3k> | I've spent my english quota for the day |
| 23:09:40 | <joepie91> | arkiver: with 'stuff that needs to stay private', are we talking configuration files with private endpoints / API credentials and such, or something else? |
| 23:09:51 | <@arkiver> | private endpoints i believe |
| 23:09:57 | <@arkiver> | i will admit it has been a while since i checked the code |
| 23:11:06 | <szczot3k> | that kinda sounds like security by obscurity which scares me |
| 23:11:22 | <joepie91> | arkiver: that seems like a fairly small problem to fix at least |
| 23:11:42 | <joepie91> | szczot3k: it's less 'security' in the technical sense |
| 23:11:54 | <joepie91> | I think I know what's being kept private and if I'm right then that makes sense |
| 23:12:14 | <joepie91> | though that should really just be in a config file outside of the code |
| 23:12:59 | <joepie91> | arkiver: do you also happen to know what's up with there apparently being very little remaining working admin panel functionality? |
| 23:14:02 | <@arkiver> | what do you think is up with that? |
| 23:14:52 | <joepie91> | well, that I don't know :) I'm trying to work out to what degree the tracker is still being maintained |
| 23:15:19 | <joepie91> | going from earlier comments in here |
| 23:15:35 | <szczot3k> | arkiver: basically we're talking about the tracker wiki article: https://wiki.archiveteam.org/index.php/Tracker (and mostly 'history') |
| 23:15:57 | <@arkiver> | the admin has not been kept up to date with what has changed behind it. it is now largely controlled directly through redis |
| 23:16:31 | <@arkiver> | i will be putting up an interface through IRC though to control projects - i hope that will help with managing project |
| 23:17:07 | <@arkiver> | if we have those controls in the channels for the projects, there's more transparency as well on what is happening |
| 23:17:20 | <joepie91> | the concern here is partly that, but also partly the ability for someone to take over tracker duties if for whatever reason the current instance disappears or falls into disrepair |
| 23:17:34 | <@arkiver> | perhaps certain actions (emergency stop, some ignores) can be allowed to be done by a larger number of people instead of just a single one |
| 23:17:50 | <joepie91> | from the description it sounds like right now there is a lot of institutional knowledge on how to operate the tracker and it's not really possible for someone to independently get the infrastructure going again |
| 23:17:53 | <steering> | honestly my bigger concern is infrastructure |
| 23:17:55 | <szczot3k> | Part of the concern is also having the tracker not be distributed |
| 23:18:01 | <steering> | we were extremely lucky to save eggdrop |
| 23:18:09 | <steering> | kiska's logs got oopsied |
| 23:18:37 | <steering> | the tracker is a bit better protected, are there good backups of it? |
| 23:19:25 | <steering> | what about some of the other stuff that one person has access to/maintains? |
| 23:19:51 | <@arkiver> | i would say eggdrop is of an entirely different than the tracker - i'm not sure how many people had access to that behind the scenes |
| 23:19:54 | <steering> | even pipelines, although I think anything we could do to add more resilience there would also have the downside of heightening the barriers to running one |
| 23:19:56 | <@arkiver> | (i did not) |
| 23:20:10 | <steering> | only fireonlive, and JAA had a couple of commands |
| 23:20:27 | <szczot3k> | Soo... no revolution? |
| 23:20:29 | <steering> | yes, I agree they are vasly different in importance and scope :) |
| 23:20:44 | <steering> | +t |
| 23:20:47 | <@arkiver> | steering: access yes, maintenance not many, but that does not mean it will suddenly drop out and never return |
| 23:21:08 | <@arkiver> | szczot3k: feel free to |
| 23:21:20 | <szczot3k> | :) |
| 23:21:47 | <@arkiver> | i somewhat get the feeling though from this discussion like AT is on the brink of death, and has been going downhill the past years(?) |
| 23:22:25 | <@arkiver> | i am probably the wrong person to say this - as i'm strongly involved in things here - but i think we have expanded projects and have very massive archiving operation running now |
| 23:22:45 | <szczot3k> | That's a spiral, not a lot of people want to invest time/effort into a failing cause, and the cause is failing, because not a lot of people invest time |
| 23:22:46 | <@arkiver> | arguably the most significant one next to some AI companies and secret services |
| 23:23:11 | <steering> | Hmm. |
| 23:23:19 | <@arkiver> | szczot3k: (again, i may be the wrong person to note this) but i think we have not been running down some spiral the past years |
| 23:23:26 | <@arkiver> | it has all only expanded |
| 23:23:29 | <steering> | I see a lot of motivated people here. |
| 23:23:39 | | Froxcey quits [Remote host closed the connection] |
| 23:23:43 | <steering> | I don't think AT is going to die any time soon. |
| 23:23:58 | <@arkiver> | mostly ups, with some downs, i think those downs do not justify claims of a downward spiral |
| 23:24:34 | | Froxcey (Froxcey) joins |
| 23:24:51 | <@arkiver> | i understand that for example OrIdow6 (maybe others?) have been feeling restricted by how things are setup currently when running project. having to run things through me or JAA when something needs to be done on the tracker is not fun/great/encouraging |
| 23:25:00 | <imer> | there is quite a barrier to helping out with actually making projects/improvements to core tools/services (vs running project containers is super easy), i think thats the main issue here? |
| 23:25:01 | <imer> | if not impossible, don't think anyone could even e.g. make small improvements to the public tracker ui for example if they wanted to |
| 23:25:02 | <szczot3k> | I'm not saying it is, I've only been 'a part of' AT for a week, but looking at the wiki doesn't give a lot of confidence in the project. Most of the things I've seen on the wiki is "We somehow survive", including the aforementioned tracker |
| 23:25:08 | <@arkiver> | i hope to fix that soon with issuing commands through IRC |
| 23:25:19 | <joepie91> | arkiver: I find it difficult to characterize; I definitely feel like archiveteam is not as *healthy*, as a project, as it has been in the past. participation seems down significantly, as does general enthusiasm, and generally Things Happening. I can't say much about how much work it is getting done, but if it is still getting a lot of work done, the sense I get from it is that it has become mostly an Internal Thing(tm) with all of the risks |
| 23:25:19 | <joepie91> | and blind spots that that entails. the culture also seems to have trouble keeping up with broader societal changes in some ways. |
| 23:25:37 | <@arkiver> | imer: right, does remind me to get quber still on github |
| 23:25:43 | <joepie91> | I want to emphasize here that "healthy as a grassroots project" and "doing lots of work" are two entirely separate metrics |
| 23:26:26 | <@arkiver> | joepie91: participation seems to be down? we have much more resources on projects than in the past. a lot of people also comment and get involved through projects like archivebot |
| 23:26:46 | <@arkiver> | i also do not agree with "Things Happening" being down |
| 23:26:46 | <szczot3k> | All of that come to having a big barrier to entry. I was determined to get through the barrier, but a lot of people will leave after taking a look on some wiki articles (the 'irc rules', the tracker, the somewhat complicated warrior setup), and their first message on #at being 'hi', and getting yelled at :) |
| 23:27:00 | <joepie91> | arkiver: oh, just noticed that something got lost in an edit. "from a distance" was supposed to be in there somewhere |
| 23:27:08 | <steering> | similar to imer, I think it's very difficult to help in some ways, and a lot of the kinda "ancillary" services that are widely relied on or helpful are just kinda random projects, that could disappear at any time for any reason |
| 23:28:14 | <@arkiver> | szczot3k: i agree the IRC rules could be changed. i think the change OrIdow6 made was due to this discussion |
| 23:28:26 | <joepie91> | arkiver: I'm more talking about general organizational participation in targeted projects (not including archivebot, which is kind of its own thing). it feels much more "there's one guy organizing things" than it used to |
| 23:28:38 | <@arkiver> | joepie91: i admit i am not very familiar with the AT and societal changes combination. do you have examples? |
| 23:29:13 | <szczot3k> | Also - it feels like many things are a one person job, RE: archiving of pclab.pl, which was started pretty late, cause of what I'm seeing as only one person being able to do the task |
| 23:29:49 | <szczot3k> | Kinda repeating, because while I'm cooking the message I don't read the backlog :D |
| 23:30:00 | | Froxcey quits [Ping timeout: 260 seconds] |
| 23:30:14 | <@arkiver> | szczot3k: is that being archived with qwarc now? i think qwarc is public (please correct me if i'm wrong) |
| 23:30:33 | | Froxcey (Froxcey) joins |
| 23:30:41 | <joepie91> | arkiver: the IRC rules would be one example (that particular style of rules is pretty broadly considered undesirable now and has been for years, even on IRC), the conflict around mastodon was another example, and another thing (non-exhaustive) that stood out to me is that there's a lot of trans folks in the general "archiving internet things for posterity" discourse but I am not seeing them here |
| 23:30:46 | <szczot3k> | AFAIK it's qwarc, from what I've gathered it seemed like only J.AA could do it |
| 23:30:59 | <joepie91> | these are just some examples off the top of my head of things that don't seem quite right |
| 23:31:37 | <@arkiver> | joepie91: i would say all people are welcome, if the IRC rules discourage that they should be changed (and after what OrIdow6 it's clear they should be changed) |
| 23:31:46 | <joepie91> | like, how is it that there are a lot of people who are voraciously 'archiving' things on archive.is but they have never heard of archiveteam, for example? |
| 23:31:59 | <@arkiver> | as for archiving - i think "archiving public stuff at risk or going away" has always been the same |
| 23:32:16 | <joepie91> | there should be a collision of social groups there, but there isn't |
| 23:32:18 | <@arkiver> | joepie91: AB is only on IRC, so not very approachable perhaps |
| 23:32:58 | <joepie91> | I think that's part of it, but I'm really talking about the people who have archive.is at the top of their bookmarks list and use it daily, who would very likely be interested in archiving whole sites and not just pages |
| 23:33:13 | <tzt> | joepie91: A lot of people know about archive.today and archive.org the WBM, but not archiveteam since it all of the projects go into the WBM and you'd only know if you checked where the archive came from |
| 23:33:16 | <szczot3k> | AB is also kind of chaotic in terms of getting stuff into AB. Things on #ab are easily missed, and only some people can use ab directly |
| 23:33:19 | | Froxcey quits [Remote host closed the connection] |
| 23:33:25 | | Froxcey (Froxcey) joins |
| 23:33:33 | <joepie91> | and like, none of these things in and of themselves are big issues or anything, but all put together I cannot help but feel like archiveteam is fading away in a sense |
| 23:33:54 | <joepie91> | at least in the public consciousness |
| 23:33:58 | <@arkiver> | i really don't have that feeling - but i'm biased |
| 23:35:13 | <joepie91> | I mean, that's kind of why I'm going off specific "things that look wrong to me" here :) you have an unusually close perspective, I have an unusually distant perspective (as I have not been very closely involved for a while), so all we have to go off is data really |
| 23:35:37 | <@arkiver> | joepie91: note that most discussions around projects nowadays happen in separate channels. #archiveteam-bs often gets less activity than those channel combined |
| 23:36:00 | <@arkiver> | since you said "from a distance" before, if you're not actively joining the channels, maybe it's not noticeable |
| 23:36:01 | <joepie91> | sure |
| 23:36:13 | <@arkiver> | but again - i may be looking to enforce my bias :P |
| 23:36:20 | <joepie91> | this was also true in the past |
| 23:36:31 | <szczot3k> | Not many people know of AB, less people understand how to use IRC (effectively), less people will hop onto the channel and we like "Hi, can you please archive hackint.org"? |
| 23:36:41 | <joepie91> | but back then there would be a fair amount of meta-discussion in -bs, around tooling and infrastructure and new projects and the initial discussion for them |
| 23:36:43 | <joepie91> | and I'm missing all of that too |
| 23:36:59 | <joepie91> | all the stuff that isn't just specific projects |
| 23:37:31 | <szczot3k> | While we're on the topic, it might be beneficial to run a hackint.org job on AB |
| 23:38:09 | <joepie91> | arkiver: to clarify, my concern here is specifically a 'coasting phase' where everything still seems fine because we're all coasting along on networks and goodwill and connections built up over the years, but nothing is replenishing those. this is often signaled by such organizational meta discussions going missing as well as social chatter |
| 23:38:41 | <joepie91> | if this is indeed the case, then everything will look fine right now, and in a few years it'll suddenly be dead when a critical part falls away and there's nothing or nobody to replace it |
| 23:38:51 | <joepie91> | I am not sure that it is the case, but it is the concern I have |
| 23:40:31 | <joepie91> | if nothing else, it is probably worth keeping an eye out for signals that this may be the case, and slowly but surely going through contingency plans to make sure that there are options for the different possible failure modes in the future :) |
| 23:40:44 | <TheTechRobo> | My 2cents: The lack of openness and maintenance on the infrastructure has been really discouraging for me in the past. |
| 23:40:54 | | etnguyen03 (etnguyen03) joins |
| 23:41:20 | <@arkiver> | joepie91: very nicely explained/summarized - thank you. i think it's not entirely the case, but there are some signals that speak for that |
| 23:41:53 | <TheTechRobo> | Like even little things like not having backfeed on universal-tracker, or requiring ancient Python versions. |
| 23:42:34 | <@arkiver> | backfeed is here https://gitea.arpa.li/ArchiveTeam/backfeed |
| 23:43:22 | <TheTechRobo> | Yeah, I found that a few months ago, but still don't know how it works. (Not familiar with Go, either, so I can't really look at the code very well.) |
| 23:45:05 | <szczot3k> | Read the -bs backlog - https://wiki.archiveteam.org/?title=Qwarc - it fits into the topic of 'we kinda need more programmers, and code standards' |
| 23:45:35 | <@arkiver> | szczot3k: why coding standards? |
| 23:45:39 | <@arkiver> | code* |
| 23:45:47 | <TheTechRobo> | I have been following -bs; all I see is discussion abou the rules |
| 23:46:14 | <szczot3k> | TheTechRobo I just mean that I've seen the wiki edit about qwarc |
| 23:46:31 | <TheTechRobo> | Ah I thought you were telling me to read the -bs backlog |
| 23:46:48 | <szczot3k> | arkiver because managing code by different people without such standards gets messy over time |
| 23:47:02 | <szczot3k> | s/read the -bs/I've read the -bs/ |
| 23:48:29 | <szczot3k> | While it's true that qwarc is open source, coming back to pclab discussion with the new knowledge - it still hase a problem of 'J.AA is the only one who really knows how to use it. ' |
| 23:49:07 | <steering> | Thanks, English, for making "read" and "read" the same word :) |
| 23:49:07 | <TheTechRobo> | qwarc has the advantage at least that there are loads of examples. I was taking a look awhile back and it didn't look too hard to write a specfile. |
| 23:50:40 | <@arkiver> | there will always be a learning curve (not talking about code standards, but qwarc and such). you do have to put in time and energy to learn to use it |
| 23:51:01 | <@arkiver> | on code standards - i see the reasons behind it yes. but then whatever doesn't fit in that should be rejected? |
| 23:52:10 | <TheTechRobo> | arkiver: Re qwarc, yes, but there should still be documentation. Examples don't fully explain everything. |
| 23:52:18 | <imer> | Oh, I can chime in on code standards! As a contributor you generally do want to follow things, but its hard to know what those things are - lowering the bar with things an IDE can flag for you automatically (e.g. via .editorconfig) is very nice |
| 23:52:41 | <szczot3k> | Last thing to put onto the pile - AT also has external challenges, like cloudflare, which I'd say is a big danger to AT's work in the perspective of a couple of years. |
| 23:52:47 | <TheTechRobo> | qwarc is one of the better ones as its API is pretty easy, but there should still realistically be *some* documentation. |
| 23:53:03 | <imer> | bit the bullet for a project I'm contributing to to add that and its been so much nicer for me to just follow what the IDE suggests instead of having to think about it |
| 23:53:09 | <@arkiver> | imer: if people use an IDE :P |
| 23:53:19 | <TheTechRobo> | szczot3k: Bypassing Cloudflare, whether through a browser or reverse-engineering, isn't impossible. https://gitea.arpa.li/JustAnotherArchivist/cloudflare-circumvent is broken but I believe it did used to work. |
| 23:53:27 | <imer> | well, cant be helped if people dont - I'd expect most do in some capacity though? |
| 23:53:35 | <imer> | most/many |
| 23:53:40 | <steering> | I don't |
| 23:53:47 | <steering> | but then I'm weird :P |
| 23:53:56 | <imer> | there is a number of masocists on this earth, yes ;) |
| 23:53:57 | <@arkiver> | me neither. started without... and still going strong |
| 23:54:16 | <@arkiver> | gedit yay (just used to it) |
| 23:54:21 | <steering> | I use vim, and have some stuff i nit that might be IDE-ish, but no reformatting or code completion or anything like that |
| 23:54:23 | <TheTechRobo> | I believe J.AA uses nano and tmux. :-) |
| 23:54:25 | <imer> | but yeah, whatever works works |
| 23:54:33 | <TheTechRobo> | I also use vim, and yeah, no styling |
| 23:54:33 | <@JAA> | I do. |
| 23:54:41 | <@JAA> | Mostly, anyway. |
| 23:54:45 | <TheTechRobo> | s/styling/reformatting/ |
| 23:55:13 | <TheTechRobo> | I do use code completion though. |
| 23:55:20 | <steering> | I think there's a bigger problem with code standards: AT builds on a lot of disparate projects |
| 23:55:21 | <imer> | things like that lower the bar though if code style is enforced |
| 23:55:38 | <steering> | (and in disparate languages) |
| 23:56:16 | <steering> | so it would really be a lot of different code styles :) |
| 23:56:27 | <szczot3k> | ALSO coming back to the 'we need an NGO' argument - some of the challenges can be alleviated by reaching out to the interested parties. I've wanted to see mpcforum.pl being archived (a big, old, popular, influential polish 'computer'/gaming community forum). It looks different when some random person who calls themself an 'internet archivist' |
| 23:56:27 | <szczot3k> | emails an entity and is like "hey, so can you ummm... turn your defenses (cloudflare) off for us? We want to download all your data", and might be different when an established entity, that is willing to get things in writing reaches out. |
| 23:56:54 | <TheTechRobo> | Reformatters are nice in languages with sane conventions. *cough cough PEP8* :-) |
| 23:56:58 | <szczot3k> | TheTechRobo: bypassing cloudflare won't always be a possibility. It's kinda their job to make sure that there's no easy way to bypass cloudflare. |
| 23:57:12 | <steering> | TheTechRobo: PEP-8 https://www.youtube.com/watch?v=QxQEf1JXq-Y |
| 23:57:17 | <TheTechRobo> | szczot3k: Yes, but it's not really a threat so much as a challenge. |
| 23:58:04 | <szczot3k> | If cloudflare really wanted I'd find a way to block warrior, and archivebot easily, probably without spending too much time on this task. |
| 23:58:25 | <szczot3k> | s/I'd/They would |
| 23:58:28 | <TheTechRobo> | steering: Exactly :D |
| 23:59:00 | <steering> | and then there's python.vim 3$%#!@*%$(#$ overruling my settings |
| 23:59:11 | <steering> | " python.vim should NOT be changing vim settings by default -_- |
| 23:59:11 | <steering> | let g:python_recommended_style = 0 |
| 23:59:27 | <szczot3k> | A fight between cloudflare, and archive team would probably be lost quite fast |