01:21:57 Law360 and LexisNexis ended their Pro Say podcast on December 21st of 2023. https://podcasts.apple.com/us/podcast/law360s-pro-say-news-analysis-on-law-and-the-legal-industry/id1240435608 https://twitter.com/AlexxLawson/status/1737989879051468954 01:21:58 nitter: https://nitter.net/AlexxLawson/status/1737989879051468954 01:32:31 PonyChan supposedly dies today - might be a good time to grab anything posted after the last grab in December? 01:34:13 https://youtu.be/gWL90wryyOw 01:34:24 I am joking btw but too good a chance not to post this in response 01:41:57 JAA: https://opensource.samsung.com/uploadSearch?searchValue=- is this archived or being worked on or in anyone's radar? 01:48:12 nicolas17: Looks like it's been mentioned a couple years ago, but that's about it. 01:49:53 looks like - doesn't return everything everything 01:51:30 using the release centre menu and the headings should do it though? 01:52:11 https://opensource.samsung.com/uploadList?menuItem=tv_n_video https://opensource.samsung.com/uploadList?menuItem=mobile https://opensource.samsung.com/uploadList?menuItem=home_appliances https://opensource.samsung.com/uploadList?menuItem=pc https://opensource.samsung.com/uploadList?menuItem=more 01:52:12 ..maybe 01:52:39 i love how every button is javascript 01:52:55 < just how god intended 01:53:27 . might be a better search term. 01:54:55 seems to include more from different categories but missing more than the menuItem combination 01:56:23 'more than the' 'stuff from the' 01:57:28 * fireonlive gives samsung the 'what the hell have you built' sticker 02:00:02 yes 02:00:15 I tried all IDs instead and found more than with any search term 02:17:45 nulldata: I put https://www.invisionapp.com/ in AB the other day but didn't do the subdomains cus there are a ton of them, most with no content 02:17:55 didn't have time to sort thru them really 02:30:19 downloading everything from opensource.samsung.com doesn't seem difficuly 02:30:31 the problem is it's all POST so it won't be suitable for WBM 02:33:20 :| hwy is it posts 02:33:40 it's a POST with a one-time (or expiring) "token" 02:34:07 ahh :c 02:50:31 What can be done about such requests then? 02:51:16 Pedrosso: They can be saved, but the Wayback Machine can't play them back 02:51:40 I dont there's much that can be done other than adding special handling to the WBM 02:52:14 I suppose then what's important is that they can be saved 03:35:33 Pedrosso: I might download everything and upload it into a plain old item in archive.org 03:37:24 not as easy to use as WBM but effective 03:39:51 JAA: re zstdwarccat, am I grokking this correctly: it can be used with just a .warc.zst (single command line arg), and it will extract the dict from the zst, write it to a tempfile, and call zstd -D with that extracted dict? 04:26:03 JAA: https://cdn.discordapp.com/attachments/286612533757083648/1193771879652147231/image.png?ex=65adedb3&is=659b78b3&hm=f1b44aecc421370a1605e42bc7a45b12ea70f142e08c3df8d6dc9a8845aa6006& 04:26:42 they have different file ID, but same filename, same file size, and assuming I didn't accidentally start downloading the same one twice, same content in what I downloaded so far 04:35:43 samsung pls 04:36:02 and they download at 200KiB/s 04:47:09 Whaaaa? What the hell? APKPure suddenly...shut...down...? https://www.claytoncountyregister.com/news2/what-happened-to-apkpure/ 04:47:19 http://apkpure.com/ 04:47:53 ...I use that and APKCombo for browse skimming through new mobile games S: 04:48:10 Would like more archivey of .APK stuff :C 04:48:21 that article looks like pure speculation, information-free, and potentially written by an AI lol 04:48:48 it does not even remotely "shed light on the situation" 04:49:06 Yeah, that unfortunately looks like an AI article, bleh 04:49:25 ...But that clues me in that APKPure is down in the first place 04:49:32 Again, use both of the websites 04:49:53 ...Might need another website as a backup~ 04:51:47 https://old.reddit.com/r/ApksApps/comments/18zoqrs/did_apkpure_just_get_taken_down/ and https://old.reddit.com/r/androidapps/comments/18z2l5o/apkpurecom_down_today/ 04:52:30 They might have been acquired? Unsure if true: https://old.reddit.com/r/ApksApps/comments/18zoqrs/did_apkpure_just_get_taken_down/kglezti/ 05:14:49 Ryz - looks like https://apkpure.net/ is up 05:16:25 Huh, unsure if a mirror or not, since the APKPure one is the .com one - https://web.archive.org/web/20240000000000*/http://apkpure.com 05:18:45 Not sure - they are both registered with Mark Monitor and behind Cloudflare. com has a private registration and net lists ELECYBER INTERNATIONAL PTE. LTD. 05:20:29 Almost a year apart on the registration 05:20:39 lol it's going to take 2 hours to download the open source notices from samsung 05:20:45 I guess it will take days to download the actual zips then 05:22:39 The YouTube channel that was linked on the .com one mentions .net in the bio links, but yet the YouTube links to a different Facebook account vs what was listed on .com 05:23:36 https://lounge.nulldata.foo/uploads/139d181123007003/image.png 05:23:41 https://www.facebook.com/APKPureOfficial/ 05:24:05 https://www.youtube.com/@apkpure6707/videos 05:26:31 nulldata: is that screenshot from just now? 05:26:40 Yeah 05:26:56 okay, so I know what "4h ago" is relative to :P 05:37:11 Hmm, yeah, so it looks like there may be a .com and .net version that's legit, at least from checking https://t.me/s/apkpurechannel 05:55:21 trying to download from the samsung site to my VPS, and speed is *worse*, I wonder if this intentional malicious compliance 05:55:58 "we provide source code for the open source components we use, complying with the license, they are 800MB zip files downloadable at 100KB/s" 07:06:34 I assume uploading 5406 files adding up to 800GiB into a single IA item is not ideal 07:11:11 and I just found part of that data is 20GB of licenses and copyright notices which compress to <0.3% so I would feel really bad uploading it uncompressed >.> 07:21:10 so far I downloaded 15GB and found 537MB of duplicated files 09:34:33 hi 09:34:44 im trying to view images under a blur 09:34:50 to archive them 09:34:56 i was told to use inspect element 09:35:01 but not sure how 09:46:12 D: if you're not sure how because you've never done that before, the web should be full of step-by-step tutorials. usually you can right-click the blurred image and should see an "inspect element" menu entry. if you click that, the window will be split and a hierarchy of the html elements will be shown to you. you can hover over them in the text so they will be highlighted on the page. 09:46:43 if you have an idea which element it could be, there should be some way to "disable" it in the development tools so it will not be rendered on the page 09:46:45 what element do i remove to remove the blur 09:47:31 sometimes the name can give you hints, but it's mostly try and error. this is hard to explain via text to someone who has never done that before ^^ 09:50:15 maybe you can find a youtube video that explains it? it should be easier to understand to see what someone is talking about while they speak 13:41:58 project10: Correct 13:48:06 nicolas17: Why am I not surprised? 15:07:08 JAA: neat, I wasn't aware that the .warc.zst had the custom dictionary embedded, thought I needed to acquire the associated zstdict through other means 15:12:44 project10: Yeah, it's in a skippable frame which isn't understood by regular tooling. 15:16:02 first implemented in Wget-AT, now somewhat officially proposed in the WARC standard as well :) 15:16:06 (well since many months) 15:35:04 Ubuntu Looking at Discontinuing Its Source ISOs: https://www.phoronix.com/news/Ubuntu-Discontinue-Source-ISOs https://news.ycombinator.com/item?id=38876167 15:40:18 oh? 15:41:14 * arkiver never understood the push for snap 15:41:52 if there is an archive of src ubuntu ISOs, let's get them 15:41:55 perhaps in #archivebot-bs 15:41:58 perhaps in #archivebot * 15:42:42 Not sure there's anything unique in those ISOs that isn't also in the package repos, but if it isn't huge, might not hurt either. 15:43:37 It's about these, for example: https://cdimage.ubuntu.com/releases/23.10/release/source/ 15:45:04 (I think, at least.) 15:46:51 how do i unblur images 15:47:00 on sites like patreon and substar 15:49:04 D: oh, that's simple. i didn't know you were talking about those. you just pay the creators :) 15:53:41 yeah! :) 15:54:54 Note that there's duplication in it. For example, https://cdimage.ubuntu.com/releases/mantic/release/source/ and https://cdimage.ubuntu.com/releases/23.10.1/release/source/ should have the same files. 15:56:45 what if the creators are dead/seem dead 15:56:57 and u want to archive their work 16:02:33 they do not handle the booking themselves. if they still have creator accounts on there, you should still be able to subscribe and download their stuff 16:03:38 i know gallery-dl supports at least patreon, so you can just pass your browser's cookies to it and it will download the creator's pictures (and maybe videos?) for you. 16:03:45 oh well 18:54:52 eh linux distro source isos are of questionable usefulness, who's gonna burn the source code into physical media? 19:16:09 i do wonder who did & why 19:16:20 i guess offline apt source comand? 19:16:22 commands 19:16:38 you can throw your enemies with it 19:16:51 ninja star! 19:16:55 JAA: what do you think about this samsung open source stuff? should I create an item with 800GB of zips? I'm not sure how to split it up :/ 19:18:02 I guess the only potential usefulness would be 'collection of all source code behind a major Linux distribution at a specific point in time', but yeah, no disagreement. 19:18:46 i think some times $manufacturers have accidentally shipped something they weren't supposed to, discovered many years later 19:18:50 internal stuff or whatever 19:19:29 nicolas17: 5.4k files in a single item might be a bit unwieldy, but if there's no reasonable way of grouping it, maybe it's the least bad approach, yeah. 19:19:31 I think the ubuntu source iso topic and the samsung open source topic are getting crossed 19:19:36 Yeah 19:20:03 well, I could also do 2500 items, each with one .zip for one device model 19:20:07 but I don't know if there's a reasonable middle ground between "all in one item" and "2500 items", such as grouping by device type or something 19:20:29 What about grouping by year? 19:20:31 Almost all files are for mobile devices, I think? 19:20:57 What kind of file-level metadata is there? 19:21:21 I think there's more metadata in the search results page than in the files themselves, but let me check 19:22:45 each row in the search results is an "upload" 19:24:17 each "upload" has one or more "attachments" 19:24:26 ( https://opensource.samsung.com/downSrcMPop?uploadId=11931 example with multiple files) 19:25:31 they can be "source" or they can be "announcement" (usually massive html or text files with licenses and copyright notices) 19:25:33 like this https://opensource.samsung.com/downAnnMPop?uploadId=11931 19:25:45 the vast majority have 1 source zip and 1 announcement html though 19:26:43 the model number and device type only show up in the search results, as far as I can tell, not in the zip contents nor in the DownSrcMPop page 19:27:33 for the "announcement" html files, there's so much redundancy that "tar | zstd -19" (in progress) gets a 0.8% compression ratio 21:39:20 JAA_ grew a tail 22:07:42 owo 23:27:52 JAA: re: transfer (from #frogger): curl --upload-file ./100MB.bin https://transfer.archivete.am 0.34s user 0.24s system 1% cpu 46.623 total 23:30:32 I wonder if this is at least partly the cause of h2ibot multi-hour processing, as it seems to upload the "cleaned" list to transfer w/o compression (so a multi-GB upload at <20Mbps) 23:31:42 also of note, h2ibot's backend doesn't seem to zst compress the files first which might be helpful for some lists 23:31:55 by backend i mean whatever arkiver is running :3