| 01:56:36 | | Wayward quits [Remote host closed the connection] |
| 02:50:06 | | OrIdow6 quits [Ping timeout: 250 seconds] |
| 03:46:00 | | DogsRNice quits [Read error: Connection reset by peer] |
| 03:47:48 | | OrIdow6 (OrIdow6) joins |
| 03:54:13 | | qw3rty_ joins |
| 03:58:00 | | qw3rty__ quits [Ping timeout: 258 seconds] |
| 05:54:49 | | qw3rty_ quits [Read error: Connection reset by peer] |
| 05:55:08 | | qw3rty_ joins |
| 06:04:39 | | Fusl quits [Excess Flood] |
| 06:04:50 | | nothere quits [Client Quit] |
| 06:04:59 | | Fusl (Fusl) joins |
| 06:05:20 | | nothere joins |
| 06:05:39 | | swebb_ quits [Ping timeout: 258 seconds] |
| 06:07:09 | | swebb joins |
| 09:01:28 | | OrIdow6 quits [Ping timeout: 250 seconds] |
| 10:40:39 | | OrIdow6 (OrIdow6) joins |
| 19:27:35 | | DogsRNice (Webuser299) joins |
| 19:33:27 | <@JAA> | TIL the WBM returns at most 2 GiB of data apparently. |
| 19:34:50 | <@JAA> | As in, trying to download a larger file (4-6 GiB tars in my case) results in a truncated file. |
| 19:35:55 | <@JAA> | I tried to verify whether two files I found on my disk are identical to the copies in the WBM, but the downloads stopped at exactly 2 GiB (and those are indeed identical). |
| 19:36:30 | <@JAA> | Does anyone know of a workaround for this? They're in liveweb, so downloading the WARC record directly isn't possible. |
| 19:42:58 | <@JAA> | The specific files are https://web.archive.org/web/20181210134758id_/https://media.offenegesetze.de/bgbl1.tar.bz2 and https://web.archive.org/web/20181210135028id_/https://media.offenegesetze.de/bgbl2.tar.bz2 in case that matters. |
| 19:58:47 | <Jake> | range requests maybe? |
| 20:05:43 | <@JAA> | Nope, looks like the WBM ignores the Range header entirely. |
| 20:06:16 | <@JAA> | Just serves the same first 2 GiB. Or well, at least it starts doing that, I didn't run the full hour-long transfer again. |
| 20:07:55 | <@JAA> | Specifically, I tried getting the first MiB starting at 2 GiB but got a full data stream; truncating to the first MiB confirms that it's the beginning of the file, not the chunk at 2 GiB. |
| 20:09:30 | <Jake> | :( Hmm, lemme think on this. |
| 20:20:26 | <Jake> | Looking at the headers I see "Warning: 299 wayback content truncated by "length"" I wonder if this means it was truncated when recorded or on playback, as I know SPN has a 2GB limit now. |
| 20:21:59 | <Jake> | (I'm trying another large item from liveweb https://web.archive.org/web/20181210152622/http://speedtest.belwue.net/10G ) |
| 20:24:00 | <@JAA> | Ah, missed that. Yeah, could be. |
| 20:54:11 | <Jake> | Trying my same link not from liveweb, but archiveteam, I get 504s, presumably trying to load all 10GB rather than 2? Trying to find something closer to 3GB to test this out, maybe some sort of OS iso? |
| 21:00:57 | <@JAA> | Here's a recent one from AB that should be 2518582021 bytes: https://web.archive.org/web/20210612040004id_/https://static.nhtsa.gov/nhtsa/downloads/CAFE/2018-27_MD-HD-Phase-2/Sensitivity%20Analysis.7z |
| 21:01:29 | <@JAA> | Er no, wrong number. 2517816128 bytes. |
| 22:07:41 | <Jake> | Yeah, not seeing a warning header on that, so my best bet is Liveweb used to cut off files at 2GB. (Trying to download the whole thing though.) |