The Usenet Megathread - No invite discussion!

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Usenet Megathread - No invite discussion!

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

mpeg4v3 posted:

I've been using UseNetNow (same company that owns BlockNews) for the past six months (since their Thanksgiving deal), and I've noticed that lately, I'm running into a LOT of incompletes, a lot of the time just-barely, too (like, 30 blocks needed after par2s out of 2,500 total). I'm just wondering if this is all just DMCA bullshit, or if anyone else running UseNetNow/BlockNews has been seeing the same sort of issues and has switched to a different provider that hasn't been having issues.

I've been having this problem too (on Blocknews), even on things that I've downloaded a few days previous and had to redownload because I'm an idiot and accidentally deleted them. Since they're suddenly missing a few bytes here and there on the RARs like 10 blocks from the PAR sets I figure it has to be intentional.

# ¿ Jun 12, 2011 19:47

Adbot: ADBOT LOVES YOU

# ¿ May 21, 2024 17:07

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

inpheaux posted:

Uhhhh, if it's in your history and you're on 0.6.x you should have a "retry" button that does exactly what you want.

I think what he wants it to do is force it to try redownloading from a different server. I've noticed that since many articles are becoming only slightly corrupted (probably due to DMCA requests?) 0.6 will just try to download it again from the same server, so all it does is redownload the corrupt file.

Then again, it may actually be getting them from a different server anyway and it's also corrupted there - I have several set up as backups but there's no easy way of telling which one it came from or which one had the corrupted file (or if it's all of them). Maybe something to look into for 0.7?

# ¿ Jun 13, 2011 01:36

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

io_burn posted:

So on a whim today I decided to setup nzb.su (previously was only using newzbin and nzbmatrix) and Sick Beard almost immediately found all this obscure stuff I had added when I first set it up then basically wrote off as "Yeah, this will never download." What reason is there that nzb.su would work so much better than the others? I mean don't all of the nzb services work basically the same behind the scenes?

So yeah, if you haven't yet, set up nzb.su. I was so pleased with their indexing that I tossed 'em $20.

Interestingly enough, I was looking for some stuff on nzbs.org and decided to check out one of the sites it suggests when it can't find something, MysterBin. Not only did I find what I was looking for (and with less missing PAR blocks than the equivalent nzb from nzbs.org, strangely), they actually have a feature I haven't seen on any other indexer: it actually searches inside RARs and poo poo instead of just matching on post titles. This cuts down on spam hits considerably (tell it to filter out *.exe) and also incorrectly labeled/purposely mislabled things. I'd definitely give it a shot if you're having trouble finding something.

# ¿ Jul 11, 2011 20:44

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

paradigmm posted:

newznab/nzb.su does this as well and additionally has thumbnail previews of video media

edit: nevermind, realized you were talking about the actual "search" function

Yeah, I can't really wrap my head around how that works without them actually downloading everything first :psypop:

Newznab has to have someone add poo poo to whatever it already has indexed, right?

# ¿ Jul 12, 2011 00:56

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Lenins Potato posted:

I was playing around with NZB Completion Checker and looking at incompletes on astraweb. I was really surprised that changing the port would make a difference on which files would be incomplete. Both NZBs happened to be around the 65-70 day mark.

This was the US server on port 119

Port 23

Port 1818

Port 8080

EU Server, only port 119 had an incomplete

I found another file that would have 4 incompletes on port 119, 1 incomplete on port 23, 1 incomplete on 1818, and no incompletes on 8080. Although, only ports 119 and 23 had a common incomplete. No incompletes on the EU server.

Between par2s and setting up the EU server as a backup, you should still be able to get all the files. Even if it is a less than ideal solution.

Is there a way for something like this to be implemented in SABnzbd (that is, check something for completion before downloading)? Seems like it would solve a lot of problems.

# ¿ Jul 19, 2011 21:22

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Sperg Victorious posted:

It takes a while for any of the programs to check. A 6Gb linux iso is probably 10 minutes or so.

That still seems preferable to me, but obviously at that rate it should be optional. I can't download faster than ~1.5 MBps, so spending 10 minutes on checking whether or not I'll be able to use something that would take literally almost 10 times that to download would be very useful.

# ¿ Jul 19, 2011 23:55

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Hogburto posted:

Avidemux is really good for merging avi files (which is pretty much always the 700MB filetype). If there are subtitles, you'll want to take care of them first, though.

Unfortunately, I don't know any console apps for me to completely automate these steps.

Newer XviD stuff is actually just standard MPEG-4 ASP (or in some weird cases, SP), which means you can just plop it in a MP4 container. To do this I wrote a script that uses MP4Box to demux the MPEG-4 tracks from the AVI files (and in most cases, fix them since AVI is a horrible lovely non-standard hack of a container that doesn't properly handle things like B-frames), and then concatenate them.

To do this do

code:

mp4box -packed -nodrop avifile.avi -aviraw video -aviraw audio
mp4box -add avifile_video.cmp -add avifile_audio.mp3 -new mp4file.mp4

and if there's more than one file

code:

mp4box mp4file_a.mp4 -cat mp4file_b.mp4 (-cat mp4file_c.mp4, etc..)

which will tack on whatever files you specify after -cat to the first one.

This has the side benefit of being a standardized format, and if you've got an iPhone or iPad, will allow them to play these files normally without re-encoding anything (that is, if the audio track is AAC or MP3).

e: I'd post the script I wrote for this since it goes farther and re-encodes the audio (ffmpeg/LAME MP3 for stereo and multichannel AAC on OS X only) and tags the titles and whatever, but it's bash so its only useful for Mac/Linux (or Windows with Interix I guess)

Beeftweeter fucked around with this message at 01:24 on Jan 19, 2012

# ¿ Jan 19, 2012 01:17

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Soup in a Bag posted:

I'd use it if you post it. I have a script that remuxes MKV -> MP4 with MP4, but I haven't gotten around to adding audio conversion yet and I don't have anything for AVI.

Sure, but there's a few requirements:

MP4Box
ffmpeg and ffprobe (last time I checked ~2 months ago libav won't work since it doesn't support CAF)
(non-OS X only; install before ffmpeg) libfaac for AAC encoding
(non-OS X only; install before ffmpeg) LAME for MP3 encoding
mkvtoolnix for converting MKV files (particularly mkvextract and only needed if you're going to be using the script for that)
libsndfile and SoX for remixing AC3 or DTS to AAC

and some caveats:

Only looks for the first English subtitle track it can find, but this is easy enough to change (look for "sindex=$" in makemp4)
If a file has an AC3 or DTS track, it has to be remixed before being converted to AAC. afconvert on OS X will read AC3 files, but channel mapping is broken on files that already have a channel layout (read: pretty much everything), and ffmpeg's channel mapping is also currently broken so the script first dumps these formats to WAV and then remixes it with SoX (which can't read AC3 or DTS natively ) and outputs this to a CAF file since libsndfile doesn't support giving that a channel map (which is good because it'd be hosed up if ffmpeg/afconvert read it), and then converts this using afconvert on OS X or ffmpeg on whatever else. This means that converting a 5.1 track will temporarily take up about 1.4 GB per hour (so around 3 GB for a typical movie). It also means that this is going to take a while (about 10 minutes on a Q9550 @ 3.2GHz with a lovely 5400-7200 RPM variable speed HDD).
iOS 5 won't read MP4 files with MP3 audio tracks, so they have to be remuxed to MOV format. This isn't really a huge deal and only takes about 30 seconds or so, but if you're going to be streaming remuxed files to iOS devices change line 4 of the post-processing script to be "ios=1".
the post-processing script will only do MKV or AVI, but if you run the regular script it'll try to remux anything that ffmpeg can read (which is pretty much anything) as long as it has a H264/MPEG-4 video track.
I try to avoid XviD/AVI poo poo whenever I can, so I haven't tested the file merging part very much (although I have tested it...), and it'll only work with files in a series with either "cdx" or "a" "b" "c" or "d" on the end (ex. lovely.Movie.cd1.avi, lovely.Movie.cd2.avi and lovely.Movie.cd3.avi). I also didn't really spend a ton of time implementing it because it's not something I use very often, so it might not work as expected.
Both scripts are really, really messy since I basically just tacked on functionality whenever I needed it.

Download:

http://dl.dropbox.com/u/8706979/makemp4
http://dl.dropbox.com/u/8706979/makemp4-pp

do chmod +x on both files, then put makemp4 in /usr/local/bin and makemp4-pp in SABnzbd's scripts folder.

# ¿ Jan 20, 2012 07:44

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Anyone else getting this today?

quote:

usnews.blocknews.net: Failed to connect: This server does not allow SSL on this port 14@usnews.blocknews.net:443

# ¿ Feb 26, 2012 08:10

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Olivil posted:

I downloaded a bunch of files that went through SAB's scanning but still refuse to unarchive using The Unarchiver on a Mac, they are apparently corrupted. Tried the download again but no dice. Is there a way to repair the archives or should I simply move on?

Are they RAR files? Some of them have built-in redundancy info but as far as I know you need to use the command-line rar utility to actually repair them.

# ¿ Feb 28, 2012 04:38

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

crm posted:

Does anybody happen to have a script that will convert stuff sickbeard pulls down to an ipad friendly format?

I'd love to maintain a list of shows that after they are grabbed, it is converted to mp4 or whatever and dumped the "automatically add to itunes" directory.

I could write one, however I am lazy and would prefer using one that already exists

I posted this a few months ago:

amf5 posted:

Sure, but there's a few requirements:

MP4Box

ffmpeg and ffprobe (last time I checked ~2 months ago libav won't work since it doesn't support CAF)

(non-OS X only; install before ffmpeg) libfaac for AAC encoding

(non-OS X only; install before ffmpeg) LAME for MP3 encoding

mkvtoolnix for converting MKV files (particularly mkvextract and only needed if you're going to be using the script for that)

libsndfile and SoX for remixing AC3 or DTS to AAC

and some caveats:

Only looks for the first English subtitle track it can find, but this is easy enough to change (look for "sindex=$" in makemp4)

If a file has an AC3 or DTS track, it has to be remixed before being converted to AAC. afconvert on OS X will read AC3 files, but channel mapping is broken on files that already have a channel layout (read: pretty much everything), and ffmpeg's channel mapping is also currently broken so the script first dumps these formats to WAV and then remixes it with SoX (which can't read AC3 or DTS natively ) and outputs this to a CAF file since libsndfile doesn't support giving that a channel map (which is good because it'd be hosed up if ffmpeg/afconvert read it), and then converts this using afconvert on OS X or ffmpeg on whatever else. This means that converting a 5.1 track will temporarily take up about 1.4 GB per hour (so around 3 GB for a typical movie). It also means that this is going to take a while (about 10 minutes on a Q9550 @ 3.2GHz with a lovely 5400-7200 RPM variable speed HDD).

iOS 5 won't read MP4 files with MP3 audio tracks, so they have to be remuxed to MOV format. This isn't really a huge deal and only takes about 30 seconds or so, but if you're going to be streaming remuxed files to iOS devices change line 4 of the post-processing script to be "ios=1".

the post-processing script will only do MKV or AVI, but if you run the regular script it'll try to remux anything that ffmpeg can read (which is pretty much anything) as long as it has a H264/MPEG-4 video track.

I try to avoid XviD/AVI poo poo whenever I can, so I haven't tested the file merging part very much (although I have tested it...), and it'll only work with files in a series with either "cdx" or "a" "b" "c" or "d" on the end (ex. lovely.Movie.cd1.avi, lovely.Movie.cd2.avi and lovely.Movie.cd3.avi). I also didn't really spend a ton of time implementing it because it's not something I use very often, so it might not work as expected.

Both scripts are really, really messy since I basically just tacked on functionality whenever I needed it.

Download:

http://dl.dropbox.com/u/8706979/makemp4
http://dl.dropbox.com/u/8706979/makemp4-pp

do chmod +x on both files, then put makemp4 in /usr/local/bin and makemp4-pp in SABnzbd's scripts folder.

# ¿ Mar 19, 2012 23:44

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

crm posted:

Dang, I'll give this a shot.

Does it populate the metadata on the files? Can you specify a list of the shows that are done?

It doesn't, primarily because I haven't been able to find an easily parsable source for metadata that doesn't completely loving suck (Subler uses TheTVDB for example, and some of the descriptions are just horrendously bad/rife with errors).

I'm not sure what you mean by "a list of the shows that are done", but if you want to do multiple files the post-processing script will do that automatically, or if you want to do it outside of SABnzbd you can do something like

code:

for movie in *.mkv; do makemp4 "$movie"; done

# ¿ Mar 20, 2012 03:50

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

crm posted:

A couple issues with this -

1) it changes / removes the existing file, which is boooooooo
2) the generated an mp4, but got an error associated with ac3 audio instead of aac

Sorry for the late reply, I haven't been keeping up with this thread too much. Anyway, I had the script delete the input file by default because the file sizes are usually pretty similar and (beyond the audio codec and container format) they should be practically the same. Either way, you can easily tell the script to not delete anything: open up makemp4 in a text editor and scroll all the way to the bottom; put a "#" (no quotes) before the line that says rm "${infile}" &>/dev/null (10th line up from the bottom/line 340). Other than deleting it at the very end of processing the script does not make any changes to the input.

I'm not sure what you mean by the second thing, could you be more specific? It shouldn't be adding or trying to add an AC3 file to the MP4.

# ¿ Apr 4, 2012 09:31

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

I've been getting some strange issues with repairing failed downloads recently: things well inside of my server's retention have had missing parts (not that unusual), but when I click on retry, they'll actually re-download fully (sometimes from the same server :wtc:

), but SABnzbd doesn't recognize this. After taking a look in the temp directory I saw what's clearly partial bits of the file (ex. 38 MB of a 50 MB RAR from the first failure; again, not that unusual), but then the same file, downloaded completely, as (for example, part 30 of something) filename.part30.rar.1, filename.part30.rar.2, etc. for however many times I mashed the retry button. If I manually rename all of the redownloaded parts, they extract and verify against the PAR sets, so it looks like SABnzbd just can't figure out what to do with them for some reason. Has this happened to anyone else?

e: edited for clarity :downs:

Beeftweeter fucked around with this message at 00:34 on May 5, 2012

# ¿ May 5, 2012 00:30

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Nonsense, they're giving everyone unlimited data! :downs:

quote:

How has Comcast "killed" its caps? "Each of these pilot approaches will effectively offer unlimited usage of our services because customers will have the ability to buy as much data as they want."

# ¿ May 17, 2012 21:58

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Michael Scott posted:

~~http://www.downforeveryoneorjustme.com/nzbs.org~~ back up

What's the best alternative, would it be something like binsearch or nzbclub? Those just seem like low-rent nzbs.org

I still think their beta site is the best out of all of them (and free), but nzbs.org has been going down and up a lot lately, and now it's throwing out internal server errors

# ¿ Jul 16, 2012 22:21

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Hello Spaceman posted:

And now there's just a default Apache placeholder page. I hope they're just reconfiguring and the server wasn't rooted and wiped

Well, the 500s said the server was running nginx, so I guess it's possible they're just switching to Apache or something.

# ¿ Jul 19, 2012 03:48

Adbot: ADBOT LOVES YOU

# ¿ May 21, 2024 17:07

Beeftweeter: Jun 28, 2005; OFFICIAL #1 GNOME FAN

Does anyone here use the Xsnews IPv6 server (reader.ipv6.xsnews.nl)? I've been getting pretty erratic speeds (~165 KBps vs 2.5 MBps normally) and spotty connectivity lately and I was wondering if it's on their end or if there's something wrong with my IPv6 setup since I'm getting a little packet loss on pings.

# ¿ Apr 12, 2013 01:57

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Usenet Megathread - No invite discussion!