Efficient Usenet Downloader
Home Page: https://nzbget.net
License: GNU General Public License v2.0
NZBGet is a binary downloader, which downloads files from Usenet based on information given in nzb-files.
NZBGet is written in C++ and is known for its performance and efficiency.
NZBGet can run on almost any device - classic PC, NAS, media player, SAT-receiver, WLAN-router, etc. The download area provides precompiled binaries for Windows, macOS, Linux (compatible with many CPUs and platform variants), FreeBSD and Android. For other platforms the program can be compiled from sources.
Testing builds had svn revision number in their names, for example 15.0-testing-r1283, where r1283 is the revision number. GIt being a distributed VCS doesn't have such revision numbers.
However we still need something to distinguish the testing releases. At best the naming scheme should remain unchanged in order to provide compatibility with existing installations. This is necessary because the built-in update routine relies on existing naming scheme.
Version names for testing releases are automatically generated during make process. That's coded in Makefile.am.
Usually nzbs contain many small rar-files with size 50-200 MB.
If however an nzb contains files of very large size (many GB) that may cause NZBGet to crash or print an error "Could not create file ..." (depending on compiler and platform) when option DirectWrite is active.
I have NZBget running on some older hardware. Downloads work fine but unrar / unpack takes a bit of time. PAR repair can take extremely long, over 30 minutes, depending on how many files need to be repaired.
I have a faster computer on the same local network. Cat 5e speeds should be enough to transfer PARs and RARs between the two quickly, for processing and sending back to the slower PC to write the finished file locally.
Some users use NZBget on very slow hardware, like the Raspberry Pi. It would greatly improve speeds for those users.
Does this exist yet? Could it exist in the future?
A problem on Drobo NAS was reported. The program was sometimes hanging during par-repair when multithreading par-repair was active (option ParThreads).
After investigation it was found the program was hanging in the system function pthread_spin_lock.
The spin locks are alternative to mutexes and can provide performance gain in certain situations. In NZBGet spin locks are used only in two cases:
Spin locks can be always replaced by mutexes. Furthermore not all platforms support spin locks (for example Mac OS X doesn't and not every Linux system has them). NZBGet uses spin locks if available, otherwise it falls back to mutexes.
Although the problem was initially reported during par-repair the easier way to reproduce is:
If the system has problem with spin locks the download will soon stop.
Current file name for source distribution archive is like nzbget-15.0.tar.gz. The name is generated automatically by make dist.
Files uploaded to GitHub releases have src in their names, for example nzbget-15.0-src.tar.gz.
TODO: Check if the command make dist can be altered to automatically add src to file name.
When generating NZBs from e.g. binsearch, they all get the same filename and NZBget ignores all but the first. Even though they have completely different contents.
NZBget should ignore the filename of the container and just check whether the contents are known. Maybe through a checksum over the whole file or the list of the articles.
As reported on forum in topic bad parsing of filenames/quotes the filename extraction from subject may give wrong results for the following example nzb:
<file poster="[email protected] (yEncBin)" date="1435841477"
subject="e6fdc3b219f4009f3afef38e15e2b90a [18/19] - "e6fdc3b219f4009f3afef38e15e2b90a" yEnc (65/68)">
--/---
<file poster="[email protected] (yEncBin)" date="1435841466"
subject="ee60718a6aa77e8aa67e8f75c416c0ee [3/11] - "ee60718a6aa77e8aa67e8f75c416c0ee" yEnc (65/68)">Sometimes it may be helpful to be able to execute extension scripts from settings page. For example pp-script EMail.py could use a button "Send test email". For other scripts something like "Validate settings" or "Cleanup database" may be useful too.
From forum topic Different exit code on client command failure:
currently the client returns 0 when a connection error occurs, which makes it hard to use it in scripts. The following command:
$ /usr/bin/nzbget --configfile /etc/nzbgetd.conf --quit Unable to send request to nzbget-server at 127.0.0.1 (port 6789)returns 0, it should not.
Update links on download page to get releases from GitHub. That eliminates the 5 second wait banner from SourceForge.
Linux installer currently doesn't detect ARM 64 bit.
The architecture name is aarch64.
When clicking on nzb-files in Windows Explorer the files must be enqueued in NZBGet.
Current solution requires users to manually add a registry entry (with command nzbget.exe -A %1) and works only if NZBGet is currently running.
I have a TL-MR3040 router running OpenWRT (hw wiki link) that I was trying to get nzbget running on.
The autoinstaller correctly identified the arch as MIPS, but got the endianness wrong, so although it looked like it installed correctly, I was getting 'syntax error' on trying to run it. (looks like the shell was trying to run the binary as a script).
Fiddled around for a while but finding answers for any recent version is difficult (the openwrt opkg only comes with 0.8.0 circa 3 years ago, and the most recent other ipkg around is v11, but information about those is limited too), until I noticed that the supported architectures includes mipsel as well as mipseb.
Ran the installer with --arch=mipseb and everything now works swimmingly.
Not sure if this is something that can be detected by the installer or not, but maybe an extra ote in the docs would be great.
Following the request on forum. The feature was already implemented but the same is needed for RSS Fetch screen.
Rename "svn_version.cpp" and function inside.
Related to #3.
After migration of the web-site to GitHub the function Check for Updates often shows error message:
Check failed. Please try again later.
After a few attempts the check succeeds eventually.
Feature request via email:
The ability to have ‘two’ settings, or configuration snapshots for the “speed” screen.
So I can easily switch between two settings, back and forth (hopefully with just one click).
Switch between “slow” and “fast”. Maybe I want tone setting to be “none” and another to be 100 KB/s.
Or one to have two news servers, and other to have just one.
To easily change between them by just clicking the icon:
And If I click on the speed instead, I get the normal modal as shown above.
After a discussion agreed upon:
The program starts in "slow" mode as dictated by config options. If a start in "fast" mode is preferred that could be possible to achieve via scheduler task which changes download rate or active servers. This however requires an extension of scheduler to allow tasks fired on program start instead of on a specified time. This alone would be a nice improvement for the scheduler.
It's currently:
<input type="text" id="S_Server1_Password" value="" class="editsmall">It should be:
<input type="password" id="S_Server1_Password" value="" class="editsmall">If NZBGet is started on boot and the config contains network paths it's possible that NZBGet starts before the paths are mounted. That causes troubles.
Related forum topic - Auto update failed (Windows) (second page).
I can think of a new option to define the paths NZBGet checks on startup. If any of the paths doesn't exist it should wait few seconds and check again. All activities should be hold until all paths became available.
#Windows:
WaitForDir=E:\Downloads;D:\
#Linux:
WaitForDir=/mnt/nas1;/mnt/nas2
And since we can refer to other options in the config it can be used like this too:
WaitForDir=${DestDir};${NzbDir}
Directories listed in WaitForDir will not be automatically created.
Password protected rar4-archives do not have a password validation record like rar5-archives. If extraction of rar4-archives fail, it's not clear whether the password is incorrect or the archive is corrupted.
In such cases NZBGet performs par-check in an attempt to repair the archive. If the par-check reports status "success, no repair necessary" NZBGet prints an unpack error:
Unpack for failed: second unpack attempt skipped due to par-check not repaired anything
It seems the message is misleading because the users don't know in details how the archives are processed and they wonder: "why the hell the unpack was skipped and why par-check didn't repair anything?"
Change the message to something more understandable for normal users.
In response to request on forum the feature proposal:
After the rss feed is read from server and before it is processed the feed content will be saved to a temporary xml-file and a new script (new option "FeedScript") will be called. The script becomes the path to the feed file as parameter. NZBGet waits until the script terminates and then reads the xml-file and processes it
The update-scripts for Linux and Windows fetch updates from http://nzbget.net/download/xxx. This will not work with GitHub's Pages (if we migrate the web-site there). For flexibility the fetch-URL should be configured via info-file. The file currently holds info about version and date of the available update. Just add another field with fetch-URL and use it from the update-scripts.
The diskstate (queue, history, statistics) may become corrupted if system crashes (not nzbget crashes).
This may happen if the diskstate file were not written from system cache to disk. The disk caches can be forcibly flushed with special system functions. There is no cross-platform standard function for that but many systems have specific functions for that purpose:
Interesting articles:
It'd be nice if the README displayed on the repository start page would be better formatted using Markdown.
I have a fresh install of NZBget 15 on Debian, and i just ran into this issue: https://www.reddit.com/r/usenet/comments/2q0t2n/nzbget_download_stuck_at_ppqueued/
I can confirm that i had paused the downloads in between, but resumed, and made sure that the shown drop down field next to the pause button have the correct values.
Please add the following 3 commands to the documentation for Ubuntu 15.04/most linux releases to run for "./configure" with all options selected
sudo apt-get install libgnutls-dev
sudo apt-get install libncurses-dev
sudo apt-get install libxml2-dev
Find new hosting for forum and migrate the forum there.
There are way too many files in the project root directory, requiring GitHub visitors to scroll the front page to see readme content. Most of files are mandatory files required by autotools, for example:
Task: try to overcome autotools requirements and to eliminate as much files as possible or move them into a subdirectory.
Create web-site on GitHub pages:
See forum topic Sending emails failes for more details.
This file isn't used anymore and should be removed from the project.
Hi!
At the NZBget Forum you've announced some time ago a fix for the white spaces in the nzbclub.com url's provided by the rss feed.
Is this still in progress? Would be great, because ist seems to be one of the better free RSS providing search engines.
RSS search query: https://www.nzbclub.com/nzbrss.aspx?q=Debian+V7.6+64+bit+OS&st=5
Generated NZB URL: http://www.nzbclub.com/nzb_get/38563232/Debian V7 6 64 bit OS.nzb
Best regards
w00dst0ck
Move documentation from old wiki to GitHub. Most articles may require reformatting for better looking.
If the active URL download fails the program waits few seconds to retry. In that time an attempt to delete the URL doesn't work properly. The download stops but the web-ui shows activity. With the second attempt the URL can be deleted.
From forum topic Show Backup icon in History:
It would be perfect if the history list could also include the backup icon that sometimes appears on the download list - this would make it very easy to find any problem items.
I would find it useful to have a quick visual reference which items are causing problems.
If the computation is too costly could you remember if the Backup icon is displayed in the Download phase and then use that in the History?
When adding nzbs the nzb-files are sometimes rejected on early stages and are not added to history. They are also not processed by pp-scripts. The cases are for example:
Every nzb-request must result in a history record and must be post-processed.
The version number shown for Development in the Check for Updates-dialog should fetch the revision number from GitHub instead of old svn repository.
When installing update via built-in update routine (Windows and Linux installers) the program downloads new setup from NZBGet download area.
The downloaded file must be verified before execution.
It was bring to my attention, that unrar shipped with NZBGet performs too slow on Synology comparing to default unrar shipped with NAS.
After investigation it turned out the x86_64 version of unrar included within Installer was compiled in debug mode and without optimizations.
Linux build script (https://github.com/nzbget/nzbget/blob/develop/linux/build-nzbget) has to be updated to work with git.
From forum topic Keep track of age of oldest article:
Could NZBget keep track of the oldest article successfully downloaded from each server and display the number of days?
Server 1: 750d
Server 2: 2500d
Server 3: 150dCould it be added as information on the Server settings page or perhaps as a new column on the Speed Limit screen?
We also need a way to reset this statistic.
The temporary pause dialog currently accepts only minutes:
Being able to enter hours+minutes would be helpful. Two cases:
Forum topic Add Hours to Custom Pause Time.
The description field may contain html-text as CDATA. The content of the field must be cleaned up from html-tags before processed by feed filter. Otherwise the filter may trigger on html-tags.
From forum topic Blank gui page after quickly switching back and forward:
Using NZBget 15.0 (on Linux and Windows) and Opera or Firefox, when quickly switching back and forward between 'messages' and 'settings' tab results into a blank page, with only the top bar with 'messages' and 'settings' on it.
How about adding a dropdown like the rss feeds just with predefined searches ?
Have a image but no idea why github doesn't want to upload it.
Something went really wrong, and we can't process that file.
I guess we need this one.
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.