Merge branch 'refs/heads/t/wmadec'
[paraslash.git] / web / manual.md
1 **Paraslash user manual**
2
3 This document describes how to install, configure and use the paraslash
4 network audio streaming system. Most chapters start with a chapter
5 overview and conclude with an example section. We try to focus on
6 general concepts and on the interaction of the various pieces of the
7 paraslash package. Hence this user manual is not meant as a replacement
8 for the manual pages that describe all command line options of each
9 paraslash executable.
10
11 ============
12 Introduction
13 ============
14
15 In this chapter we give an [overview](#Overview) of the interactions of
16 the two main programs contained in the paraslash package, followed by
17 [brief descriptions](#The.paraslash.executables) of all executables.
18
19 Overview
20 --------
21
22 The core functionality of the para suite is provided by two main
23 executables, para_server and para_audiod. The former maintains a
24 database of audio files and streams these files to para_audiod which
25 receives and plays the stream.
26
27 In a typical setting, both para_server and para_audiod act as
28 background daemons whose functionality is controlled by client
29 programs: the para_audioc client controls para_audiod over a local
30 socket while the para_client program connects to para_server over a
31 local or remote networking connection.
32
33 Typically, these two daemons run on different hosts but a local setup
34 is also possible.
35
36 A simplified picture of a typical setup is as follows
37
38 server_host client_host
39 ~~~~~~~~~~~ ~~~~~~~~~~~
40
41 +-----------+ audio stream +-----------+
42 |para_server| -----------------------------> |para_audiod|
43 +-----------+ +-----------+
44 ^ ^
45 | |
46 | | connect
47 | |
48 | |
49 | +-----------+
50 | |para_audioc|
51 | +-----------+
52 |
53 |
54 | connect +-----------+
55 +-------------------------------------- |para_client|
56 +-----------+
57 The paraslash executables
58 -------------------------
59
60 ### para_server ###
61
62 para_server streams binary audio data (MP3, ...) over local and/or
63 remote networks. It listens on a TCP port and accepts commands such
64 as play, stop, pause, next from authenticated clients. There are
65 many more commands though, see the man page of para_server for a
66 description of all commands.
67
68 It supports three built-in network streaming protocols
69 (senders/receivers): HTTP, DCCP, or UDP. This is explained in more
70 detail in the section on [networking](#Networking).
71
72 The built-in audio file selector of paraslash is used to manage your
73 audio files. It maintains statistics on the usage of all available
74 audio files such as last-played time, and the number of times each
75 file was selected.
76
77 Additional information may be added to the database to allow
78 fine-grained selection based on various properties of the audio file,
79 including information found in (ID3) tags. However, old-fashioned
80 playlists are also supported.
81
82 It is also possible to store images (album covers) and lyrics in the
83 database and associate these to the corresponding audio files.
84
85 The section on the [audio file selector](#The.audio.file.selector)
86 discusses this topic.
87
88
89 ### para_client ###
90
91 The client program to connect to para_server. paraslash commands
92 are sent to para_server and the response is dumped to STDOUT. This
93 can be used by any scripting language to produce user interfaces with
94 little programming effort.
95
96 All connections between para_server and para_client are encrypted
97 with a symmetric session key. For each user of paraslash you must
98 create a public/secret RSA key pair for authentication.
99
100 If para_client is started without non-option arguments, an interactive
101 session (shell) is started. Command history and command completion are
102 supported through libreadline.
103
104 ### para_audiod ###
105
106 The local daemon that collects information from para_server.
107
108 It runs on the client side and connects to para_server. As soon as
109 para_server announces the availability of an audio stream, para_audiod
110 starts an appropriate receiver, any number of filters and a paraslash
111 writer to play the stream.
112
113 Moreover, para_audiod listens on a local socket and sends status
114 information about para_server and para_audiod to local clients on
115 request. Access via this local socket may be restricted by using Unix
116 socket credentials, if available.
117
118
119 ### para_audioc ###
120
121 The client program which talks to para_audiod. Used to control
122 para_audiod, to receive status info, or to grab the stream at any
123 point of the decoding process. Like para_client, para_audioc supports
124 interactive sessions on systems with libreadline.
125
126 ### para_recv ###
127
128 A command line HTTP/DCCP/UDP stream grabber. The http mode is
129 compatible with arbitrary HTTP streaming sources (e.g. icecast).
130 In addition to the three network streaming modes, para_recv can also
131 operate in local (afh) mode. In this mode it writes the content of
132 an audio file on the local file system in complete chunks to stdout,
133 optionally 'just in time'. This allows to cut an audio file without
134 first decoding it, and it enables third-party software which is unaware
135 of the particular audio format to send complete frames in real time.
136
137 ### para_filter ###
138
139 A filter program that reads from STDIN and writes to STDOUT.
140 Like para_recv, this is an atomic building block which can be used to
141 assemble higher-level audio receiving facilities. It combines several
142 different functionalities in one tool: decoders for multiple audio
143 formats and a number of processing filters, among these a normalizer
144 for audio volume.
145
146 ### para_afh ###
147
148 A small stand-alone program that prints tech info about the given
149 audio file to STDOUT. It can be instructed to print a "chunk table",
150 an array of offsets within the audio file.
151
152 ### para_write ###
153
154 A modular audio stream writer. It supports a simple file writer
155 output plug-in and optional WAV/raw players for ALSA (Linux) and for
156 coreaudio (Mac OS). para_write can also be used as a stand-alone WAV
157 or raw audio player.
158
159 ### para_play ###
160
161 A command line audio player.
162
163 ### para_gui ###
164
165 Curses-based gui that presents status information obtained in a curses
166 window. Appearance can be customized via themes. para_gui provides
167 key-bindings for the most common server commands and new key-bindings
168 can be added easily.
169
170 ### para_fade ###
171
172 An alarm clock and volume-fader for OSS and ALSA.
173
174 ===========
175 Quick start
176 ===========
177
178 This chapter lists the [necessary software](#Requirements)
179 that must be installed to compile the paraslash package, describes
180 how to [compile and install](#Installation) the paraslash
181 source code and the steps that have to be performed in order to
182 [set up](#Configuration) a typical server and client.
183
184 Requirements
185 ------------
186 ### For the impatient ###
187
188 git clone git://git.tuebingen.mpg.de/osl
189 cd osl && make && sudo make install && sudo ldconfig
190 sudo apt-get install autoconf libssl-dev help2man gengetopt \
191 libmad0-dev libid3tag0-dev libasound2-dev libvorbis-dev \
192 libfaad-dev libspeex-dev libFLAC-dev libsamplerate-dev realpath \
193 libasound2-dev libao-dev libreadline-dev libncurses-dev \
194 libopus-dev
195
196 ### Detailed description ###
197
198 In any case you will need
199
200 - [libosl](http://people.tuebingen.mpg.de/maan/osl/). The _object
201 storage layer_ library is used by para_server. To clone the source
202 code repository, execute
203
204 git clone git://git.tuebingen.mpg.de/osl
205
206 - [gcc](ftp://ftp.gnu.org/pub/gnu/gcc) or
207 [clang](http://clang.llvm.org). All gcc versions >= 4.2 are currently
208 supported. Clang version 1.1 or newer should work as well.
209
210 - [gnu make](ftp://ftp.gnu.org/pub/gnu/make) is also shipped with the
211 disto. On BSD systems the gnu make executable is often called gmake.
212
213 - [bash](ftp://ftp.gnu.org/pub/gnu/bash). Some scripts which run
214 during compilation require the _Bourne again shell_. It is most
215 likely already installed.
216
217 - [gengetopt](ftp://ftp.gnu.org/pub/gnu/gengetopt/) is needed to
218 generate the C code for the command line parsers of all paraslash
219 executables.
220
221 - [help2man](ftp://ftp.gnu.org/pub/gnu/help2man) is used to create
222 the man pages.
223
224 Optional:
225
226 - [openssl](http://www.openssl.org/) or
227 [libgcrypt](ftp://ftp.gnupg.org/gcrypt/libgcrypt/). At least one
228 of these two libraries is needed as the backend for cryptographic
229 routines on both the server and the client side. Both openssl and
230 libgcrypt are usually shipped with the distro, but you might have
231 to install the development package (`libssl-dev` or `libgcrypt-dev`
232 on debian systems) as well.
233
234 - [libmad](http://www.underbit.com/products/mad/). To compile in MP3
235 support for paraslash, the development package must be installed. It
236 is called `libmad0-dev` on debian-based systems. Note that libmad is
237 not necessary on the server side, i.e., for sending MP3 files.
238
239 - [libid3tag](http://www.underbit.com/products/mad/). For version-2
240 ID3 tag support, you willl need the libid3tag development package
241 `libid3tag0-dev`. Without libid3tag, only version-1 tags are
242 recognized. The mp3 tagger also needs this library for modifying
243 (id3v1 and id3v2) tags.
244
245 - [ogg vorbis](http://www.xiph.org/downloads/). For ogg vorbis streams
246 you need libogg, libvorbis, libvorbisfile. The corresponding Debian
247 packages are called `libogg-dev` and `libvorbis-dev`.
248
249 - [libfaad](http://www.audiocoding.com/). For aac files (m4a) you
250 need libfaad (`libfaad-dev`).
251
252 - [speex](http://www.speex.org/). In order to stream or decode speex
253 files, libspeex (`libspeex-dev`) is required.
254
255 - [flac](http://flac.sourceforge.net/). To stream or decode files
256 encoded with the _Free Lossless Audio Codec_, libFLAC (`libFLAC-dev`)
257 must be installed.
258
259 - [libsamplerate](http://www.mega-nerd.com/SRC/index.html). The
260 resample filter will only be compiled if this library is
261 installed. Debian package: `libsamplerate-dev`.
262
263 - [alsa-lib](ftp://ftp.alsa-project.org/pub/lib/). On Linux, you will
264 need to have the ALSA development package `libasound2-dev` installed.
265
266 - [libao](http://downloads.xiph.org/releases/ao/). Needed to build
267 the ao writer (ESD, PulseAudio,...). Debian package: `libao-dev`.
268
269 - [curses](ftp://ftp.gnu.org/pub/gnu/ncurses). Needed for
270 para_gui. Debian package: `libncurses-dev`.
271
272 - [GNU
273 Readline](http://cnswww.cns.cwru.edu/php/chet/readline/rltop.html). If
274 this library (`libreadline-dev`) is installed, para_client, para_audioc
275 and para_play support interactive sessions.
276
277 Installation
278 ------------
279 To build the sources from a tarball, execute
280
281 ./configure && make
282
283 To build from git or a gitweb snapshot, run this command instead:
284
285 ./autogen.sh
286
287 There should be no errors but probably some warnings about missing
288 packages which usually implies that not all audio formats will be
289 supported. If headers or libs are installed at unusual locations you
290 might need to tell the configure script where to find them. Try
291
292 ./configure --help
293
294 to see a list of options. If the paraslash package was compiled
295 successfully, execute (optionally)
296
297 make test
298
299 to run the paraslash test suite. If all tests pass, execute as root
300
301 make install
302
303 to install executables under /usr/local/bin and the man pages under
304 /usr/local/man.
305
306 Configuration
307 -------------
308
309 ### Create a paraslash user ###
310
311 In order to control para_server at runtime you must create a paraslash
312 user. As authentication is based on the RSA crypto system you'll have
313 to create an RSA key pair. If you already have a user and an RSA key
314 pair, you may skip this step.
315
316 In this section we'll assume a typical setup: You would like to run
317 para_server on some host called server_host as user foo, and you want
318 to connect to para_server from another machine called client_host as
319 user bar.
320
321 As foo@server_host, create ~/.paraslash/server.users by typing the
322 following commands:
323
324 user=bar
325 target=~/.paraslash/server.users
326 key=~/.paraslash/id_rsa.pub.$user
327 perms=AFS_READ,AFS_WRITE,VSS_READ,VSS_WRITE
328 mkdir -p ~/.paraslash
329 echo "user $user $key $perms" >> $target
330
331 Next, change to the "bar" account on client_host and generate the
332 key pair with the commands
333
334 ssh-keygen -q -t rsa -b 2048 -N '' -f $key
335
336 This generates the two files id_rsa and id_rsa.pub in ~/.ssh. Note
337 that para_server won't accept keys shorter than 2048 bits. Moreover,
338 para_client rejects private keys which are world-readable.
339
340 para_server only needs to know the public key of the key pair just
341 created. Copy this public key to server_host:
342
343 src=~/.ssh/id_rsa.pub
344 dest=.paraslash/id_rsa.pub.$LOGNAME
345 scp $src foo@server_host:$dest
346
347 Finally, tell para_client to connect to server_host:
348
349 conf=~/.paraslash/client.conf
350 echo 'hostname server_host' > $conf
351
352
353 ### Start para_server ###
354
355 For this first try, we'll use the info loglevel to make the output
356 of para_server more verbose.
357
358 para_server -l info
359
360 Now you can use para_client to connect to the server and issue
361 commands. Open a new shell as bar@client_host and try
362
363 para_client help
364 para_client si
365
366 to retrieve the list of available commands and some server info.
367 Don't proceed if this doesn't work.
368
369 ### Create and populate the database ###
370
371 An empty database is created with
372
373 para_client init
374
375 This initializes a couple of empty tables under
376 ~/.paraslash/afs_database-0.4. You normally don't need to look at these
377 tables, but it's good to know that you can start from scratch with
378
379 rm -rf ~/.paraslash/afs_database-0.4
380
381 in case something went wrong.
382
383 Next, you need to add some audio files to that database so that
384 para_server knows about them. Choose an absolute path to a directory
385 containing some audio files and add them to the audio file table:
386
387 para_client add /my/mp3/dir
388
389 This might take a while, so it is a good idea to start with a directory
390 containing not too many files. Note that the table only contains data
391 about the audio files found, not the files themselves.
392
393 You may print the list of all known audio files with
394
395 para_client ls
396
397 ### Configure para_audiod ###
398
399 We will have to tell para_audiod that it should receive the audio
400 stream from server_host via http:
401
402 para_audiod -l info -r '.:http -i server_host'
403
404 You should now be able to listen to the audio stream once para_server
405 starts streaming. To activate streaming, execute
406
407 para_client play
408
409 Since no playlist has been specified yet, the "dummy" mode which
410 selects all known audio files is activated automatically. See the
411 section on the [audio file selector](#The.audio.file.selector) for how
412 to use playlists and moods to specify which files should be streamed
413 in which order.
414
415 Troubleshooting
416 ---------------
417
418 If you receive a socket related error on server or audiod startup,
419 make sure you have write permissions to the /var/paraslash directory:
420
421 sudo chown $LOGNAME /var/paraslash
422
423 Alternatively, use the --afs-socket (para_server) or --socket
424 (para_audiod) option to specify a different socket pathname.
425
426 To identify streaming problems try to receive, decode and play the
427 stream manually using para_recv, para_filter and para_write as follows.
428 For simplicity we assume that you're running Linux/ALSA and that only
429 MP3 files have been added to the database.
430
431 para_recv -r 'http -i server_host' > file.mp3
432 # (interrupt with CTRL+C after a few seconds)
433 ls -l file.mp3 # should not be empty
434 para_filter -f mp3dec -f wav < file.mp3 > file.wav
435 ls -l file.wav # should be much bigger than file.mp3
436 para_write -w alsa < file.wav
437
438 Double check what is logged by para_server and use the --loglevel
439 option of para_recv, para_filter and para_write to increase verbosity.
440
441 ===============
442 User management
443 ===============
444
445 para_server uses a challenge-response mechanism to authenticate
446 requests from incoming connections, similar to ssh's public key
447 authentication method. Authenticated connections are encrypted using
448 a stream cipher, either RC4 or AES in integer counter mode.
449
450 In this chapter we briefly describe RSA, RC4 and AES, and sketch the
451 [authentication handshake](#Client-server.authentication)
452 between para_client and para_server. User management is discussed
453 in the section on [the user_list file](#The.user_list.file).
454 These sections are all about communication between the client and the
455 server. Connecting para_audiod is a different matter and is described
456 in a [separate section](#Connecting.para_audiod).
457
458 RSA, RC4, AES
459 -------------
460
461 RSA is an asymmetric block cipher which is used in many applications,
462 including ssh and gpg. An RSA key consists in fact of two keys,
463 called the public key and the private key. A message can be encrypted
464 with either key and only the counterpart of that key can decrypt
465 the message. While RSA can be used for both signing and encrypting
466 a message, paraslash uses RSA only for the latter purpose. The
467 RSA public key encryption and signatures algorithms are defined in
468 detail in RFC 2437.
469
470 RC4 is a stream cipher, i.e. the input is XORed with a pseudo-random
471 key stream to produce the output. Decryption uses the same function
472 calls as encryption. While RC4 supports variable key lengths,
473 paraslash uses a fixed length of 256 bits, which is considered a
474 strong encryption by today's standards. Since the same key must never
475 be used twice, a different, randomly-generated key is used for every
476 new connection.
477
478 AES, the advanced encryption standard, is a well-known symmetric block
479 cipher, i.e. a transformation operating on fixed-length blocks which
480 is determined by a single key for both encryption and decryption. Any
481 block cipher can be turned into a stream cipher by generating
482 a pseudo-random key stream by encrypting successive values of a
483 counter. The AES_CTR128 stream cipher used in paraslash is obtained
484 in this way from the AES block cipher with a 128 bit block size.
485
486 Client-server authentication
487 ----------------------------
488
489 The authentication handshake between para_client and para_server goes
490 as follows:
491
492 - para_client connects to para_server and sends an authentication
493 request for a user. It does so by connecting to TCP port 2990 of the
494 server host. This port is called the para_server _control port_.
495
496 - para_server accepts the connection and forks a child process which
497 handles the incoming request. The parent process keeps listening on the
498 control port while the child process (also called para_server below)
499 continues as follows.
500
501 - para_server loads the RSA public key of that user, fills a
502 fixed-length buffer with random bytes, encrypts that buffer using the
503 public key and sends the encrypted buffer to the client. The first
504 part of the buffer is the challenge which is used for authentication
505 while the second part is the session key.
506
507 - para_client receives the encrypted buffer and decrypts it with the
508 user's private key, thereby obtaining the challenge buffer and the
509 session key. It sends the SHA1 hash value of the challenge back to
510 para_server and stores the session key for further use.
511
512 - para_server also computes the SHA1 hash of the challenge and compares
513 it against what was sent back by the client.
514
515 - If the two hashes do not match, the authentication has failed and
516 para_server closes the connection.
517
518 - Otherwise the user is considered authenticated and the client is
519 allowed to proceed by sending a command to be executed. From this
520 point on the communication is encrypted using the stream cipher with
521 the session key known to both peers.
522
523 paraslash relies on the quality of the pseudo-random bytes provided
524 by the crypto library (openssl or libgcrypt), on the security of the
525 implementation of the RSA, RC4 and AES crypto routines and on the
526 infeasibility to invert the SHA1 function.
527
528 Neither para_server or para_client create RSA keys on their
529 own. This has to be done once for each user as sketched in
530 [Quick start](#Quick.start) and discussed in more detail
531 [below](#The.user_list.file).
532
533 The user_list file
534 ------------------
535
536 At startup para_server reads the user list file which contains one
537 line per user. The default location of the user list file may be
538 changed with the --user-list option.
539
540 There should be at least one user in this file. Each user must have
541 an RSA key pair. The public part of the key is needed by para_server
542 while the private key is needed by para_client. Each line of the
543 user list file must be of the form
544
545 user <username> <key> <perms>
546
547 where _username_ is an arbitrary string (usually the user's login
548 name), _key_ is the full path to that user's public RSA key, and
549 _perms_ is a comma-separated list of zero or more of the following
550 permission bits:
551
552 +---------------------------------------------------------+
553 | AFS_READ | read the contents of the databases |
554 +-----------+---------------------------------------------+
555 | AFS_WRITE | change database contents |
556 +-----------+---------------------------------------------+
557 | VSS_READ | obtain information about the current stream |
558 +-----------+---------------------------------------------+
559 | VSS_WRITE | change the current stream |
560 +---------------------------------------------------------+
561
562 The permission bits specify which commands the user is allowed to
563 execute. The output of
564
565 para_client help
566
567 contains in the third column the permissions needed to execute the
568 command.
569
570 It is possible to make para_server reread the user_list file by
571 executing the paraslash "hup" command or by sending SIGHUP to the
572 PID of para_server.
573
574 Connecting para_audiod
575 ----------------------
576
577 para_audiod listens on a Unix domain socket. Those sockets are
578 for local communication only, so only local users can connect to
579 para_audiod. The default is to let any user connect but this can be
580 restricted on platforms that support UNIX socket credentials which
581 allow para_audiod to obtain the Unix credentials of the connecting
582 process.
583
584 Use para_audiod's --user-allow option to allow connections only for
585 a limited set of users.
586
587 =======================
588 The audio file selector
589 =======================
590
591 paraslash comes with a sophisticated audio file selector (AFS),
592 whose main task is to determine which file to stream next, based on
593 information on the audio files stored in a database. It communicates
594 also with para_client whenever an AFS command is executed, for example
595 to answer a database query.
596
597 Besides the traditional playlists, AFS supports audio file selection
598 based on _moods_ which act as a filter that limits the set of all
599 known audio files to those which satisfy certain criteria. It also
600 maintains tables containing images (e.g. album cover art) and lyrics
601 that can be associated with one or more audio files.
602
603 AFS employs [libosl](http://people.tuebingen.mpg.de/maan/osl/), the
604 object storage layer library, as the backend library for storing
605 information on audio files, playlists, etc. This library offers
606 functionality similar to a relational database, but is much more
607 lightweight than a full database backend.
608
609 In this chapter we sketch the setup of the [AFS
610 process](#The.AFS.process) during server startup and proceed with the
611 description of the [layout](#Database.layout) of the various database
612 tables. The section on [playlists and moods](#Playlists.and.moods)
613 explains these two audio file selection mechanisms in detail
614 and contains pratical examples. The way [file renames and content
615 changes](#File.renames.and.content.changes) are detected is discussed
616 briefly before the [Troubleshooting](#Troubleshooting) section
617 concludes the chapter.
618
619 The AFS process
620 ---------------
621
622 On startup, para_server forks to create the AFS process which opens
623 the OSL database tables. The server process communicates with the
624 AFS process via pipes and shared memory. Usually, the AFS process
625 awakes only briefly whenever the current audio file changes. The AFS
626 process determines the next audio file, opens it, verifies it has
627 not been changed since it was added to the database and passes the
628 open file descriptor to the server process, along with audio file
629 meta-data such as file name, duration, audio format and so on. The
630 server process then starts to stream the audio file.
631
632 The AFS process also accepts connections from local clients via
633 a well-known socket. However, only child processes of para_server
634 may connect through this socket. All server commands that have the
635 AFS_READ or AFS_WRITE permission bits use this mechanism to query or
636 change the database.
637
638 Database layout
639 ---------------
640
641 ### The audio file table ###
642
643 This is the most important and usually also the largest table of the
644 AFS database. It contains the information needed to stream each audio
645 file. In particular the following data is stored for each audio file.
646
647 - SHA1 hash value of the audio file contents. This is computed once
648 when the file is added to the database. Whenever AFS selects this
649 audio file for streaming the hash value is recomputed and checked
650 against the value stored in the database to detect content changes.
651
652 - The time when this audio file was last played.
653
654 - The number of times the file has been played so far.
655
656 - The attribute bitmask.
657
658 - The image id which describes the image associated with this audio
659 file.
660
661 - The lyrics id which describes the lyrics associated with this
662 audio file.
663
664 - The audio format id (MP3, OGG, ...).
665
666 - An amplification value that can be used by the amplification filter
667 to pre-amplify the decoded audio stream.
668
669 - The chunk table. It describes the location and the timing of the
670 building blocks of the audio file. This is used by para_server to
671 send chunks of the file at appropriate times.
672
673 - The duration of the audio file.
674
675 - Tag information contained in the audio file (ID3 tags, Vorbis
676 comments, ...).
677
678 - The number of channels
679
680 - The encoding bitrate.
681
682 - The sampling frequency.
683
684 To add or refresh the data contained in the audio file table, the _add_
685 command is used. It takes the full path of either an audio file or a
686 directory. In the latter case, the directory is traversed recursively
687 and all files which are recognized as valid audio files are added to
688 the database.
689
690 ### The attribute table ###
691
692 The attribute table contains two columns, _name_ and _bitnum_. An
693 attribute is simply a name for a certain bit number in the attribute
694 bitmask of the audio file table.
695
696 Each of the 64 bits of the attribute bitmask can be set for each
697 audio file individually. Hence up to 64 different attributes may be
698 defined. For example, "pop", "rock", "blues", "jazz", "instrumental",
699 "german_lyrics", "speech", whatever. You are free to choose as
700 many attributes as you like and there are no naming restrictions
701 for attributes.
702
703 A new attribute "test" is created by
704
705 para_client addatt test
706 and
707 para_client lsatt
708
709 lists all available attributes. You can set the "test" attribute for
710 an audio file by executing
711
712 para_client setatt test+ /path/to/the/audio/file
713
714 Similarly, the "test" bit can be removed from an audio file with
715
716 para_client setatt test- /path/to/the/audio/file
717
718 Instead of a path you may use a shell wildcard pattern. The attribute
719 is applied to all audio files matching this pattern:
720
721 para_client setatt test+ '/test/directory/*'
722
723 The command
724
725 para_client -- ls -l=v
726
727 gives you a verbose listing of your audio files also showing which
728 attributes are set.
729
730 In case you wonder why the double-dash in the above command is needed:
731 It tells para_client to not interpret the options after the dashes. If
732 you find this annoying, just say
733
734 alias para='para_client --'
735
736 and be happy. In what follows we shall use this alias.
737
738 The "test" attribute can be dropped from the database with
739
740 para rmatt test
741
742 Read the output of
743
744 para help ls
745 para help setatt
746
747 for more information and a complete list of command line options to
748 these commands.
749
750 ### Blob tables ###
751
752 The image, lyrics, moods and playlists tables are all blob tables.
753 Blob tables consist of three columns each: The identifier which is
754 a positive non-negative number that is auto-incremented, the name
755 (an arbitrary string) and the content (the blob).
756
757 All blob tables support the same set of actions: cat, ls, mv, rm
758 and add. Of course, _add_ is used for adding new blobs to the table
759 while the other actions have the same meaning as the corresponding
760 Unix commands. The paraslash commands to perform these actions are
761 constructed as the concatenation of the table name and the action. For
762 example addimg, catimg, lsimg, mvimg, rmimg are the commands that
763 manipulate or query the image table.
764
765 The add variant of these commands is special as these commands read
766 the blob contents from stdin. To add an image to the image table the
767 command
768
769 para addimg image_name < file.jpg
770
771 can be used.
772
773 Note that the images and lyrics are not interpreted at all, and also
774 the playlist and the mood blobs are only investigated when the mood
775 or playlist is activated with the select command.
776
777 ### The score table ###
778
779 Unlike all other tables the contents of the score table remain in
780 memory and are never stored on disk. The score table contains two
781 columns: The SHA1 hash value (of an audio file) and its current
782 score.
783
784 However, only those files which are admissible for the current mood
785 or playlist are contained in the score table. The audio file selector
786 always chooses the row with the highest score as the file to stream
787 next. While doing so, it computes the new score and updates the
788 last_played and the num_played fields in the audio file table.
789
790 The score table is recomputed by the select command which loads a
791 mood or playlist. Audio files are chosen for streaming from the rows
792 of the score table on a highest-score-first basis.
793
794
795 Playlists and moods
796 -------------------
797
798 Playlists and moods offer two different ways of specifying the set of
799 admissible files. A playlist in itself describes a set of admissible
800 files. A mood, in contrast, describes the set of admissible files in
801 terms of attributes and other type of information available in the
802 audio file table. As an example, a mood can define a filename pattern,
803 which is then matched against the names of audio files in the table.
804
805 ### Playlists ###
806
807 Playlists are accommodated in the playlist table of the afs database,
808 using the aforementioned blob format for tables. A new playlist is
809 created with the addpl command by specifying the full (absolute)
810 paths of all desired audio files, separated by newlines. Example:
811
812 find /my/mp3/dir -name "*.mp3" | para addpl my_playlist
813
814 If _my_playlist_ already exists it is overwritten. To activate the
815 new playlist, execute
816
817 para select p/my_playlist
818
819 The audio file selector will assign scores to each entry of the list,
820 in descending order so that files will be selected in order. If a
821 file could not be opened for streaming, its entry is removed from
822 the score table (but not from the playlist).
823
824 ### Moods ###
825
826 A mood consists of a unique name and its *mood definition*, which is
827 a set of *mood lines* containing expressions in terms of attributes
828 and other data contained in the database.
829
830 At any time at most one mood can be *active* which means that
831 para_server is going to select only files from that subset of
832 admissible files.
833
834 So in order to create a mood definition one has to write a set of
835 mood lines. Mood lines come in three flavours: Accept lines, deny
836 lines and score lines.
837
838 The general syntax of the three types of mood lines is
839
840
841 accept [with score <score>] [if] [not] <mood_method> [options]
842 deny [with score <score>] [if] [not] <mood_method> [options]
843 score <score> [if] [not] <mood_method> [options]
844
845
846 Here <score> is either an integer or the string "random" which assigns
847 a random score to all matching files. The score value changes the
848 order in which admissible files are going to be selected, but is of
849 minor importance for this introduction.
850
851 So we concentrate on the first two forms, i.e. accept and deny
852 lines. As usual, everything in square brackets is optional, i.e.
853 accept/deny lines take the following form when ignoring scores:
854
855 accept [if] [not] <mood_method> [options]
856
857 and analogously for the deny case. The "if" keyword is only syntactic
858 sugar and has no function. The "not" keyword just inverts the result,
859 so the essence of a mood line is the mood method part and the options
860 following thereafter.
861
862 A *mood method* is realized as a function which takes an audio file
863 and computes a number from the data contained in the database.
864 If this number is non-negative, we say the file *matches* the mood
865 method. The file matches the full mood line if it either
866
867 - matches the mood method and the "not" keyword is not given,
868 or
869 - does not match the mood method, but the "not" keyword is given.
870
871 The set of admissible files for the whole mood is now defined as those
872 files which match at least one accept mood line, but no deny mood line.
873 More formally, an audio file F is admissible if and only if
874
875 (F ~ AL1 or F ~ AL2...) and not (F ~ DL1 or F ~ DN2 ...)
876
877 where AL1, AL2... are the accept lines, DL1, DL2... are the deny
878 lines and "~" means "matches".
879
880 The cases where no mood lines of accept/deny type are defined need
881 special treatment:
882
883 - Neither accept nor deny lines: This treats all files as
884 admissible (in fact, that is the definition of the dummy mood
885 which is activated automatically if no moods are available).
886
887 - Only accept lines: A file is admissible iff it matches at
888 least one accept line:
889
890 F ~ AL1 or F ~ AL2 or ...
891
892 - Only deny lines: A file is admissible iff it matches no
893 deny line:
894
895 not (F ~ DL1 or F ~ DN2 ...)
896
897
898
899 ### List of mood_methods ###
900
901 no_attributes_set
902
903 Takes no arguments and matches an audio file if and only if no
904 attributes are set.
905
906 is_set <attribute_name>
907
908 Takes the name of an attribute and matches iff that attribute is set.
909
910 path_matches <pattern>
911
912 Takes a filename pattern and matches iff the path of the audio file
913 matches the pattern.
914
915 artist_matches <pattern>
916 album_matches <pattern>
917 title_matches <pattern>
918 comment_matches <pattern>
919
920 Takes an extended regular expression and matches iff the text of the
921 corresponding tag of the audio file matches the pattern. If the tag
922 is not set, the empty string is matched against the pattern.
923
924 year ~ <num>
925 bitrate ~ <num>
926 frequency ~ <num>
927 channels ~ <num>
928 num_played ~ <num>
929 image_id ~ <num>
930 lyrics_id ~ <num>
931
932 Takes a comparator ~ of the set {<, =, <=, >, >=, !=} and a number
933 <num>. Matches an audio file iff the condition <val> ~ <num> is
934 satisfied where val is the corresponding value of the audio file
935 (value of the year tag, bitrate in kbit/s, etc.).
936
937 The year tag is special as its value is undefined if the audio file
938 has no year tag or the content of the year tag is not a number. Such
939 audio files never match. Another difference is the special treatment
940 if the year tag is a two-digit number. In this case either 1900 or
941 2000 is added to the tag value, depending on whether the number is
942 greater than 2000 plus the current year.
943
944
945 ### Mood usage ###
946
947 To create a new mood called "my_mood", write its definition into
948 some temporary file, say "tmpfile", and add it to the mood table
949 by executing
950
951 para addmood my_mood < tmpfile
952
953 If the mood definition is really short, you may just pipe it to the
954 client instead of using temporary files. Like this:
955
956 echo "$MOOD_DEFINITION" | para addmood my_mood
957
958 There is no need to keep the temporary file since you can always use
959 the catmood command to get it back:
960
961 para catmood my_mood
962
963 A mood can be activated by executing
964
965 para select m/my_mood
966
967 Once active, the list of admissible files is shown by the ls command
968 if the "-a" switch is given:
969
970 para ls -a
971
972
973 ### Example mood definition ###
974
975 Suppose you have defined attributes "punk" and "rock" and want to define
976 a mood containing only Punk-Rock songs. That is, an audio file should be
977 admissible if and only if both attributes are set. Since
978
979 punk and rock
980
981 is obviously the same as
982
983 not (not punk or not rock)
984
985 (de Morgan's rule), a mood definition that selects only Punk-Rock
986 songs is
987
988 deny if not is_set punk
989 deny if not is_set rock
990
991
992
993 File renames and content changes
994 --------------------------------
995
996 Since the audio file selector knows the SHA1 of each audio file that
997 has been added to the afs database, it recognizes if the content of
998 a file has changed, e.g. because an ID3 tag was added or modified.
999 Also, if a file has been renamed or moved to a different location,
1000 afs will detect that an entry with the same hash value already exists
1001 in the audio file table.
1002
1003 In both cases it is enough to just re-add the new file. In the
1004 first case (file content changed), the audio table is updated, while
1005 metadata such as the num_played and last_played fields, as well as
1006 the attributes, remain unchanged. In the other case, when the file
1007 is moved or renamed, only the path information is updated, all other
1008 data remains as before.
1009
1010 It is possible to change the behaviour of the add command by using the
1011 "-l" (lazy add) or the "-f" (force add) option.
1012
1013 Troubleshooting
1014 ---------------
1015
1016 Use the debug loglevel (-l debug) to show debugging info. All paraslash
1017 executables have a brief online help which is displayed when -h is
1018 given. The --detailed-help option prints the full help text.
1019
1020 If para_server crashed or was killed by SIGKILL (signal 9), it
1021 may refuse to start again because of "dirty osl tables". In this
1022 case you'll have to run the oslfsck program of libosl to fix your
1023 database:
1024
1025 oslfsck -fd ~/.paraslash/afs_database-0.4
1026
1027 However, make sure para_server isn't running before executing oslfsck.
1028
1029 If you don't mind to recreate your database you can start
1030 from scratch by removing the entire database directory, i.e.
1031
1032 rm -rf ~/.paraslash/afs_database-0.4
1033
1034 Be aware that this removes all attribute definitions, all playlists
1035 and all mood definitions and requires to re-initialize the tables.
1036
1037 Although oslfsck fixes inconsistencies in database tables it doesn't
1038 care about the table contents. To check for invalid table contents, use
1039
1040 para_client check
1041
1042 This prints out references to missing audio files as well as invalid
1043 playlists and mood definitions.
1044
1045 Similarly, para_audiod refuses to start if its socket file exists, since
1046 this indicates that another instance of para_audiod is running. After
1047 a crash a stale socket file might remain and you must run
1048
1049 para_audiod --force
1050
1051 once to fix it up.
1052
1053 =======================================
1054 Audio formats and audio format handlers
1055 =======================================
1056
1057 Audio formats
1058 -------------
1059
1060 The following audio formats are supported by paraslash:
1061
1062 ### MP3 ###
1063
1064 Mp3, MPEG-1 Audio Layer 3, is a common audio format for audio storage,
1065 designed as part of its MPEG-1 standard. An MP3 file is made up of
1066 multiple MP3 frames, which consist of a header and a data block. The
1067 size of an MP3 frame depends on the bit rate and on the number
1068 of channels. For a typical CD-audio file (sample rate of 44.1 kHz
1069 stereo), encoded with a bit rate of 128 kbit, an MP3 frame is about
1070 400 bytes large.
1071
1072 ### OGG/Vorbis ###
1073
1074 OGG is a standardized audio container format, while Vorbis is an
1075 open source codec for lossy audio compression. Since Vorbis is most
1076 commonly made available via the OGG container format, it is often
1077 referred to as OGG/Vorbis. The OGG container format divides data into
1078 chunks called OGG pages. A typical OGG page is about 4KB large. The
1079 Vorbis codec creates variable-bitrate (VBR) data, where the bitrate
1080 may vary considerably.
1081
1082 ### OGG/Speex ###
1083
1084 Speex is an open-source speech codec that is based on CELP (Code
1085 Excited Linear Prediction) coding. It is designed for voice
1086 over IP applications, has modest complexity and a small memory
1087 footprint. Wideband and narrowband (telephone quality) speech are
1088 supported. As for Vorbis audio, Speex bit-streams are often stored
1089 in OGG files. As of 2012 this codec is considered obsolete since the
1090 Oppus codec, described below, surpasses its performance in all areas.
1091
1092 ### OGG/Opus ###
1093
1094 Opus is a lossy audio compression format standardized through RFC
1095 6716 in 2012. It combines the speech-oriented SILK codec and the
1096 low-latency CELT (Constrained Energy Lapped Transform) codec. Like
1097 OGG/Vorbis and OGG/Speex, Opus data is usually encapsulated in OGG
1098 containers. All known software patents which cover Opus are licensed
1099 under royalty-free terms.
1100
1101 ### AAC ###
1102
1103 Advanced Audio Coding (AAC) is a standardized, lossy compression
1104 and encoding scheme for digital audio which is the default audio
1105 format for Apple's iPhone, iPod, iTunes. Usually MPEG-4 is used as
1106 the container format and audio files encoded with AAC have the .m4a
1107 extension. A typical AAC frame is about 700 bytes large.
1108
1109 ### WMA ###
1110
1111 Windows Media Audio (WMA) is an audio data compression technology
1112 developed by Microsoft. A WMA file is usually encapsulated in the
1113 Advanced Systems Format (ASF) container format, which also specifies
1114 how meta data about the file is to be encoded. The bit stream of WMA
1115 is composed of superframes, each containing one or more frames of
1116 2048 samples. For 16 bit stereo a WMA superframe is about 8K large.
1117
1118 ### FLAC ###
1119
1120 The Free Lossless Audio Codec (FLAC) compresses audio without quality
1121 loss. It gives better compression ratios than a general purpose
1122 compressor like zip or bzip2 because FLAC is designed specifically
1123 for audio. A FLAC-encoded file consists of frames of varying size, up
1124 to 16K. Each frame starts with a header that contains all information
1125 necessary to decode the frame.
1126
1127 Meta data
1128 ---------
1129
1130 Unfortunately, each audio format has its own conventions how meta
1131 data is added as tags to the audio file.
1132
1133 For MP3 files, ID3, version 1 and 2 are widely used. ID3 version 1
1134 is rather simple but also very limited as it supports only artist,
1135 title, album, year and comment tags. Each of these can only be at most
1136 32 characters long. ID3, version 2 is much more flexible but requires
1137 a separate library being installed for paraslash to support it.
1138
1139 Ogg vorbis, ogg speex and flac files contain meta data as Vorbis
1140 comments, which are typically implemented as strings of the form
1141 "[TAG]=[VALUE]". Unlike ID3 version 1 tags, one may use whichever
1142 tags are appropriate for the content.
1143
1144 AAC files usually use the MPEG-4 container format for storing meta
1145 data while WMA files wrap meta data as special objects within the
1146 ASF container format.
1147
1148 paraslash only tracks the most common tags that are supported by
1149 all tag variants: artist, title, year, album, comment. When a file
1150 is added to the AFS database, the meta data of the file is extracted
1151 and stored in the audio file table.
1152
1153 Chunks and chunk tables
1154 -----------------------
1155
1156 paraslash uses the word "chunk" as common term for the building blocks
1157 of an audio file. For MP3 files, a chunk is the same as an MP3 frame,
1158 while for OGG files a chunk is an OGG page, etc. Therefore the chunk
1159 size varies considerably between audio formats, from a few hundred
1160 bytes (MP3) up to 16K (FLAC).
1161
1162 The chunk table contains the offsets within the audio file that
1163 correspond to the chunk boundaries of the file. Like the meta data,
1164 the chunk table is computed and stored in the database whenever an
1165 audio file is added.
1166
1167 The paraslash senders (see below) always send complete chunks. The
1168 granularity for seeking is therefore determined by the chunk size.
1169
1170 Audio format handlers
1171 ---------------------
1172
1173 For each audio format paraslash contains an audio format handler whose
1174 first task is to tell whether a given file is a valid audio file of
1175 this type. If so, the audio file handler extracts some technical data
1176 (duration, sampling rate, number of channels etc.), computes the
1177 chunk table and reads the meta data.
1178
1179 The audio format handler code is linked into para_server and executed
1180 via the _add_ command. The same code is also available as a stand-alone
1181 tool, para_afh, which prints the technical data, the chunk table
1182 and the meta data of a file. Moreover, all audio format handlers are
1183 combined in the afh receiver which is part of para_recv and para_play.
1184
1185 ==========
1186 Networking
1187 ==========
1188
1189 Paraslash uses different network connections for control and data.
1190 para_client communicates with para_server over a dedicated TCP control
1191 connection. To transport audio data, separate data connections are
1192 used. For these data connections, a variety of transports (UDP, DCCP,
1193 HTTP) can be chosen.
1194
1195 The chapter starts with the [control
1196 service](#The.paraslash.control.service), followed by a section
1197 on the various [streaming protocols](#Streaming.protocols)
1198 in which the data connections are described. The way
1199 audio file headers are embedded into the stream is discussed
1200 [briefly](#Streams.with.headers.and.headerless.streams) before the
1201 [example section](#Networking.examples) which illustrates typical
1202 commands for real-life scenarios.
1203
1204 Both IPv4 and IPv6 are supported.
1205
1206 The paraslash control service
1207 -----------------------------
1208
1209 para_server is controlled at runtime via the paraslash control
1210 connection. This connection is used for server commands (play, stop,
1211 ...) as well as for afs commands (ls, select, ...).
1212
1213 The server listens on a TCP port and accepts connections from clients
1214 that connect the open port. Each connection causes the server to fork
1215 off a client process which inherits the connection and deals with that
1216 client only. In this classical accept/fork approach the server process
1217 is unaffected if the child dies or goes crazy for whatever reason. In
1218 fact, the child process can not change address space of server process.
1219
1220 The section on [client-server
1221 authentication](#Client-server.authentication) above described the
1222 early connection establishment from the crypto point of view. Here
1223 it is described what happens after the connection (including crypto
1224 setup) has been established. There are four processes involved during
1225 command dispatch as sketched in the following diagram.
1226
1227 server_host client_host
1228 ~~~~~~~~~~~ ~~~~~~~~~~~
1229
1230 +-----------+ connect +-----------+
1231 |para_server|<------------------------------ |para_client|
1232 +-----------+ +-----------+
1233 | ^
1234 | fork +---+ |
1235 +----------> |AFS| |
1236 | +---+ |
1237 | ^ |
1238 | | |
1239 | | connect (cookie) |
1240 | | |
1241 | | |
1242 | fork +-----+ inherited connection |
1243 +---------->|child|<--------------------------+
1244 +-----+
1245
1246 Note that the child process is not a child of the afs process,
1247 so communication of these two processes has to happen via local
1248 sockets. In order to avoid abuse of the local socket by unrelated
1249 processes, a magic cookie is created once at server startup time just
1250 before the server process forks off the AFS process. This cookie is
1251 known to the server, AFS and the child, but not to unrelated processes.
1252
1253 There are two different kinds of commands: First there are commands
1254 that cause the server to respond with some answer such as the list
1255 of all audio files. All but the addblob commands (addimg, addlyr,
1256 addpl, addmood) are of this kind. The addblob commands add contents
1257 to the database, so they need to transfer data the other way round,
1258 from the client to the server.
1259
1260 There is no knowledge about the server commands built into para_client,
1261 so it does not know about addblob commands. Instead, the server sends
1262 a special "awaiting data" packet for these commands. If the client
1263 receives this packet, it sends STDIN to the server, otherwise it
1264 dumps data from the server to STDOUT.
1265
1266 Streaming protocols
1267 -------------------
1268
1269 A network (audio) stream usually consists of one streaming source,
1270 the _sender_, and one or more _receivers_ which read data over the
1271 network from the streaming source.
1272
1273 Senders are thus part of para_server while receivers are part of
1274 para_audiod. Moreover, there is the stand-alone tool para_recv which
1275 can be used to manually download a stream, either from para_server
1276 or from a web-based audio streaming service.
1277
1278 The following three streaming protocols are supported by paraslash:
1279
1280 - HTTP. Recommended for public streams that can be played by any
1281 player like mpg123, xmms, itunes, winamp, etc. The HTTP sender is
1282 supported on all operating systems and all platforms.
1283
1284 - DCCP. Recommended for LAN streaming. DCCP is currently available
1285 only for Linux.
1286
1287 - UDP. Recommended for multicast LAN streaming.
1288
1289 See the Appendix on [network protocols](/#Network.protocols)
1290 for brief descriptions of the various protocols relevant for network
1291 audio streaming with paraslash.
1292
1293 It is possible to activate more than one sender simultaneously.
1294 Senders can be controlled at run time and via config file and command
1295 line options.
1296
1297 Note that audio connections are _not_ encrypted. Transport or Internet
1298 layer encryption should be used if encrypted data connections are
1299 needed.
1300
1301 Since DCCP and TCP are both connection-oriented protocols, connection
1302 establishment/teardown and access control are very similar between
1303 these two streaming protocols. UDP is the most lightweight option,
1304 since in contrast to TCP/DCCP it is connectionless. It is also the
1305 only protocol supporting IP multicast.
1306
1307 The HTTP and the DCCP sender listen on a (TCP/DCCP) port waiting for
1308 clients to connect and establish a connection via some protocol-defined
1309 handshake mechanism. Both senders maintain two linked lists each:
1310 The list of all clients which are currently connected, and the list
1311 of access control entries which determines who is allowed to connect.
1312 IP-based access control may be configured through config file and
1313 command line options and via the "allow" and "deny" sender subcommands.
1314
1315 Upon receiving a GET request from the client, the HTTP sender sends
1316 back a status line and a message. The body of this message is the
1317 audio stream. This is common practice and is supported by many popular
1318 clients which can thus be used to play a stream offered by para_server.
1319 For DCCP things are a bit simpler: No messages are exchanged between
1320 the receiver and sender. The client simply connects and the sender
1321 starts to stream.
1322
1323 DCCP is an experimental protocol which offers a number of new features
1324 not available for TCP. Both ends can negotiate these features using
1325 a built-in negotiation mechanism. In contrast to TCP/HTTP, DCCP is
1326 datagram-based (no retransmissions) and thus should not be used over
1327 lossy media (e.g. WiFi networks). One useful feature offered by DCCP
1328 is access to a variety of different congestion-control mechanisms
1329 called CCIDs. Two different CCIDs are available per default on Linux:
1330
1331
1332 - _CCID 2_. A Congestion Control mechanism similar to that of TCP. The
1333 sender maintains a congestion window and halves this window in response
1334 to congestion.
1335
1336
1337 - _CCID-3_. Designed to be fair when competing for bandwidth.
1338 It has lower variation of throughput over time compared with TCP,
1339 which makes it suitable for streaming media.
1340
1341 Unlike the HTTP and DCCP senders, the UDP sender maintains only a
1342 single list, the _target list_. This list describes the set of clients
1343 to which the stream is sent. There is no list for access control and
1344 no "allow" and "deny" commands for the UDP sender. Instead, the "add"
1345 and "delete" commands can be used to modify the target list.
1346
1347 Since both UDP and DCCP offer an unreliable datagram-based transport,
1348 additional measures are necessary to guard against disruptions over
1349 networks that are lossy or which may be subject to interference (as
1350 is for instance the case with WiFi). Paraslash uses FEC (Forward
1351 Error Correction) to guard against packet losses and reordering. The
1352 stream is FEC-encoded before it is sent through the UDP socket and
1353 must be decoded accordingly on the receiver side.
1354
1355 The packet size and the amount of redundancy introduced by FEC can
1356 be configured via the FEC parameters which are dictated by server
1357 and may also be configured through the "sender" command. The FEC
1358 parameters are encoded in the header of each network packet, so no
1359 configuration is necessary on the receiver side. See the section on
1360 [FEC](#Forward.error.correction) below.
1361
1362 Streams with headers and headerless streams
1363 -------------------------------------------
1364
1365 For OGG/Vorbis, OGG/Speex and wma streams, some of the information
1366 needed to decode the stream is only contained in the audio file
1367 header of the container format but not in each data chunk. Clients
1368 must be able to obtain this information in case streaming starts in
1369 the middle of the file or if para_audiod is started while para_server
1370 is already sending a stream.
1371
1372 This is accomplished in different ways, depending on the streaming
1373 protocol. For connection-oriented streams (HTTP, DCCP) the audio file
1374 header is sent prior to audio file data. This technique however does
1375 not work for the connectionless UDP transport. Hence the audio file
1376 header is periodically being embedded into the UDP audio data stream.
1377 By default, the header is resent after five seconds. The receiver has
1378 to wait until the next header arrives before it can start decoding
1379 the stream.
1380
1381 Networking examples
1382 -------------------
1383
1384 The "si" (server info) command lists some information about the
1385 currently running server process.
1386
1387 -> Show PIDs, number of connected clients, uptime, and more:
1388
1389 para_client si
1390
1391 The sender command of para_server prints information about senders,
1392 like the various access control lists, and it allows to (de-)activate
1393 senders and to change the access permissions at runtime.
1394
1395 -> List all senders
1396
1397 para_client sender
1398
1399 -> Obtain general help for the sender command:
1400
1401 para_client help sender
1402
1403 -> Get help for a specific sender (contains further examples):
1404
1405 s=http # or dccp or udp
1406 para_client sender $s help
1407
1408 -> Show status of the http sender
1409
1410 para_client sender http status
1411
1412 By default para_server activates both the HTTP and th DCCP sender on
1413 startup. This can be changed via command line options or para_server's
1414 config file.
1415
1416 -> List config file options for senders:
1417
1418 para_server -h
1419
1420 All senders share the "on" and "off" commands, so senders may be
1421 activated and deactivated independently of each other.
1422
1423 -> Switch off the http sender:
1424
1425 para_client sender http off
1426
1427 -> Receive a DCCP stream using CCID2 and write the output into a file:
1428
1429 host=foo.org; ccid=2; filename=bar
1430 para_recv --receiver "dccp --host $host --ccid $ccid" > $filename
1431
1432 Note the quotes around the arguments for the dccp receiver. Each
1433 receiver has its own set of command line options and its own command
1434 line parser, so arguments for the dccp receiver must be protected
1435 from being interpreted by para_recv.
1436
1437 -> Start UDP multicast, using the default multicast address:
1438
1439 para_client sender udp add 224.0.1.38
1440
1441 -> Receive FEC-encoded multicast stream and write the output into a file:
1442
1443 filename=foo
1444 para_recv -r udp > $filename
1445
1446 -> Add an UDP unicast for a client to the target list of the UDP sender:
1447
1448 t=client.foo.org
1449 para_client sender udp add $t
1450
1451 -> Receive this (FEC-encoded) unicast stream:
1452
1453 filename=foo
1454 para_recv -r 'udp -i 0.0.0.0' > $filename
1455
1456 -> Create a minimal config for para_audiod for HTTP streams:
1457
1458 c=$HOME/.paraslash/audiod.conf.min; s=server.foo.com
1459 echo receiver \".:http -i $s\" > $c
1460 para_audiod --config $c
1461
1462 =======
1463 Filters
1464 =======
1465
1466 A paraslash filter is a module which transforms an input stream into
1467 an output stream. Filters are included in the para_audiod executable
1468 and in the stand-alone tool para_filter which usually contains the
1469 same modules.
1470
1471 While para_filter reads its input stream from STDIN and writes
1472 the output to STDOUT, the filter modules of para_audiod are always
1473 connected to a receiver which produces the input stream and a writer
1474 which absorbs the output stream.
1475
1476 Some filters depend on a specific library and are not compiled in
1477 if this library was not found at compile time. To see the list of
1478 supported filters, run para_filter and para_audiod with the --help
1479 option. The output looks similar to the following:
1480
1481 Available filters:
1482 compress wav amp fecdec wmadec prebuffer oggdec aacdec mp3dec
1483
1484 Out of these filter modules, a chain of filters can be constructed,
1485 much in the way Unix pipes can be chained, and analogous to the use
1486 of modules in gstreamer: The output of the first filter becomes the
1487 input of the second filter. There is no limitation on the number of
1488 filters and the same filter may occur more than once.
1489
1490 Like receivers, each filter has its own command line options which
1491 must be quoted to protect them from the command line options of
1492 the driving application (para_audiod or para_filter). Example:
1493
1494 para_filter -f 'mp3dec --ignore-crc' -f 'compress --damp 1'
1495
1496 For para_audiod, each audio format has its own set of filters. The
1497 name of the audio format for which the filter should be applied can
1498 be used as the prefix for the filter option. Example:
1499
1500 para_audiod -f 'mp3:prebuffer --duration 300'
1501
1502 The "mp3" prefix above is actually interpreted as a POSIX extended
1503 regular expression. Therefore
1504
1505 para_audiod -f '.:prebuffer --duration 300'
1506
1507 activates the prebuffer filter for all supported audio formats (because
1508 "." matches all audio formats) while
1509
1510 para_audiod -f 'wma|ogg:prebuffer --duration 300'
1511
1512 activates it only for wma and ogg streams.
1513
1514 Decoders
1515 --------
1516
1517 For each supported audio format there is a corresponding filter
1518 which decodes audio data in this format to 16 bit PCM data which
1519 can be directly sent to the sound device or any other software that
1520 operates on undecoded PCM data (visualizers, equalizers etc.). Such
1521 filters are called _decoders_ in general, and xxxdec is the name of
1522 the paraslash decoder for the audio format xxx. For example, the mp3
1523 decoder is called mp3dec.
1524
1525 Note that the output of the decoder is about 10 times larger than
1526 its input. This means that filters that operate on the decoded audio
1527 stream have to deal with much more data than filters that transform
1528 the audio stream before it is fed to the decoder.
1529
1530 Paraslash relies on external libraries for most decoders, so these
1531 libraries must be installed for the decoder to be included in the
1532 executables. For example, the mp3dec filter depends on the mad library.
1533
1534 Forward error correction
1535 ------------------------
1536
1537 As already mentioned [earlier](#Streaming.protocols), paraslash
1538 uses forward error correction (FEC) for the unreliable UDP and
1539 DCCP transports. FEC is a technique which was invented already in
1540 1960 by Reed and Solomon and which is widely used for the parity
1541 calculations of storage devices (RAID arrays). It is based on the
1542 algebraic concept of finite fields, today called Galois fields, in
1543 honour of the mathematician Galois (1811-1832). The FEC implementation
1544 of paraslash is based on code by Luigi Rizzo.
1545
1546 Although the details require a sound knowledge of the underlying
1547 mathematics, the basic idea is not hard to understand: For positive
1548 integers k and n with k < n it is possible to compute for any k given
1549 data bytes d_1, ..., d_k the corresponding r := n -k parity bytes p_1,
1550 ..., p_r such that all data bytes can be reconstructed from *any*
1551 k bytes of the set
1552
1553 {d_1, ..., d_k, p_1, ..., p_r}.
1554
1555 FEC-encoding for unreliable network transports boils down to slicing
1556 the audio stream into groups of k suitably sized pieces called _slices_
1557 and computing the r corresponding parity slices. This step is performed
1558 in para_server which then sends both the data and the parity slices
1559 over the unreliable network connection. If the client was able
1560 to receive at least k of the n = k + r slices, it can reconstruct
1561 (FEC-decode) the original audio stream.
1562
1563 From these observations it is clear that there are three different
1564 FEC parameters: The slice size, the number of data slices k, and the
1565 total number of slices n. It is crucial to choose the slice size
1566 such that no fragmentation of network packets takes place because
1567 FEC only guards against losses and reordering but fails if slices are
1568 received partially.
1569
1570 FEC decoding in paralash is performed through the fecdec filter which
1571 usually is the first filter (there can be other filters before fecdec
1572 if these do not alter the audio stream).
1573
1574 Volume adjustment (amp and compress)
1575 ------------------------------------
1576
1577 The amp and the compress filter both adjust the volume of the audio
1578 stream. These filters operate on uncompressed audio samples. Hence
1579 they are usually placed directly after the decoding filter. Each
1580 sample is multiplied with a scaling factor (>= 1) which makes amp
1581 and compress quite expensive in terms of computing power.
1582
1583 ### amp ###
1584
1585 The amp filter amplifies the audio stream by a fixed scaling factor
1586 that must be known in advance. For para_audiod this factor is derived
1587 from the amplification field of the audio file's entry in the audio
1588 file table while para_filter uses the value given at the command line.
1589
1590 The optimal scaling factor F for an audio file is the largest real
1591 number F >= 1 such that after multiplication with F all samples still
1592 fit into the sample interval [-32768, 32767]. One can use para_filter
1593 in combination with the sox utility to compute F:
1594
1595 para_filter -f mp3dec -f wav < file.mp3 | sox -t wav - -e stat -v
1596
1597 The amplification value V which is stored in the audio file table,
1598 however, is an integer between 0 and 255 which is connected to F
1599 through the formula
1600
1601 V = (F - 1) * 64.
1602
1603 To store V in the audio file table, the command
1604
1605 para_client -- touch -a=V file.mp3
1606
1607 is used. The reader is encouraged to write a script that performs
1608 these computations :)
1609
1610 ### compress ###
1611
1612 Unlike the amplification filter, the compress filter adjusts the volume
1613 of the audio stream dynamically without prior knowledge about the peak
1614 value. It maintains the maximal volume of the last n samples of the
1615 audio stream and computes a suitable amplification factor based on that
1616 value and the various configuration options. It tries to chose this
1617 factor such that the adjusted volume meets the desired target level.
1618
1619 Note that it makes sense to combine amp and compress.
1620
1621 Misc filters (wav and prebuffer)
1622 --------------------------------
1623
1624 These filters are rather simple and do not modify the audio stream at
1625 all. The wav filter is only useful with para_filter and in connection
1626 with a decoder. It asks the decoder for the number of channels and the
1627 sample rate of the stream and adds a Microsoft wave header containing
1628 this information at the beginning. This allows to write wav files
1629 rather than raw PCM files (which do not contain any information about
1630 the number of channels and the sample rate).
1631
1632 The prebuffer filter simply delays the output until the given time has
1633 passed (starting from the time the first byte was available in its
1634 input queue) or until the given amount of data has accumulated. It
1635 is mainly useful for para_audiod if the standard parameters result
1636 in buffer underruns.
1637
1638 Both filters require almost no additional computing time, even when
1639 operating on uncompressed audio streams, since data buffers are simply
1640 "pushed down" rather than copied.
1641
1642 Examples
1643 --------
1644
1645 -> Decode an mp3 file to wav format:
1646
1647 para_filter -f mp3dec -f wav < file.mp3 > file.wav
1648
1649 -> Amplify a raw audio file by a factor of 1.5:
1650
1651 para_filter -f amp --amp 32 < foo.raw > bar.raw
1652
1653 ======
1654 Output
1655 ======
1656
1657 Once an audio stream has been received and decoded to PCM format,
1658 it can be sent to a sound device for playback. This part is performed
1659 by paraslash _writers_ which are described in this chapter.
1660
1661 Writers
1662 -------
1663
1664 A paraslash writer acts as a data sink that consumes but does not
1665 produce audio data. Paraslash writers operate on the client side and
1666 are contained in para_audiod and in the stand-alone tool para_write.
1667
1668 The para_write program reads uncompressed audio data from STDIN. If
1669 this data starts with a wav header, sample rate, sample format and
1670 channel count are read from the header. Otherwise CD audio (44.1KHz
1671 16 bit little endian, stereo) is assumed but this can be overridden
1672 by command line options. para_audiod, on the other hand, obtains
1673 the sample rate and the number of channels from the decoder.
1674
1675 Like receivers and filters, each writer has an individual set of
1676 command line options, and for para_audiod writers can be configured
1677 per audio format separately. It is possible to activate more than
1678 one writer for the same stream simultaneously.
1679
1680 OS-dependent APIs
1681 -----------------
1682
1683 Unfortunately, the various flavours of Unix on which paraslash
1684 runs on have different APIs for opening a sound device and starting
1685 playback. Hence for each such API there is a paraslash writer that
1686 can play the audio stream via this API.
1687
1688 - *ALSA*. The _Advanced Linux Sound Architecture_ is only available on
1689 Linux systems. Although there are several mid-layer APIs in use by
1690 the various Linux distributions (ESD, Jack, PulseAudio), paraslash
1691 currently supports only the low-level ALSA API which is not supposed
1692 to be change. ALSA is very feature-rich, in particular it supports
1693 software mixing via its DMIX plugin. ALSA is the default writer on
1694 Linux systems.
1695
1696 - *OSS*. The _Open Sound System_ is the only API on \*BSD Unixes and
1697 is also available on Linux systems, usually provided by ALSA as an
1698 emulation for backwards compatibility. This API is rather simple but
1699 also limited. For example only one application can open the device
1700 at any time. The OSS writer is activated by default on BSD Systems.
1701
1702 - *OSX*. Mac OS X has yet another API called CoreAudio. The OSX writer
1703 for this API is only compiled in on such systems and is of course
1704 the default there.
1705
1706 - *FILE*. The file writer allows to capture the audio stream and
1707 write the PCM data to a file on the file system rather than playing
1708 it through a sound device. It is supported on all platforms and is
1709 always compiled in.
1710
1711 - *AO*. _Libao_ is a cross-platform audio library which supports a wide
1712 variety of platforms including PulseAudio (gnome), ESD (Enlightened
1713 Sound Daemon), AIX, Solaris and IRIX. The ao writer plays audio
1714 through an output plugin of libao.
1715
1716 Examples
1717 --------
1718
1719 -> Use the OSS writer to play a wav file:
1720
1721 para_write --writer oss < file.wav
1722
1723 -> Enable ALSA software mixing for mp3 streams:
1724
1725 para_audiod --writer 'mp3:alsa -d plug:swmix'
1726
1727
1728 ===
1729 Gui
1730 ===
1731
1732 para_gui executes an arbitrary command which is supposed to print
1733 status information to STDOUT. It then displays this information in
1734 a curses window. By default the command
1735
1736 para_audioc -- stat -p
1737
1738 is executed, but this can be customized via the --stat-cmd option. In
1739 particular it possible to use
1740
1741 para_client -- stat -p
1742
1743 to make para_gui work on systems on which para_audiod is not running.
1744
1745 Key bindings
1746 ------------
1747
1748 It is possible to bind keys to arbitrary commands via custom
1749 key-bindings. Besides the internal keys which can not be changed (help,
1750 quit, loglevel, version...), the following flavours of key-bindings
1751 are supported:
1752
1753 - external: Shutdown curses before launching the given command.
1754 Useful for starting other ncurses programs from within para_gui,
1755 e.g. aumix or dialog scripts. Or, use the mbox output format to write
1756 a mailbox containing one mail for each (admissible) file the audio
1757 file selector knows about. Then start mutt from within para_gui to
1758 browse your collection!
1759
1760 - display: Launch the command and display its stdout in para_gui's
1761 bottom window.
1762
1763 - para: Like display, but start "para_client <specified command>"
1764 instead of "<specified command>".
1765
1766 The general form of a key binding is
1767
1768 key_map k:m:c
1769
1770 which maps key k to command c using mode m. Mode may be x, d or p
1771 for external, display and paraslash commands, respectively.
1772
1773 Themes
1774 ------
1775
1776 Currently there are only two themes for para_gui. It is easy, however,
1777 to add more themes. To create a new theme one has to define the
1778 position, color and geometry for for each status item that should be
1779 shown by this theme. See gui_theme.c for examples.
1780
1781 The "." and "," keys are used to switch between themes.
1782
1783 Examples
1784 --------
1785
1786 -> Show server info:
1787
1788 key_map "i:p:si"
1789
1790 -> Jump to the middle of the current audio file by pressing F5:
1791
1792 key_map "<F5>:p:jmp 50"
1793
1794 -> vi-like bindings for jumping around:
1795
1796 key_map "l:p:ff 10"
1797 key_map "h:p:ff 10-"
1798 key_map "w:p:ff 60"
1799 key_map "b:p:ff 60-"
1800
1801 -> Print the current date and time:
1802
1803 key_map "D:d:date"
1804
1805 -> Call other curses programs:
1806
1807 key_map "U:x:aumix"
1808 key_map "!:x:/bin/bash"
1809 key_map "^E:x:/bin/sh -c 'vi ~/.paraslash/gui.conf'"
1810
1811 ===========
1812 Development
1813 ===========
1814
1815 Tools
1816 -----
1817
1818 In order to compile the sources from the git repository (rather than
1819 from tar balls) and for contributing non-trivial changes to the
1820 paraslash project, some additional tools should be installed on a
1821 developer machine.
1822
1823 - [git](http://git.or.cz/). As described in more detail
1824 [below](#Git.branches), the git source code management tool is used for
1825 paraslash development. It is necessary for cloning the git repository
1826 and for getting updates.
1827
1828 - [m4](ftp://ftp.gnu.org/pub/gnu/m4/). Some input files for gengetopt
1829 are generated from templates by the m4 macro processor.
1830
1831 - [autoconf](ftp://ftp.gnu.org/pub/gnu/autoconf/) GNU autoconf creates
1832 the configure file which is shipped in the tarballs but has to be
1833 generated when compiling from git.
1834
1835 - [discount](http://www.pell.portland.or.us/~orc/Code/discount). The
1836 HTML version of this manual and some of the paraslash web pages are
1837 written in the Markdown markup language and are translated into html
1838 with the converter of the *Discount* package.
1839
1840 - [doxygen](http://www.stack.nl/~dimitri/doxygen/). The documentation
1841 of paraslash's C sources uses the doxygen documentation system. The
1842 conventions for documenting the source code is described in the
1843 [Doxygen section](#Doxygen).
1844
1845 - [global](ftp://ftp.gnu.org/pub/gnu/global). This is used to generate
1846 browsable HTML from the C sources. It is needed by doxygen.
1847
1848 Git branches
1849 ------------
1850
1851 Paraslash has been developed using the git source code management
1852 tool since 2006. Development is organized roughly in the same spirit
1853 as the git development itself, as described below.
1854
1855 The following text passage is based on "A note from the maintainer",
1856 written by Junio C Hamano, the maintainer of git.
1857
1858 There are four branches in the paraslash repository that track the
1859 source tree: "master", "maint", "next", and "pu".
1860
1861 The "master" branch is meant to contain what is well tested and
1862 ready to be used in a production setting. There could occasionally be
1863 minor breakages or brown paper bag bugs but they are not expected to
1864 be anything major, and more importantly quickly and easily fixable.
1865 Every now and then, a "feature release" is cut from the tip of this
1866 branch, named with three dotted decimal digits, like 0.4.2.
1867
1868 Whenever changes are about to be included that will eventually lead to
1869 a new major release (e.g. 0.5.0), a "maint" branch is forked off from
1870 "master" at that point. Obvious, safe and urgent fixes after the major
1871 release are applied to this branch and maintenance releases are cut
1872 from it. New features never go to this branch. This branch is also
1873 merged into "master" to propagate the fixes forward.
1874
1875 A trivial and safe enhancement goes directly on top of "master".
1876 New development does not usually happen on "master", however.
1877 Instead, a separate topic branch is forked from the tip of "master",
1878 and it first is tested in isolation; Usually there are a handful such
1879 topic branches that are running ahead of "master". The tip of these
1880 branches is not published in the public repository to keep the number
1881 of branches that downstream developers need to worry about low.
1882
1883 The quality of topic branches varies widely. Some of them start out as
1884 "good idea but obviously is broken in some areas" and then with some
1885 more work become "more or less done and can now be tested by wider
1886 audience". Luckily, most of them start out in the latter, better shape.
1887
1888 The "next" branch is to merge and test topic branches in the latter
1889 category. In general, this branch always contains the tip of "master".
1890 It might not be quite rock-solid production ready, but is expected to
1891 work more or less without major breakage. The maintainer usually uses
1892 the "next" version of paraslash for his own pleasure, so it cannot
1893 be _that_ broken. The "next" branch is where new and exciting things
1894 take place.
1895
1896 The two branches "master" and "maint" are never rewound, and "next"
1897 usually will not be either (this automatically means the topics that
1898 have been merged into "next" are usually not rebased, and you can find
1899 the tip of topic branches you are interested in from the output of
1900 "git log next"). You should be able to safely build on top of them.
1901
1902 However, at times "next" will be rebuilt from the tip of "master" to
1903 get rid of merge commits that will never be in "master". The commit
1904 that replaces "next" will usually have the identical tree, but it
1905 will have different ancestry from the tip of "master".
1906
1907 The "pu" (proposed updates) branch bundles the remainder of the
1908 topic branches. The "pu" branch, and topic branches that are only in
1909 "pu", are subject to rebasing in general. By the above definition
1910 of how "next" works, you can tell that this branch will contain quite
1911 experimental and obviously broken stuff.
1912
1913 When a topic that was in "pu" proves to be in testable shape, it
1914 graduates to "next". This is done with
1915
1916 git checkout next
1917 git merge that-topic-branch
1918
1919 Sometimes, an idea that looked promising turns out to be not so good
1920 and the topic can be dropped from "pu" in such a case.
1921
1922 A topic that is in "next" is expected to be polished to perfection
1923 before it is merged to "master". Similar to the above, this is
1924 done with
1925
1926 git checkout master
1927 git merge that-topic-branch
1928 git branch -d that-topic-branch
1929
1930 Note that being in "next" is not a guarantee to appear in the next
1931 release (being in "master" is such a guarantee, unless it is later
1932 found seriously broken and reverted), nor even in any future release.
1933
1934 Coding Style
1935 ------------
1936
1937 The preferred coding style for paraslash coincides more or less
1938 with the style of the Linux kernel. So rather than repeating what is
1939 written [there](http://www.kernel.org/doc/Documentation/CodingStyle),
1940 here are the most important points.
1941
1942 - Burn the GNU coding standards.
1943 - Never use spaces for indentation.
1944 - Tabs are 8 characters, and thus indentations are also 8 characters.
1945 - Don't put multiple assignments on a single line.
1946 - Avoid tricky expressions.
1947 - Don't leave whitespace at the end of lines.
1948 - The limit on the length of lines is 80 columns.
1949 - Use K&R style for placing braces and spaces:
1950
1951 if (x is true) {
1952 we do y
1953 }
1954
1955 - Use a space after (most) keywords.
1956 - Do not add spaces around (inside) parenthesized expressions.
1957 - Use one space around (on each side of) most binary and ternary operators.
1958 - Do not use cute names like ThisVariableIsATemporaryCounter, call it tmp.
1959 - Mixed-case names are frowned upon.
1960 - Descriptive names for global variables are a must.
1961 - Avoid typedefs.
1962 - Functions should be short and sweet, and do just one thing.
1963 - The number of local variables shouldn't exceed 10.
1964 - Gotos are fine if they improve readability and reduce nesting.
1965 - Don't use C99-style "// ..." comments.
1966 - Names of macros defining constants and labels in enums are capitalized.
1967 - Enums are preferred when defining several related constants.
1968 - Always use the paraslash wrappers for allocating memory.
1969 - If the name of a function is an action or an imperative.
1970 command, the function should return an error-code integer
1971 (<0 means error, >=0 means success). If the name is a
1972 predicate, the function should return a "succeeded" boolean.
1973
1974 Doxygen
1975 -------
1976
1977 Doxygen is a documentation system for various programming
1978 languages. The API reference on the paraslash web page is generated
1979 by doxygen.
1980
1981 It is more illustrative to look at the source code for examples than
1982 to describe the conventions in this manual, so we only describe which
1983 parts of the code need doxygen comments, but leave out details on
1984 documentation conventions.
1985
1986 As a rule, only the public part of the C source is documented with
1987 Doxygen. This includes structures, defines and enumerations in header
1988 files as well as public (non-static) C functions. These should be
1989 documented completely. For example, each parameter and the return
1990 value of a public function should get a descriptive doxygen comment.
1991
1992 No doxygen comments are necessary for static functions and for
1993 structures and enumerations in C files (which are used only within
1994 this file). This does not mean, however, that those entities need
1995 no documentation at all. Instead, common sense should be applied to
1996 document what is not obvious from reading the code.
1997
1998 ========
1999 Appendix
2000 ========
2001
2002 Network protocols
2003 -----------------
2004
2005 ### IP ###
2006
2007 The _Internet Protocol_ is the primary networking protocol used for
2008 the Internet. All protocols described below use IP as the underlying
2009 layer. Both the prevalent IPv4 and the next-generation IPv6 variant
2010 are being deployed actively worldwide.
2011
2012 ### Connection-oriented and connectionless protocols ###
2013
2014 Connectionless protocols differ from connection-oriented ones in
2015 that state associated with the sending/receiving endpoints is treated
2016 implicitly. Connectionless protocols maintain no internal knowledge
2017 about the state of the connection. Hence they are not capable of
2018 reacting to state changes, such as sudden loss or congestion on the
2019 connection medium. Connection-oriented protocols, in contrast, make
2020 this knowledge explicit. The connection is established only after
2021 a bidirectional handshake which requires both endpoints to agree
2022 on the state of the connection, and may also involve negotiating
2023 specific parameters for the particular connection. Maintaining an
2024 up-to-date internal state of the connection also in general means
2025 that the sending endpoints perform congestion control, adapting to
2026 qualitative changes of the connection medium.
2027
2028 ### Reliability ###
2029
2030 In IP networking, packets can be lost, duplicated, or delivered
2031 out of order, and different network protocols handle these
2032 problems in different ways. We call a transport-layer protocol
2033 _reliable_, if it turns the unreliable IP delivery into an ordered,
2034 duplicate- and loss-free delivery of packets. Sequence numbers
2035 are used to discard duplicates and re-arrange packets delivered
2036 out-of-order. Retransmission is used to guarantee loss-free
2037 delivery. Unreliable protocols, in contrast, do not guarantee ordering
2038 or data integrity.
2039
2040 ### Classification ###
2041
2042 With these definitions the protocols which are used by paraslash for
2043 steaming audio data may be classified as follows.
2044
2045 - HTTP/TCP: connection-oriented, reliable,
2046 - UDP: connectionless, unreliable,
2047 - DCCP: connection-oriented, unreliable.
2048
2049 Below we give a short descriptions of these protocols.
2050
2051 ### TCP ###
2052
2053 The _Transmission Control Protocol_ provides reliable, ordered delivery
2054 of a stream and a classic window-based congestion control. In contrast
2055 to UDP and DCCP (see below), TCP does not have record-oriented or
2056 datagram-based syntax, i.e. it provides a stream which is unaware
2057 and independent of any record (packet) boundaries. TCP is used
2058 extensively by many application layers. Besides HTTP (the Hypertext
2059 Transfer Protocol), also FTP (the File Transfer protocol), SMTP (Simple
2060 Mail Transfer Protocol), SSH (Secure Shell) all sit on top of TCP.
2061
2062 ### UDP ###
2063
2064 The _User Datagram Protocol_ is the simplest transport-layer protocol,
2065 built as a thin layer directly on top of IP. For this reason, it offers
2066 the same best-effort service as IP itself, i.e. there is no detection
2067 of duplicate or reordered packets. Being a connectionless protocol,
2068 only minimal internal state about the connection is maintained, which
2069 means that there is no protection against packet loss or network
2070 congestion. Error checking and correction (if at all) are performed
2071 in the application.
2072
2073 ### DCCP ###
2074
2075 The _Datagram Congestion Control Protocol_ combines the
2076 connection-oriented state maintenance known from TCP with the
2077 unreliable, datagram-based transport of UDP. This means that it
2078 is capable of reacting to changes in the connection by performing
2079 congestion control, offering multiple alternative approaches. But it
2080 is bound to datagram boundaries (the maximum packet size supported
2081 by a medium), and like UDP it lacks retransmission to protect
2082 against loss. Due to the use of sequence numbers, it is however
2083 able to react to loss (interpreted as a congestion indication) and
2084 to ignore out-of-order and duplicate packets. Unlike TCP it allows
2085 to negotiate specific, binding features for a connection, such as
2086 the choice of congestion control: classic, window-based congestion
2087 control known from TCP is available as CCID-2, rate-based, "smooth"
2088 congestion control is offered as CCID-3.
2089
2090 ### HTTP ###
2091
2092 The _Hypertext Transfer Protocol_ is an application layer protocol
2093 on top of TCP. It is spoken by web servers and is most often used
2094 for web services. However, as can be seen by the many Internet radio
2095 stations and YouTube/Flash videos, http is by far not limited to the
2096 delivery of web pages only. Being a simple request/response based
2097 protocol, the semantics of the protocol also allow the delivery of
2098 multimedia content, such as audio over http.
2099
2100 ### Multicast ###
2101
2102 IP multicast is not really a protocol but a technique for one-to-many
2103 communication over an IP network. The challenge is to deliver
2104 information to a group of destinations simultaneously using the
2105 most efficient strategy to send the messages over each link of the
2106 network only once. This has benefits for streaming multimedia: the
2107 standard one-to-one unicast offered by TCP/DCCP means that n clients
2108 listening to the same stream also consume n-times the resources,
2109 whereas multicast requires to send the stream just once, irrespective
2110 of the number of receivers. Since it would be costly to maintain state
2111 for each listening receiver, multicast often implies connectionless
2112 transport, which is the reason that it is currently only available
2113 via UDP.
2114
2115 Abstract socket namespace
2116 -------------------------
2117 UNIX domain sockets are a traditional way to communicate between
2118 processes on the same machine. They are always reliable (see above)
2119 and don't reorder datagrams. Unlike TCP and UDP, UNIX domain sockets
2120 support passing open file descriptors or process credentials to
2121 other processes.
2122
2123 The usual way to set up a UNIX domain socket (as obtained from
2124 socket(2)) for listening is to first bind the socket to a file system
2125 pathname and then call listen(2), then accept(2). Such sockets are
2126 called _pathname sockets_ because bind(2) creates a special socket
2127 file at the specified path. Pathname sockets allow unrelated processes
2128 to communicate with the listening process by binding to the same path
2129 and calling connect(2).
2130
2131 There are two problems with pathname sockets:
2132
2133 * The listing process must be able to (safely) create the
2134 socket special in a directory which is also accessible to
2135 the connecting process.
2136
2137 * After an unclean shutdown of the listening process, a stale
2138 socket special may reside on the file system.
2139
2140 The abstract socket namespace is a non-portable Linux feature which
2141 avoids these problems. Abstract sockets are still bound to a name,
2142 but the name has no connection with file system pathnames.
2143
2144 License
2145 -------
2146
2147 Paraslash is licensed under the GPL, version 2. Most of the code
2148 base has been written from scratch, and those parts are GPL V2
2149 throughout. Notable exceptions are FEC and the WMA decoder. See the
2150 corresponding source files for licencing details for these parts. Some
2151 code sniplets of several other third party software packages have
2152 been incorporated into the paraslash sources, for example log message
2153 coloring was taken from the git sources. These third party software
2154 packages are all published under the GPL or some other license
2155 compatible to the GPL.
2156
2157 Acknowledgements
2158 ----------------
2159
2160 Many thanks to Gerrit Renker who read an early draft of this manual
2161 and contributed significant improvements.
2162
2163 ==========
2164 References
2165 ==========
2166
2167 Articles
2168 --------
2169 - [Polynomial Codes over Certain Finite
2170 Fields](http://kom.aau.dk/~heb/kurser/NOTER/KOFA01.PDF) by Reed, Irving
2171 S.; Solomon, Gustave (1960), Journal of the Society for Industrial
2172 and Applied Mathematics (SIAM) 8 (2): 300-304, doi:10.1137/0108018)
2173
2174 RFCs
2175 ----
2176
2177 - [RFC 768](http://www.ietf.org/rfc/rfc768.txt) (1980): User Datagram
2178 Protocol
2179
2180 - [RFC 791](http://www.ietf.org/rfc/rfc791.txt) (1981): Internet
2181 Protocol
2182
2183 - [RFC 2437](http://www.ietf.org/rfc/rfc2437.txt) (1998): RSA
2184 Cryptography Specifications
2185
2186 - [RFC 4340](http://www.ietf.org/rfc/rfc4340.txt) (2006): Datagram
2187 Congestion Control Protocol (DCCP)
2188
2189 - [RFC 4341](http://www.ietf.org/rfc/rfc4341.txt) (2006): Congestion
2190 Control ID 2: TCP-like Congestion Control
2191
2192 - [RFC 4342](http://www.ietf.org/rfc/rfc4342.txt) (2006): Congestion
2193 Control ID 3: TCP-Friendly Rate Control (TFRC)
2194
2195 - [RFC 6716](http://www.ietf.org/rfc/rfc6716.txt) (2012): Definition
2196 of the Opus Audio Codec
2197
2198 Application web pages
2199 ---------------------
2200
2201 - [paraslash](http://people.tuebingen.mpg.de/maan/paraslash/)
2202 - [xmms](http://xmms2.org/wiki/Main_Page)
2203 - [mpg123](http://www.mpg123.de/)
2204 - [gstreamer](http://gstreamer.freedesktop.org/)
2205 - [icecast](http://www.icecast.org/)
2206 - [Audio Compress](http://beesbuzz.biz/code/audiocompress.php)
2207
2208 External documentation
2209 ----------------------
2210
2211 - [The mathematics of
2212 Raid6](http://kernel.org/pub/linux/kernel/people/hpa/raid6.pdf)
2213 by H. Peter Anvin
2214
2215 - [Effective Erasure Codes for reliable Computer Communication
2216 Protocols](http://info.iet.unipi.it/~luigi/fec_ccr.ps.gz) by Luigi
2217 Rizzo
2218
2219 Code
2220 ----
2221 - [Original FEC
2222 implementation](http://info.iet.unipi.it/~luigi/vdm.tar.gz) by
2223 Luigi Rizzo)