Make web pages conform to the HTML standard again.
[paraslash.git] / web / manual.m4
1 dnl To generate the html version, execute
2 dnl m4 web/manual.m4 | grutatxt --toc
3
4 define(`LOCAL_LINK_NAME', `translit(`$1', `A-Z
5 ', `a-z__')')
6 define(`REMOVE_NEWLINE', `translit(`$1',`
7 ', ` ')')
8
9 define(`REFERENCE', ./``#''`LOCAL_LINK_NAME($1)' (`REMOVE_NEWLINE($2)'))
10 define(`XREFERENCE', `$1' (`REMOVE_NEWLINE($2)'))
11 define(`EMPH', ``_''`REMOVE_NEWLINE($1)'``_'')
12
13 Paraslash user manual
14 =====================
15
16 This document describes how to install, configure and use the paraslash
17 network audio streaming system. Most chapters start with a chapter
18 overview and conclude with an example section. We try to focus on
19 general concepts and on the interaction of the various pieces of the
20 paraslash package. Hence this user manual is not meant as a replacement
21 for the manual pages that describe all command line options of each
22 paraslash executable.
23
24 ------------
25 Introduction
26 ------------
27
28 In this chapter we give an REFERENCE(Overview, overview) of the
29 interactions of the two main programs contained in the paraslash
30 package, followed by REFERENCE(The paraslash executables, brief
31 descriptions) of all executables.
32
33 Overview
34 ~~~~~~~~
35
36 The core functionality of the para suite is provided by two main
37 executables, para_server and para_audiod. The former maintains a
38 database of audio files and streams these files to para_audiod which
39 receives and plays the stream.
40
41 In a typical setting, both para_server and para_audiod act as
42 background daemons whose functionality is controlled by client
43 programs: the para_audioc client controls para_audiod over a local
44 socket while the para_client program connects to para_server over a
45 local or remote networking connection.
46
47 Typically, these two daemons run on different hosts but a local setup
48 is also possible.
49
50 A simplified picture of a typical setup is as follows
51 <<
52 <pre>
53 server_host client_host
54 ~~~~~~~~~~~ ~~~~~~~~~~~
55
56 +-----------+ audio stream +-----------+
57 |para_server| -----------------------------> |para_audiod|
58 +-----------+ +-----------+
59 ^ ^
60 | |
61 | | connect
62 | |
63 | |
64 | +-----------+
65 | |para_audioc|
66 | +-----------+
67 |
68 |
69 | connect +-----------+
70 +-------------------------------------- |para_client|
71 +-----------+
72 </pre>
73 >>
74
75 The paraslash executables
76 ~~~~~~~~~~~~~~~~~~~~~~~~~
77
78 *para_server*
79
80 para_server streams binary audio data (MP3, OGG/Vorbis, OGG/Speex,
81 M4A, WMA files) over local and/or remote networks. It listens on a
82 TCP port and accepts commands such as play, stop, pause, next from
83 authenticated clients. There are many more commands though, see the
84 man page of para_server for a description of all commands.
85
86 It supports three built-in network streaming protocols
87 (senders/receivers): HTTP, DCCP, or UDP. This is explained in more
88 detail in the section on REFERENCE(Networking, networking).
89
90 The built-in audio file selector of paraslash is used to manage your
91 audio files. It maintains statistics on the usage of all available
92 audio files such as last-played time, and the number of times each
93 file was selected.
94
95 Additional information may be added to the database to allow
96 fine-grained selection based on various properties of the audio file,
97 including information found in (ID3) tags. However, old-fashioned
98 playlists are also supported.
99
100 It is also possible to store images (album covers) and lyrics in the
101 database and associate these to the corresponding audio files.
102
103 The section on the REFERENCE(The audio file selector, audio file
104 selector) discusses this topic.
105
106
107 *para_client*
108
109 The client program to connect to para_server. paraslash commands
110 are sent to para_server and the response is dumped to STDOUT. This
111 can be used by any scripting language to produce user interfaces with
112 little programming effort.
113
114 All connections between para_server and para_client are encrypted
115 with a symmetric RC4 session key. For each user of paraslash you must
116 create a public/secret RSA key pair for authentication.
117
118 If para_client is started without non-option arguments, an interactive
119 session (shell) is started. Command history and command completion are
120 supported through libreadline.
121
122 *para_audiod*
123
124 The local daemon that collects information from para_server.
125
126 It runs on the client side and connects to para_server. As soon as
127 para_server announces the availability of an audio stream, para_audiod
128 starts an appropriate receiver, any number of filters and a paraslash
129 writer to play the stream.
130
131 Moreover, para_audiod listens on a local socket and sends status
132 information about para_server and para_audiod to local clients on
133 request. Access via this local socket may be restricted by using Unix
134 socket credentials, if available.
135
136
137 *para_audioc*
138
139 The client program which talks to para_audiod. Used to control
140 para_audiod, to receive status info, or to grab the stream at any
141 point of the decoding process. Like para_client, para_audioc supports
142 interactive sessions on systems with libreadline.
143
144 *para_recv*
145
146 A command line HTTP/DCCP/UDP stream grabber. The http mode is
147 compatible with arbitrary HTTP streaming sources (e.g. icecast).
148
149 *para_filter*
150
151 A filter program that reads from STDIN and writes to STDOUT.
152 Like para_recv, this is an atomic building block which can be used to
153 assemble higher-level audio receiving facilities. It combines several
154 different functionalities in one tool: decoders for multiple audio
155 formats and a number of processing filters, among these a normalizer
156 for audio volume.
157
158 *para_afh*
159
160 A small stand-alone program that prints tech info about the given
161 audio file to STDOUT. It can be instructed to print a "chunk table",
162 an array of offsets within the audio file or to write the content of
163 the audio file in complete chunks 'just in time'.
164
165 This allows third-party streaming software that is unaware of the
166 particular audio format to send complete frames in real time.
167
168 *para_write*
169
170 A modular audio stream writer. It supports a simple file writer
171 output plug-in and optional WAV/raw players for ALSA (Linux) and for
172 coreaudio (Mac OS). para_write can also be used as a stand-alone WAV
173 or raw audio player.
174
175
176 *para_gui*
177
178 Curses-based gui that presents status information obtained in a curses
179 window. Appearance can be customized via themes. para_gui provides
180 key-bindings for the most common server commands and new key-bindings
181 can be added easily.
182
183
184 *para_fade*
185
186 An (OSS-only) alarm clock and volume-fader.
187
188 -----------
189 Quick start
190 -----------
191
192 This chapter lists the REFERENCE(Requirements, necessary software)
193 that must be installed to compile the paraslash package, describes
194 how to REFERENCE(Installation, compile and install) the paraslash
195 source code and the steps that have to be performed in order to
196 REFERENCE(Quick start, set up) a typical server and client.
197
198 Requirements
199 ~~~~~~~~~~~~
200
201 In any case you'll need
202
203 - XREFERENCE(http://systemlinux.org/~maan/osl/, libosl).
204 The _object storage layer_ library is used by para_server. To
205 clone the source code repository, execute
206
207 git clone git://git.tuebingen.mpg.de/osl
208
209 - XREFERENCE(ftp://ftp.gnu.org/pub/gnu/gcc, gcc) or
210 XREFERENCE(http://clang.llvm.org, clang). All gcc versions
211 >= 3.3 are currently supported. Clang version 1.1 or newer
212 should work as well.
213
214 - XREFERENCE(ftp://ftp.gnu.org/pub/gnu/make, gnu make) is
215 also shipped with the disto. On BSD systems the gnu make
216 executable is often called gmake.
217
218 - XREFERENCE(ftp://ftp.gnu.org/pub/gnu/bash, bash). Some
219 scripts which run during compilation require the EMPH(Bourne
220 again shell). It is most likely already installed.
221
222 - XREFERENCE(http://www.openssl.org/, openssl) or
223 XREFERENCE(ftp://ftp.gnupg.org/gcrypt/libgcrypt/, libgcrypt).
224 At least one of these two libraries is needed as the backend
225 for cryptographic routines on both the server and the client
226 side. Both openssl and libgcrypt are usually shipped with the
227 distro, but you might have to install the development package
228 (libssl-dev or libgcrypt-dev on debian systems) as well.
229
230 - XREFERENCE(ftp://ftp.gnu.org/pub/gnu/gengetopt/, gengetopt)
231 is needed to generate the C code for the command line parsers
232 of all paraslash executables.
233
234 - XREFERENCE(ftp://ftp.gnu.org/pub/gnu/help2man, help2man)
235 is used to create the man pages.
236
237 Optional:
238
239 - XREFERENCE(http://www.underbit.com/products/mad/, libmad).
240 To compile in MP3 support for paraslash, the development
241 package must be installed. It is called libmad0-dev on
242 debian-based systems. Note that libmad is not necessary on
243 the server side, i.e. for sending MP3 files.
244
245 - XREFERENCE(http://www.underbit.com/products/mad/,
246 libid3tag). For version-2 ID3 tag support, you'll need
247 the libid3tag development package libid3tag0-dev. Without
248 libid3tag, only version one tags are recognized.
249
250 - XREFERENCE(http://www.xiph.org/downloads/, ogg vorbis).
251 For ogg vorbis streams you'll need libogg, libvorbis,
252 libvorbisfile. The corresponding Debian packages are called
253 libogg-dev and libvorbis-dev.
254
255 - XREFERENCE(http://www.audiocoding.com/, libfaad). For aac
256 files (m4a) you'll need libfaad (libfaad-dev).
257
258 - XREFERENCE(http://www.speex.org/, speex). In order to stream
259 or decode speex files, libspeex (libspeex-dev) is required.
260
261 - XREFERENCE(http://flac.sourceforge.net/, flac). To stream
262 or decode files encoded with the _Free Lossless Audio Codec_,
263 libFLAC (libFLAC-dev) must be installed.
264
265 - XREFERENCE(ftp://ftp.alsa-project.org/pub/lib/, alsa-lib). On
266 Linux, you'll need to have ALSA's development package
267 libasound2-dev installed.
268
269 - XREFERENCE(http://downloads.xiph.org/releases/ao/,
270 libao). Needed to build the ao writer (ESD, PulseAudio,...).
271 Debian package: libao-dev.
272
273 - XREFERENCE(http://cnswww.cns.cwru.edu/php/chet/readline/rltop.html,
274 GNU Readline). If this library (libreadline-dev) is installed,
275 para_client and para_audioc support interactive sessions.
276
277 Installation
278 ~~~~~~~~~~~~
279
280 First make sure all non-optional packages listed in the section on
281 REFERENCE(Requirements, required software) are installed on your
282 system.
283
284 You don't need everything listed there. In particular, MP3, OGG/Vorbis,
285 OGG/Speex and AAC support are all optional. The configure script will
286 detect what is installed on your system and will only try to build
287 those executables that can be built with your setup.
288
289 Note that no special decoder library (not even the MP3 decoding library
290 libmad) is needed for para_server if you only want to stream MP3 or WMA
291 files. Also, it's fine to use para_server on a box without sound card.
292
293 Next, install the paraslash package on all machines, you'd like this
294 software to run on:
295
296 (./configure && make) > /dev/null
297
298 There should be no errors but probably some warnings about missing
299 packages which usually implies that not all audio formats will be
300 supported. If headers or libs are installed at unusual locations you
301 might need to tell the configure script where to find them. Try
302
303 ./configure --help
304
305 to see a list of options. If the paraslash package was compiled
306 successfully, execute (optionally)
307
308 make test
309
310 to run the paraslash test suite. If all tests pass, execute as root
311
312 make install
313
314 to install executables under /usr/local/bin and the man pages under
315 /usr/local/man.
316
317 Configuration
318 ~~~~~~~~~~~~~
319
320 *Step 1*: Create a paraslash user
321
322 In order to control para_server at runtime you must create a paraslash
323 user. As authentication is based on the RSA crypto system you'll have
324 to create an RSA key pair. If you already have a user and an RSA key
325 pair, you may skip this step.
326
327 In this section we'll assume a typical setup: You would like to run
328 para_server on some host called server_host as user foo, and you want
329 to connect to para_server from another machine called client_host as
330 user bar.
331
332 As foo@server_host, create ~/.paraslash/server.users by typing the
333 following commands:
334
335 user=bar
336 target=~/.paraslash/server.users
337 key=~/.paraslash/id_rsa.pub.$user
338 perms=AFS_READ,AFS_WRITE,VSS_READ,VSS_WRITE
339 mkdir -p ~/.paraslash
340 echo "user $user $key $perms" >> $target
341
342 Next, change to the "bar" account on client_host and generate the
343 key pair with the commands
344
345 ssh-keygen -t rsa -b 2048
346 # hit enter twice to create a key with no passphrase
347
348 This generates the two files id_rsa and id_rsa.pub in ~/.ssh. Note
349 that paraslash can also read keys generated by the "openssl genrsa"
350 command. However, since keys created with ssh-keygen can also be used
351 for ssh, this method is recommended.
352
353 Note that para_server refuses to use a key if it is shorter than 2048
354 bits. In particular, the RSA keys of paraslash 0.3.x will not work
355 with version 0.4.x. Moreover, para_client refuses to use a (private)
356 key which is world-readable.
357
358 para_server only needs to know the public key of the key pair just
359 created. Copy this public key to server_host:
360
361 src=~/.ssh/id_rsa.pub
362 dest=.paraslash/id_rsa.pub.$LOGNAME
363 scp $src foo@server_host:$dest
364
365 Finally, tell para_client to connect to server_host:
366
367 conf=~/.paraslash/client.conf
368 echo 'hostname server_host' > $conf
369
370
371 *Step 2*: Start para_server
372
373 Before starting the server make sure you have write permissions to
374 the directory /var/paraslash that has been created during installation:
375
376 sudo chown $LOGNAME /var/paraslash
377
378 Alternatively, use the --afs_socket Option to specify a different
379 location for the AFS command socket.
380
381 For this first try, we'll use the info loglevel to make the output
382 of para_server more verbose.
383
384 para_server -l info
385
386 Now you can use para_client to connect to the server and issue
387 commands. Open a new shell as bar@client_host and try
388
389 para_client help
390 para_client si
391
392 to retrieve the list of available commands and some server info.
393 Don't proceed if this doesn't work.
394
395 *Step 3*: Create and populate the database
396
397 An empty database is created with
398
399 para_client init
400
401 This initializes a couple of empty tables under
402 ~/.paraslash/afs_database-0.4. You normally don't need to look at these
403 tables, but it's good to know that you can start from scratch with
404
405 rm -rf ~/.paraslash/afs_database-0.4
406
407 in case something went wrong.
408
409 Next, you need to add some audio files to that database so that
410 para_server knows about them. Choose an absolute path to a directory
411 containing some audio files and add them to the audio file table:
412
413 para_client add /my/mp3/dir
414
415 This might take a while, so it is a good idea to start with a directory
416 containing not too many files. Note that the table only contains data
417 about the audio files found, not the files themselves.
418
419 You may print the list of all known audio files with
420
421 para_client ls
422
423 *Step 4*: Configure para_audiod
424
425 para_audiod needs to create a "well-known" socket for the clients to
426 connect to. The default path for this socket is
427
428 /var/paraslash/audiod_socket.$HOSTNAME
429
430 In order to make this directory writable for para_audiod, execute
431 as bar@client_host
432
433 sudo chown $LOGNAME /var/paraslash
434
435
436 We will also have to tell para_audiod that it should receive the
437 audio stream from server_host via http:
438
439 para_audiod -l info -r '.:http -i server_host'
440
441 You should now be able to listen to the audio stream once para_server
442 starts streaming. To activate streaming, execute
443
444 para_client play
445
446 Since no playlist has been specified yet, the "dummy" mode which
447 selects all known audio files is activated automatically. See the
448 section on the REFERENCE(The audio file selector, audio file selector)
449 for how to use playlists and moods to specify which files should be
450 streamed in which order.
451
452 *Troubleshooting*
453
454 It did not work? To find out why, try to receive, decode and play the
455 stream manually using para_recv, para_filter and para_write as follows.
456
457 For simplicity we assume that you're running Linux/ALSA and that only
458 MP3 files have been added to the database.
459
460 para_recv -r 'http -i server_host' > file.mp3
461 # (interrupt with CTRL+C after a few seconds)
462 ls -l file.mp3 # should not be empty
463 para_filter -f mp3dec -f wav < file.mp3 > file.wav
464 ls -l file.wav # should be much bigger than file.mp3
465 para_write -w alsa < file.wav
466
467 Double check what is logged by para_server and use the --loglevel
468 option of para_recv, para_filter and para_write to increase verbosity.
469
470 ---------------
471 User management
472 ---------------
473
474 para_server uses a challenge-response mechanism to authenticate
475 requests from incoming connections, similar to ssh's public key
476 authentication method. Authenticated connections are encrypted using
477 the RC4 stream cipher.
478
479 In this chapter we briefly describe RSA and RC4 and sketch the
480 REFERENCE(Client-server authentication, authentication handshake)
481 between para_client and para_server. User management is discussed
482 in the section on REFERENCE(The user_list file, the user_list file).
483 These sections are all about communication between the client and the
484 server. Connecting para_audiod is a different matter and is described
485 in a REFERENCE(Connecting para_audiod, separate section).
486
487
488
489 RSA and RC4
490 ~~~~~~~~~~~
491
492 RSA is an asymmetric block cipher which is used in many applications,
493 including ssh and gpg. An RSA key consists in fact of two keys,
494 called the public key and the private key. A message can be encrypted
495 with either key and only the counterpart of that key can decrypt
496 the message. While RSA can be used for both signing and encrypting
497 a message, paraslash only uses RSA only for the latter purpose. The
498 RSA public key encryption and signatures algorithms are defined in
499 detail in RFC 2437.
500
501 RC4 is a stream cipher, i.e. the input is XORed with a pseudo-random
502 key stream to produce the output. Decryption uses the same function
503 calls as encryption. While RC4 supports variable key lengths,
504 paraslash uses a fixed length of 256 bits, which is considered a
505 strong encryption by today's standards. Since the same key must never
506 be used twice, a different, randomly-generated key is used for every
507 new connection.
508
509 Client-server authentication
510 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
511
512 The authentication handshake between para_client and para_server goes
513 as follows:
514
515 - para_client connects to para_server and sends an
516 authentication request for a user. It does so by connecting
517 to TCP port 2990 of the server host. This port is called the
518 para_server _control port_.
519
520 - para_server accepts the connection and forks a child process
521 which handles the incoming request. The parent process keeps
522 listening on the control port while the child process (also
523 called para_server below) continues as follows.
524
525 - para_server loads the RSA public key of that user, fills a
526 fixed-length buffer with random bytes, encrypts that buffer
527 using the public key and sends the encrypted buffer to the
528 client. The first part of the buffer is the challenge which
529 is used for authentication while the second part is the RC4
530 session key.
531
532 - para_client receives the encrypted buffer and decrypts it
533 with the user's private key, thereby obtaining the challenge
534 buffer and the session key. It sends the SHA1 hash value of
535 the challenge back to para_server and stores the session key
536 for further use.
537
538 - para_server also computes the SHA1 hash of the challenge
539 and compares it against what was sent back by the client.
540
541 - If the two hashes do not match, the authentication has
542 failed and para_server closes the connection.
543
544 - Otherwise the user is considered authenticated and the client
545 is allowed to proceed by sending a command to be executed. From
546 this point on the communication is encrypted using the RC4
547 stream cipher with the session key known to both peers.
548
549 paraslash relies on the quality of the pseudo-random bytes provided
550 by the crypto library (openssl or libgcrypt), on the security of
551 the implementation of the RSA and RC4 crypto routines and on the
552 infeasibility to invert the SHA1 function.
553
554 Neither para_server or para_client create RSA keys on their own. This
555 has to be done once for each user as sketched in REFERENCE(Quick start,
556 Quick start) and discussed in more detail REFERENCE(The user_list
557 file, below).
558
559 The user_list file
560 ~~~~~~~~~~~~~~~~~~
561
562 At startup para_server reads the user list file which contains one
563 line per user. The default location of the user list file may be
564 changed with the --user_list option.
565
566 There should be at least one user in this file. Each user must have
567 an RSA key pair. The public part of the key is needed by para_server
568 while the private key is needed by para_client. Each line of the
569 user list file must be of the form
570
571 user <username> <key> <perms>
572
573 where _username_ is an arbitrary string (usually the user's login
574 name), _key_ is the full path to that user's public RSA key, and
575 _perms_ is a comma-separated list of zero or more of the following
576 permission bits:
577
578 +---------------------------------------------------------+
579 | AFS_READ | read the contents of the databases |
580 +-----------+---------------------------------------------+
581 | AFS_WRITE | change database contents |
582 +-----------+---------------------------------------------+
583 | VSS_READ | obtain information about the current stream |
584 +-----------+---------------------------------------------+
585 | VSS_WRITE | change the current stream |
586 +---------------------------------------------------------+
587
588 The permission bits specify which commands the user is allowed to
589 execute. The output of
590
591 para_client help
592
593 contains in the third column the permissions needed to execute the
594 command.
595
596 It is possible to make para_server reread the user_list file by
597 executing the paraslash "hup" command or by sending SIGHUP to the
598 PID of para_server.
599
600
601 Connecting para_audiod
602 ~~~~~~~~~~~~~~~~~~~~~~
603
604 para_audiod listens on a Unix domain socket. Those sockets are
605 for local communication only, so only local users can connect to
606 para_audiod. The default is to let any user connect but this can be
607 restricted on platforms that support UNIX socket credentials which
608 allow para_audiod to obtain the Unix credentials of the connecting
609 process.
610
611 Use para_audiod's --user_allow option to allow connections only for
612 a limited set of users.
613
614 -----------------------
615 The audio file selector
616 -----------------------
617
618 paraslash comes with a sophisticated audio file selector (AFS),
619 whose main task is to determine which file to stream next, based on
620 information on the audio files stored in a database. It communicates
621 also with para_client whenever an AFS command is executed, for example
622 to answer a database query.
623
624 Besides the traditional playlists, AFS supports audio file selection
625 based on _moods_ which act as a filter that limits the set of all
626 known audio files to those which satisfy certain criteria. It also
627 maintains tables containing images (e.g. album cover art) and lyrics
628 that can be associated with one or more audio files.
629
630 AFS uses XREFERENCE(http://systemlinux.org/~maan/osl/, libosl), the
631 object storage layer library, as the backend library for storing
632 information on audio files, playlists, etc. This library offers
633 functionality similar to a relational database, but is much more
634 lightweight than a full database backend.
635
636 In this chapter we sketch the setup of the REFERENCE(The AFS process,
637 AFS process) during server startup and proceed with the description
638 of the REFERENCE(Database layout, layout) of the various database
639 tables. The section on REFERENCE(Playlists and moods, playlists
640 and moods) explains these two audio file selection mechanisms
641 in detail and contains pratical examples. The way REFERENCE(File
642 renames and content changes, file renames and content changes) are
643 detected is discussed briefly before the REFERENCE(Troubleshooting,
644 Troubleshooting) section concludes the chapter.
645
646 The AFS process
647 ~~~~~~~~~~~~~~~
648
649 On startup, para_server forks to create the AFS process which opens
650 the OSL database tables. The server process communicates with the
651 AFS process via pipes and shared memory. Usually, the AFS process
652 awakes only briefly whenever the current audio file changes. The AFS
653 process determines the next audio file, opens it, verifies it has
654 not been changed since it was added to the database and passes the
655 open file descriptor to the server process, along with audio file
656 meta-data such as file name, duration, audio format and so on. The
657 server process then starts to stream the audio file.
658
659 The AFS process also accepts connections from local clients via
660 a well-known socket. However, only child processes of para_server
661 may connect through this socket. All server commands that have the
662 AFS_READ or AFS_WRITE permission bits use this mechanism to query or
663 change the database.
664
665 Database layout
666 ~~~~~~~~~~~~~~~
667
668 *The audio file table*
669
670 This is the most important and usually also the largest table of the
671 AFS database. It contains the information needed to stream each audio
672 file. In particular the following data is stored for each audio file.
673
674 - SHA1 hash value of the audio file contents. This is computed
675 once when the file is added to the database. Whenever AFS
676 selects this audio file for streaming the hash value is
677 recomputed and checked against the value stored in the
678 database to detect content changes.
679
680 - The time when this audio file was last played.
681
682 - The number of times the file has been played so far.
683
684 - The attribute bitmask.
685
686 - The image id which describes the image associated with this
687 audio file.
688
689 - The lyrics id which describes the lyrics associated with
690 this audio file.
691
692 - The audio format id (MP3, OGG, ...).
693
694 - An amplification value that can be used by the amplification
695 filter to pre-amplify the decoded audio stream.
696
697 - The chunk table. It describes the location and the timing
698 of the building blocks of the audio file. This is used by
699 para_server to send chunks of the file at appropriate times.
700
701 - The duration of the audio file.
702
703 - Tag information contained in the audio file (ID3 tags,
704 Vorbis comments, ...).
705
706 - The number of channels
707
708 - The encoding bitrate.
709
710 - The sampling frequency.
711
712 To add or refresh the data contained in the audio file table, the _add_
713 command is used. It takes the full path of either an audio file or a
714 directory. In the latter case, the directory is traversed recursively
715 and all files which are recognized as valid audio files are added to
716 the database.
717
718 *The attribute table*
719
720 The attribute table contains two columns, _name_ and _bitnum_. An
721 attribute is simply a name for a certain bit number in the attribute
722 bitmask of the audio file table.
723
724 Each of the 64 bits of the attribute bitmask can be set for each
725 audio file individually. Hence up to 64 different attributes may be
726 defined. For example, "pop", "rock", "blues", "jazz", "instrumental",
727 "german_lyrics", "speech", whatever. You are free to choose as
728 many attributes as you like and there are no naming restrictions
729 for attributes.
730
731 A new attribute "test" is created by
732
733 para_client addatt test
734 and
735 para_client lsatt
736
737 lists all available attributes. You can set the "test" attribute for
738 an audio file by executing
739
740 para_client setatt test+ /path/to/the/audio/file
741
742 Similarly, the "test" bit can be removed from an audio file with
743
744 para_client setatt test- /path/to/the/audio/file
745
746 Instead of a path you may use a shell wildcard pattern. The attribute
747 is applied to all audio files matching this pattern:
748
749 para_client setatt test+ '/test/directory/*'
750
751 The command
752
753 para_client -- ls -lv
754
755 gives you a verbose listing of your audio files also showing which
756 attributes are set.
757
758 In case you wonder why the double-dash in the above command is needed:
759 It tells para_client to not interpret the options after the dashes. If
760 you find this annoying, just say
761
762 alias para='para_client --'
763
764 and be happy. In what follows we shall use this alias.
765
766 The "test" attribute can be dropped from the database with
767
768 para rmatt test
769
770 Read the output of
771
772 para help ls
773 para help setatt
774
775 for more information and a complete list of command line options to
776 these commands.
777
778 *Blob tables*
779
780 The image, lyrics, moods and playlists tables are all blob tables.
781 Blob tables consist of three columns each: The identifier which is
782 a positive non-negative number that is auto-incremented, the name
783 (an arbitrary string) and the content (the blob).
784
785 All blob tables support the same set of actions: cat, ls, mv, rm
786 and add. Of course, _add_ is used for adding new blobs to the table
787 while the other actions have the same meaning as the corresponding
788 Unix commands. The paraslash commands to perform these actions are
789 constructed as the concatenation of the table name and the action. For
790 example addimg, catimg, lsimg, mvimg, rmimg are the commands that
791 manipulate or query the image table.
792
793 The add variant of these commands is special as these commands read
794 the blob contents from stdin. To add an image to the image table the
795 command
796
797 para addimg image_name < file.jpg
798
799 can be used.
800
801 Note that the images and lyrics are not interpreted at all, and also
802 the playlist and the mood blobs are only investigated when the mood
803 or playlist is activated with the select command.
804
805 *The score table*
806
807 Unlike all other tables the contents of the score table remain in
808 memory and are never stored on disk. The score table contains two
809 columns: The SHA1 hash value (of an audio file) and its current
810 score.
811
812 However, only those files which are admissible for the current mood
813 or playlist are contained in the score table. The audio file selector
814 always chooses the row with the highest score as the file to stream
815 next. While doing so, it computes the new score and updates the
816 last_played and the num_played fields in the audio file table.
817
818 The score table is recomputed by the select command which loads a
819 mood or playlist. Audio files are chosen for streaming from the rows
820 of the score table on a highest-score-first basis.
821
822
823 Playlists and moods
824 ~~~~~~~~~~~~~~~~~~~
825
826 Playlists and moods offer two different ways of specifying the set of
827 admissible files. A playlist in itself describes a set of admissible
828 files. A mood, in contrast, describes the set of admissible files in
829 terms of attributes and other type of information available in the
830 audio file table. As an example, a mood can define a filename pattern,
831 which is then matched against the names of audio files in the table.
832
833 *Playlists*
834
835 Playlists are accommodated in the playlist table of the afs database,
836 using the aforementioned blob format for tables. A new playlist is
837 created with the addpl command by specifying the full (absolute)
838 paths of all desired audio files, separated by newlines. Example:
839
840 find /my/mp3/dir -name "*.mp3" | para addpl my_playlist
841
842 If _my_playlist_ already exists it is overwritten. To activate the
843 new playlist, execute
844
845 para select p/my_playlist
846
847 The audio file selector will assign scores to each entry of the list,
848 in descending order so that files will be selected in order. If a
849 file could not be opened for streaming, its entry is removed from
850 the score table (but not from the playlist).
851
852 *Moods*
853
854 A mood consists of a unique name and its *mood definition*, which is
855 a set of *mood lines* containing expressions in terms of attributes
856 and other data contained in the database.
857
858 At any time at most one mood can be *active* which means that
859 para_server is going to select only files from that subset of
860 admissible files.
861
862 So in order to create a mood definition one has to write a set of
863 mood lines. Mood lines come in three flavours: Accept lines, deny
864 lines and score lines.
865
866 The general syntax of the three types of mood lines is
867
868
869 accept [with score <score>] [if] [not] <mood_method> [options]
870 deny [with score <score>] [if] [not] <mood_method> [options]
871 score <score> [if] [not] <mood_method> [options]
872
873
874 Here <score> is either an integer or the string "random" which assigns
875 a random score to all matching files. The score value changes the
876 order in which admissible files are going to be selected, but is of
877 minor importance for this introduction.
878
879 So we concentrate on the first two forms, i.e. accept and deny
880 lines. As usual, everything in square brackets is optional, i.e.
881 accept/deny lines take the following form when ignoring scores:
882
883 accept [if] [not] <mood_method> [options]
884
885 and analogously for the deny case. The "if" keyword is only syntactic
886 sugar and has no function. The "not" keyword just inverts the result,
887 so the essence of a mood line is the mood method part and the options
888 following thereafter.
889
890 A *mood method* is realized as a function which takes an audio file
891 and computes a number from the data contained in the database.
892 If this number is non-negative, we say the file *matches* the mood
893 method. The file matches the full mood line if it either
894
895 - matches the mood method and the "not" keyword is not given,
896 or
897 - does not match the mood method, but the "not" keyword is given.
898
899 The set of admissible files for the whole mood is now defined as those
900 files which match at least one accept mood line, but no deny mood line.
901 More formally, an audio file F is admissible if and only if
902
903 (F ~ AL1 or F ~ AL2...) and not (F ~ DL1 or F ~ DN2 ...)
904
905 where AL1, AL2... are the accept lines, DL1, DL2... are the deny
906 lines and "~" means "matches".
907
908 The cases where no mood lines of accept/deny type are defined need
909 special treatment:
910
911 - Neither accept nor deny lines: This treats all files as
912 admissible (in fact, that is the definition of the dummy mood
913 which is activated automatically if no moods are available).
914
915 - Only accept lines: A file is admissible iff it matches at
916 least one accept line:
917
918 F ~ AL1 or F ~ AL2 or ...
919
920 - Only deny lines: A file is admissible iff it matches no
921 deny line:
922
923 not (F ~ DL1 or F ~ DN2 ...)
924
925
926
927 *List of mood_methods*
928
929 no_attributes_set
930
931 Takes no arguments and matches an audio file if and only if no
932 attributes are set.
933
934 is_set <attribute_name>
935
936 Takes the name of an attribute and matches iff that attribute is set.
937
938 path_matches <pattern>
939
940 Takes a filename pattern and matches iff the path of the audio file
941 matches the pattern.
942
943 artist_matches <pattern>
944 album_matches <pattern>
945 title_matches <pattern>
946 comment_matches <pattern>
947
948 Takes an extended regular expression and matches iff the text of the
949 corresponding tag of the audio file matches the pattern. If the tag
950 is not set, the empty string is matched against the pattern.
951
952 year ~ <num>
953 bitrate ~ <num>
954 frequency ~ <num>
955 channels ~ <num>
956 num_played ~ <num>
957
958 Takes a comparator ~ of the set {<, =, <=, >, >=, !=} and a number
959 <num>. Matches an audio file iff the condition <val> ~ <num> is
960 satisfied where val is the corresponding value of the audio file
961 (value of the year tag, bitrate in kbit/s, frequency in Hz, channel
962 count, play count).
963
964 The year tag is special as its value is undefined if the audio file
965 has no year tag or the content of the year tag is not a number. Such
966 audio files never match. Another difference is the special treatment
967 if the year tag is a two-digit number. In this case either 1900 or
968 2000 is added to the tag value, depending on whether the number is
969 greater than 2000 plus the current year.
970
971
972 *Mood usage*
973
974 To create a new mood called "my_mood", write its definition into
975 some temporary file, say "tmpfile", and add it to the mood table
976 by executing
977
978 para addmood my_mood < tmpfile
979
980 If the mood definition is really short, you may just pipe it to the
981 client instead of using temporary files. Like this:
982
983 echo "$MOOD_DEFINITION" | para addmood my_mood
984
985 There is no need to keep the temporary file since you can always use
986 the catmood command to get it back:
987
988 para catmood my_mood
989
990 A mood can be activated by executing
991
992 para select m/my_mood
993
994 Once active, the list of admissible files is shown by the ls command
995 if the "-a" switch is given:
996
997 para ls -a
998
999
1000 *Example mood definition*
1001
1002 Suppose you have defined attributes "punk" and "rock" and want to define
1003 a mood containing only Punk-Rock songs. That is, an audio file should be
1004 admissible if and only if both attributes are set. Since
1005
1006 punk and rock
1007
1008 is obviously the same as
1009
1010 not (not punk or not rock)
1011
1012 (de Morgan's rule), a mood definition that selects only Punk-Rock
1013 songs is
1014
1015 deny if not is_set punk
1016 deny if not is_set rock
1017
1018
1019
1020 File renames and content changes
1021 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1022
1023 Since the audio file selector knows the SHA1 of each audio file that
1024 has been added to the afs database, it recognizes if the content of
1025 a file has changed, e.g. because an ID3 tag was added or modified.
1026 Also, if a file has been renamed or moved to a different location,
1027 afs will detect that an entry with the same hash value already exists
1028 in the audio file table.
1029
1030 In both cases it is enough to just re-add the new file. In the
1031 first case (file content changed), the audio table is updated, while
1032 metadata such as the num_played and last_played fields, as well as
1033 the attributes, remain unchanged. In the other case, when the file
1034 is moved or renamed, only the path information is updated, all other
1035 data remains as before.
1036
1037 It is possible to change the behaviour of the add command by using the
1038 "-l" (lazy add) or the "-f" (force add) option.
1039
1040 Troubleshooting
1041 ~~~~~~~~~~~~~~~
1042
1043 Use the debug loglevel (option -l debug for most commands) to show
1044 debugging info. Almost all paraslash executables have a brief online
1045 help which is displayed by using the -h switch. The --detailed-help
1046 option prints the full help text.
1047
1048 If para_server crashed or was killed by SIGKILL (signal 9), it
1049 may refuse to start again because of "dirty osl tables". In this
1050 case you'll have to run the oslfsck program of libosl to fix your
1051 database. It might be necessary to use --force (even if your name
1052 isn't Luke). However, make sure para_server isn't running before
1053 executing oslfsck --force.
1054
1055 If you don't mind to recreate your database you can start
1056 from scratch by removing the entire database directory, i.e.
1057
1058 rm -rf ~/.paraslash/afs_database-0.4
1059
1060 Be aware that this removes all attribute definitions, all playlists
1061 and all mood definitions and requires to re-initialize the tables.
1062
1063 Although oslfsck fixes inconsistencies in database tables it doesn't
1064 care about the table contents. To check for invalid table contents, use
1065
1066 para_client check
1067
1068 This prints out references to missing audio files as well as invalid
1069 playlists and mood definitions.
1070
1071 ---------------------------------------
1072 Audio formats and audio format handlers
1073 ---------------------------------------
1074
1075 Audio formats
1076 ~~~~~~~~~~~~~
1077
1078 The following audio formats are supported by paraslash:
1079
1080 *MP3*
1081
1082 Mp3, MPEG-1 Audio Layer 3, is a common audio format for audio storage,
1083 designed as part of its MPEG-1 standard. An MP3 file is made up of
1084 multiple MP3 frames, which consist of a header and a data block. The
1085 size of an MP3 frame depends on the bit rate and on the number
1086 of channels. For a typical CD-audio file (sample rate of 44.1 kHz
1087 stereo), encoded with a bit rate of 128 kbit, an MP3 frame is about
1088 400 bytes large.
1089
1090 *OGG/Vorbis*
1091
1092 OGG is a standardized audio container format, while Vorbis is an
1093 open source codec for lossy audio compression. Since Vorbis is most
1094 commonly made available via the OGG container format, it is often
1095 referred to as OGG/Vorbis. The OGG container format divides data into
1096 chunks called OGG pages. A typical OGG page is about 4KB large. The
1097 Vorbis codec creates variable-bitrate (VBR) data, where the bitrate
1098 may vary considerably.
1099
1100 *OGG/Speex*
1101
1102 Speex is an open-source speech codec that is based on CELP (Code
1103 Excited Linear Prediction) coding. It is designed for voice
1104 over IP applications, has modest complexity and a small memory
1105 footprint. Wideband and narrowband (telephone quality) speech are
1106 supported. As for Vorbis audio, Speex bit-streams are often stored
1107 in OGG files.
1108
1109 *AAC*
1110
1111 Advanced Audio Coding (AAC) is a standardized, lossy compression
1112 and encoding scheme for digital audio which is the default audio
1113 format for Apple's iPhone, iPod, iTunes. Usually MPEG-4 is used as
1114 the container format and audio files encoded with AAC have the .m4a
1115 extension. A typical AAC frame is about 700 bytes large.
1116
1117 *WMA*
1118
1119 Windows Media Audio (WMA) is an audio data compression technology
1120 developed by Microsoft. A WMA file is usually encapsulated in the
1121 Advanced Systems Format (ASF) container format, which also specifies
1122 how meta data about the file is to be encoded. The bit stream of WMA
1123 is composed of superframes, each containing one or more frames of
1124 2048 samples. For 16 bit stereo a WMA superframe is about 8K large.
1125
1126 *FLAC*
1127
1128 The Free Lossless Audio Codec (FLAC) compresses audio without quality
1129 loss. It gives better compression ratios than a general purpose
1130 compressor like zip or bzip2 because FLAC is designed specifically
1131 for audio. A FLAC-encoded file consits of frames of varying size, up
1132 to 16K. Each frame starts with a header that contains all information
1133 necessary to decode the frame.
1134
1135 Meta data
1136 ~~~~~~~~~
1137
1138 Unfortunately, each audio format has its own conventions how meta
1139 data is added as tags to the audio file.
1140
1141 For MP3 files, ID3, version 1 and 2 are widely used. ID3 version 1
1142 is rather simple but also very limited as it supports only artist,
1143 title, album, year and comment tags. Each of these can only be at most
1144 32 characters long. ID3, version 2 is much more flexible but requires
1145 a separate library being installed for paraslash to support it.
1146
1147 Ogg vorbis, ogg speex and flac files contain meta data as Vorbis
1148 comments, which are typically implemented as strings of the form
1149 "[TAG]=[VALUE]". Unlike ID3 version 1 tags, one may use whichever
1150 tags are appropriate for the content.
1151
1152 AAC files usually use the MPEG-4 container format for storing meta
1153 data while WMA files wrap meta data as special objects within the
1154 ASF container format.
1155
1156 paraslash only tracks the most common tags that are supported by
1157 all tag variants: artist, title, year, album, comment. When a file
1158 is added to the AFS database, the meta data of the file is extracted
1159 and stored in the audio file table.
1160
1161 Chunks and chunk tables
1162 ~~~~~~~~~~~~~~~~~~~~~~~
1163
1164 paraslash uses the word "chunk" as common term for the building blocks
1165 of an audio file. For MP3 files, a chunk is the same as an MP3 frame,
1166 while for OGG files a chunk is an OGG page, etc. Therefore the chunk
1167 size varies considerably between audio formats, from a few hundred
1168 bytes (MP3) up to 16K (FLAC).
1169
1170 The chunk table contains the offsets within the audio file that
1171 correspond to the chunk boundaries of the file. Like the meta data,
1172 the chunk table is computed and stored in the database whenever an
1173 audio file is added.
1174
1175 The paraslash senders (see below) always send complete chunks. The
1176 granularity for seeking is therefore determined by the chunk size.
1177
1178 Audio format handlers
1179 ~~~~~~~~~~~~~~~~~~~~~
1180
1181 For each audio format paraslash contains an audio format handler whose
1182 first task is to tell whether a given file is a valid audio file of
1183 this type. If so, the audio file handler extracts some technical data
1184 (duration, sampling rate, number of channels etc.), computes the
1185 chunk table and reads the meta data.
1186
1187 The audio format handler code is linked into para_server and executed
1188 via the _add_ command. The same code is also available as a stand-alone
1189 tool, para_afh, which can be used to print the technical data, the
1190 chunk table and the meta data of a file. Furthermore, one can use
1191 para_afh to cut an audio file, i.e. to select some of its chunks to
1192 produce a new file containing only these chunks.
1193
1194 ----------
1195 Networking
1196 ----------
1197
1198 Paraslash uses different network connections for control and data.
1199 para_client communicates with para_server over a dedicated TCP control
1200 connection. To transport audio data, separate data connections are
1201 used. For these data connections, a variety of transports (UDP, DCCP,
1202 HTTP) can be chosen.
1203
1204 The chapter starts with the REFERENCE(The paraslash control
1205 service, control service), followed by a section on the various
1206 REFERENCE(Streaming protocols, streaming protocols) in which the data
1207 connections are described. The way audio file headers are embedded into
1208 the stream is discussed REFERENCE(Streams with headers and headerless
1209 streams, briefly) before the REFERENCE(Networking examples, example
1210 section) which illustrates typical commands for real-life scenarios.
1211
1212 Both IPv4 and IPv6 are supported.
1213
1214 The paraslash control service
1215 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1216
1217 para_server is controlled at runtime via the paraslash control
1218 connection. This connection is used for server commands (play, stop,
1219 ...) as well as for afs commands (ls, select, ...).
1220
1221 The server listens on a TCP port and accepts connections from clients
1222 that connect the open port. Each connection causes the server to fork
1223 off a client process which inherits the connection and deals with that
1224 client only. In this classical accept/fork approach the server process
1225 is unaffected if the child dies or goes crazy for whatever reason. In
1226 fact, the child process can not change address space of server process.
1227
1228 The section on REFERENCE(Client-server authentication, client-server
1229 authentication) above described the early connection establishment
1230 from the crypto point of view. Here it is described what happens
1231 after the connection (including crypto setup) has been established.
1232 There are four processes involved during command dispatch as sketched
1233 in the following diagram.
1234
1235 <<
1236 <pre>
1237 server_host client_host
1238 ~~~~~~~~~~~ ~~~~~~~~~~~
1239
1240 +-----------+ connect +-----------+
1241 |para_server|<------------------------------ |para_client|
1242 +-----------+ +-----------+
1243 | ^
1244 | fork +---+ |
1245 +----------> |AFS| |
1246 | +---+ |
1247 | ^ |
1248 | | |
1249 | | connect (cookie) |
1250 | | |
1251 | | |
1252 | fork +-----+ inherited connection |
1253 +---------->|child|<--------------------------+
1254 +-----+
1255 </pre>
1256 >>
1257
1258 Note that the child process is not a child of the afs process,
1259 so communication of these two processes has to happen via local
1260 sockets. In order to avoid abuse of the local socket by unrelated
1261 processes, a magic cookie is created once at server startup time just
1262 before the server process forks off the AFS process. This cookie is
1263 known to the server, AFS and the child, but not to unrelated processes.
1264
1265 There are two different kinds of commands: First there are commands
1266 that cause the server to respond with some answer such as the list
1267 of all audio files. All but the addblob commands (addimg, addlyr,
1268 addpl, addmood) are of this kind. The addblob commands add contents
1269 to the database, so they need to transfer data the other way round,
1270 from the client to the server.
1271
1272 There is no knowledge about the server commands built into para_client,
1273 so it does not know about addblob commands. Instead, it inspects the
1274 first data package sent by the server for a magic string. If this
1275 string was found, it sends STDIN to the server, otherwise it dumps
1276 data from the server to STDOUT.
1277
1278 Streaming protocols
1279 ~~~~~~~~~~~~~~~~~~~
1280
1281 A network (audio) stream usually consists of one streaming source,
1282 the _sender_, and one or more _receivers_ which read data over the
1283 network from the streaming source.
1284
1285 Senders are thus part of para_server while receivers are part of
1286 para_audiod. Moreover, there is the stand-alone tool para_recv which
1287 can be used to manually download a stream, either from para_server
1288 or from a web-based audio streaming service.
1289
1290 The following three streaming protocols are supported by paraslash:
1291
1292 - HTTP. Recommended for public streams that can be played by
1293 any player like mpg123, xmms, itunes, winamp, etc. The HTTP
1294 sender is supported on all operating systems and all platforms.
1295
1296 - DCCP. Recommended for LAN streaming. DCCP is currently
1297 available only for Linux.
1298
1299 - UDP. Recommended for multicast LAN streaming.
1300
1301 See the Appendix on REFERENCE(Network protocols, network protocols)
1302 for brief descriptions of the various protocols relevant for network
1303 audio streaming with paraslash.
1304
1305 It is possible to activate more than one sender simultaneously.
1306 Senders can be controlled at run time and via config file and command
1307 line options.
1308
1309 Note that audio connections are _not_ encrypted. Transport or Internet
1310 layer encryption should be used if encrypted data connections are
1311 needed.
1312
1313 Since DCCP and TCP are both connection-oriented protocols, connection
1314 establishment/teardown and access control are very similar between
1315 these two streaming protocols. UDP is the most lightweight option,
1316 since in contrast to TCP/DCCP it is connectionless. It is also the
1317 only protocol supporting IP multicast.
1318
1319 The HTTP and the DCCP sender listen on a (TCP/DCCP) port waiting for
1320 clients to connect and establish a connection via some protocol-defined
1321 handshake mechanism. Both senders maintain two linked lists each:
1322 The list of all clients which are currently connected, and the list
1323 of access control entries which determines who is allowed to connect.
1324 IP-based access control may be configured through config file and
1325 command line options and via the "allow" and "deny" sender subcommands.
1326
1327 Upon receiving a GET request from the client, the HTTP sender sends
1328 back a status line and a message. The body of this message is the
1329 audio stream. This is common practice and is supported by many popular
1330 clients which can thus be used to play a stream offered by para_server.
1331 For DCCP things are a bit simpler: No messages are exchanged between
1332 the receiver and sender. The client simply connects and the sender
1333 starts to stream.
1334
1335 DCCP is an experimental protocol which offers a number of new features
1336 not available for TCP. Both ends can negotiate these features using
1337 a built-in negotiation mechanism. In contrast to TCP/HTTP, DCCP is
1338 datagram-based (no retransmissions) and thus should not be used over
1339 lossy media (e.g. WiFi networks). One useful feature offered by DCCP
1340 is access to a variety of different congestion-control mechanisms
1341 called CCIDs. Two different CCIDs are available per default on Linux:
1342
1343
1344 - _CCID 2_. A Congestion Control mechanism similar to that
1345 of TCP. The sender maintains a congestion window and halves
1346 this window in response to congestion.
1347
1348
1349 - _CCID-3_. Designed to be fair when competing for bandwidth.
1350 It has lower variation of throughput over time compared with
1351 TCP, which makes it suitable for streaming media.
1352
1353 Unlike the HTTP and DCCP senders, the UDP sender maintains only a
1354 single list, the _target list_. This list describes the set of clients
1355 to which the stream is sent. There is no list for access control and
1356 no "allow" and "deny" commands for the UDP sender. Instead, the "add"
1357 and "delete" commands can be used to modify the target list.
1358
1359 Since both UDP and DCCP offer an unreliable datagram-based transport,
1360 additional measures are necessary to guard against disruptions over
1361 networks that are lossy or which may be subject to interference (as
1362 is for instance the case with WiFi). Paraslash uses FEC (Forward
1363 Error Correction) to guard against packet losses and reordering. The
1364 stream is FEC-encoded before it is sent through the UDP socket and
1365 must be decoded accordingly on the receiver side.
1366
1367 The packet size and the amount of redundancy introduced by FEC can
1368 be configured via the FEC parameters which are dictated by server
1369 and may also be configured through the "sender" command. The FEC
1370 parameters are encoded in the header of each network packet, so no
1371 configuration is necessary on the receiver side. See the section on
1372 REFERENCE(Forward error correction, FEC) below.
1373
1374 Streams with headers and headerless streams
1375 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1376
1377 For OGG/Vorbis, OGG/Speex and wma streams, some of the information
1378 needed to decode the stream is only contained in the audio file
1379 header of the container format but not in each data chunk. Clients
1380 must be able to obtain this information in case streaming starts in
1381 the middle of the file or if para_audiod is started while para_server
1382 is already sending a stream.
1383
1384 This is accomplished in different ways, depending on the streaming
1385 protocol. For connection-oriented streams (HTTP, DCCP) the audio file
1386 header is sent prior to audio file data. This technique however does
1387 not work for the connectionless UDP transport. Hence the audio file
1388 header is periodically being embedded into the UDP audio data stream.
1389 By default, the header is resent after five seconds. The receiver has
1390 to wait until the next header arrives before it can start decoding
1391 the stream.
1392
1393 Examples
1394 ~~~~~~~~
1395
1396 The sender command of para_server allows to (de-)activate senders
1397 and to change the access permissions senders at runtime. The "si"
1398 (server info) command is used to list the streaming options of the
1399 currently running server as well as the various sender access lists.
1400
1401 -> Show client/target/access lists:
1402
1403 para_client si
1404
1405 -> Obtain general help for the sender command:
1406
1407 para_client help sender
1408
1409 -> Get help for a specific sender (contains further examples):
1410
1411 s=http # or dccp or udp
1412 para_client sender $s help
1413
1414 By default para_server activates both the HTTP and th DCCP sender on
1415 startup. This can be changed via command line options or para_server's
1416 config file.
1417
1418 -> List config file options for senders:
1419
1420 para_server -h
1421
1422 All senders share the "on" and "off" commands, so senders may be
1423 activated and deactivated independently of each other.
1424
1425 -> Switch off the http sender:
1426
1427 para_client sender http off
1428
1429 -> Receive a DCCP stream using CCID2 and write the output into a file:
1430
1431 host=foo.org; ccid=2; filename=bar
1432 para_recv --receiver "dccp --host $host --ccid $ccid" > $filename
1433
1434 Note the quotes around the arguments for the dccp receiver. Each
1435 receiver has its own set of command line options and its own command
1436 line parser, so arguments for the dccp receiver must be protected
1437 from being interpreted by para_recv.
1438
1439 -> Start UDP multicast, using the default multicast address:
1440
1441 para_client sender udp add 224.0.1.38
1442
1443 -> Receive FEC-encoded multicast stream and write the output into a file:
1444
1445 filename=foo
1446 para_recv -r udp > $filename
1447
1448 -> Add an UDP unicast for a client to the target list of the UDP sender:
1449
1450 t=client.foo.org
1451 para_client sender udp add $t
1452
1453 -> Receive this (FEC-encoded) unicast stream:
1454
1455 filename=foo
1456 para_recv -r 'udp -i 0.0.0.0' > $filename
1457
1458 -> Create a minimal config for para_audiod for HTTP streams:
1459
1460 c=$HOME/.paraslash/audiod.conf.min; s=server.foo.com
1461 echo receiver \".:http -i $s\" > $c
1462 para_audiod --config $c
1463
1464 -------
1465 Filters
1466 -------
1467
1468 A paraslash filter is a module which transforms an input stream into
1469 an output stream. Filters are included in the para_audiod executable
1470 and in the stand-alone tool para_filter which usually contains the
1471 same modules.
1472
1473 While para_filter reads its input stream from STDIN and writes
1474 the output to STDOUT, the filter modules of para_audiod are always
1475 connected to a receiver which produces the input stream and a writer
1476 which absorbs the output stream.
1477
1478 Some filters depend on a specific library being installed and are
1479 not compiled in if this library was not found at compile time. To
1480 see the list of supported filters, run para_filter and para_audiod
1481 with the --help option. The output looks similar to the following:
1482
1483 Available filters:
1484 compress wav amp fecdec wmadec prebuffer oggdec aacdec mp3dec
1485
1486 Out of these filter modules, a chain of filters can be constructed,
1487 much in the way Unix pipes can be chained, and analogous to the use
1488 of modules in gstreamer: The output of the first filter becomes the
1489 input of the second filter. There is no limitation on the number of
1490 filters and the same filter may occur more than once.
1491
1492 Like receivers, each filter has its own command line options which
1493 must be quoted to protect them from the command line options of
1494 the driving application (para_audiod or para_filter). Example:
1495
1496 para_filter -f 'mp3dec --ignore-crc' -f 'compress --damp 1'
1497
1498 For para_audiod, each audio format has its own set of filters. The
1499 name of the audio format for which the filter should be applied can
1500 be used as the prefix for the filter option. Example:
1501
1502 para_audiod -f 'mp3:prebuffer --duration 300'
1503
1504 The "mp3" prefix above is actually interpreted as a POSIX extended
1505 regular expression. Therefore
1506
1507 para_audiod -f '.:prebuffer --duration 300'
1508
1509 activates the prebuffer filter for all supported audio formats (because
1510 "." matches all audio formats) while
1511
1512 para_audiod -f 'wma|ogg:prebuffer --duration 300'
1513
1514 activates it only for wma and ogg streams.
1515
1516 Decoders
1517 ~~~~~~~~
1518
1519 For each supported audio format there is a corresponding filter
1520 which decodes audio data in this format to 16 bit PCM data which
1521 can be directly sent to the sound device or any other software that
1522 operates on undecoded PCM data (visualizers, equalizers etc.). Such
1523 filters are called _decoders_ in general, and xxxdec is the name of
1524 the paraslash decoder for the audio format xxx. For example, the mp3
1525 decoder filter is called mp3dec.
1526
1527 Note that the output of the decoder is about 10 times larger than
1528 its input. This means that filters that operate on the decoded audio
1529 stream have to deal with much more data than filters that transform
1530 the audio stream before it is fed to the decoder.
1531
1532 Paraslash relies on external libraries for most decoders, so these
1533 libraries must be installed for the decoder to be included in the
1534 para_filter and para_audiod executables. The oggdec filter depends
1535 on the libogg and libvorbis libraries for example.
1536
1537 Forward error correction
1538 ~~~~~~~~~~~~~~~~~~~~~~~~
1539
1540 As already mentioned REFERENCE(Streaming protocols, earlier),
1541 paraslash uses forward error correction (FEC) for the unreliable UDP
1542 and DCCP transports. FEC is a technique which was invented already
1543 in 1960 by Reed and Solomon and which is widely used for the parity
1544 calculations of storage devices (RAID arrays). It is based on the
1545 algebraic concept of finite fields, today called Galois fields, in
1546 honour of the mathematician Galois (1811-1832). The FEC implementation
1547 of paraslash is based on code by Luigi Rizzo.
1548
1549 Although the details require a sound knowledge of the underlying
1550 mathematics, the basic idea is not hard to understand: For positive
1551 integers k and n with k < n it is possible to compute for any k given
1552 data bytes d_1, ..., d_k the corresponding r := n -k parity bytes p_1,
1553 ..., p_r such that all data bytes can be reconstructed from *any*
1554 k bytes of the set
1555
1556 {d_1, ..., d_k, p_1, ..., p_r}.
1557
1558 FEC-encoding for unreliable network transports boils down to slicing
1559 the audio stream into groups of k suitably sized pieces called _slices_
1560 and computing the r corresponding parity slices. This step is performed
1561 in para_server which then sends both the data and the parity slices
1562 over the unreliable network connection. If the client was able
1563 to receive at least k of the n = k + r slices, it can reconstruct
1564 (FEC-decode) the original audio stream.
1565
1566 From these observations it is clear that there are three different
1567 FEC parameters: The slice size, the number of data slices k, and the
1568 total number of slices n. It is crucial to choose the slice size
1569 such that no fragmentation of network packets takes place because
1570 FEC only guards against losses and reordering but fails if slices are
1571 received partially.
1572
1573 FEC decoding in paralash is performed through the fecdec filter which
1574 usually is the first filter (there can be other filters before fecdec
1575 if these do not alter the audio stream).
1576
1577
1578 Volume adjustment (amp and compress)
1579 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1580
1581 The amp and the compress filter both adjust the volume of the audio
1582 stream. These filters operate on uncompressed audio samples. Hence
1583 they are usually placed directly after the decoding filter. Each
1584 sample is multiplied with a scaling factor (>= 1) which makes amp
1585 and compress quite expensive in terms of computing power.
1586
1587 *amp*
1588
1589 The amp filter amplifies the audio stream by a fixed scaling factor
1590 that must be known in advance. For para_audiod this factor is derived
1591 from the amplification field of the audio file's entry in the audio
1592 file table while para_filter uses the value given at the command line.
1593
1594 The optimal scaling factor F for an audio file is the largest real
1595 number F >= 1 such that after multiplication with F all samples still
1596 fit into the sample interval [-32768, 32767]. One can use para_filter
1597 in combination with the sox utility to compute F:
1598
1599 para_filter -f mp3dec -f wav < file.mp3 | sox -t wav - -e stat -v
1600
1601 The amplification value V which is stored in the audio file table,
1602 however, is an integer between 0 and 255 which is connected to F
1603 through the formula
1604
1605 V = (F - 1) * 64.
1606
1607 To store V in the audio file table, the command
1608
1609 para_client -- touch -a=V file.mp3
1610
1611 is used. The reader is encouraged to write a script that performs
1612 these computations :)
1613
1614 *compress*
1615
1616 Unlike the amplification filter, the compress filter adjusts the volume
1617 of the audio stream dynamically without prior knowledge about the peak
1618 value. It maintains the maximal volume of the last n samples of the
1619 audio stream and computes a suitable amplification factor based on that
1620 value and the various configuration options. It tries to chose this
1621 factor such that the adjusted volume meets the desired target level.
1622
1623 Note that it makes sense to combine amp and compress.
1624
1625 Misc filters (wav and prebuffer)
1626 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1627
1628 These filters are rather simple and do not modify the audio stream at
1629 all. The wav filter is only useful with para_filter and in connection
1630 with a decoder. It asks the decoder for the number of channels and the
1631 sample rate of the stream and adds a Microsoft wave header containing
1632 this information at the beginning. This allows to write wav files
1633 rather than raw PCM files (which do not contain any information about
1634 the number of channels and the sample rate).
1635
1636 The prebuffer filter simply delays the output until the given time has
1637 passed (starting from the time the first byte was available in its
1638 input queue) or until the given amount of data has accumulated. It
1639 is mainly useful for para_audiod if the standard parameters result
1640 in buffer underruns.
1641
1642 Both filters require almost no additional computing time, even when
1643 operating on uncompressed audio streams, since data buffers are simply
1644 "pushed down" rather than copied.
1645
1646 Examples
1647 ~~~~~~~~
1648
1649 -> Decode an mp3 file to wav format:
1650
1651 para_filter -f mp3dec -f wav < file.mp3 > file.wav
1652
1653 -> Amplify a raw audio file by a factor of 1.5:
1654
1655 para_filter -f amp --amp 32 < foo.raw > bar.raw
1656
1657 ------
1658 Output
1659 ------
1660
1661 Once an audio stream has been received and decoded to PCM format,
1662 it can be sent to a sound device for playback. This part is performed
1663 by paraslash _writers_ which are described in this chapter.
1664
1665 Writers
1666 ~~~~~~~
1667
1668 A paraslash writer acts as a data sink that consumes but does not
1669 produce audio data. Paraslash writers operate on the client side and
1670 are contained in para_audiod and in the stand-alone tool para_write.
1671
1672 The para_write program reads uncompressed audio data from STDIN. If
1673 this data starts with a wav header, sample rate, sample format and
1674 channel count are read from the header. Otherwise CD audio (44.1KHz
1675 16 bit little endian, stereo) is assumed but this can be overridden
1676 by command line options. para_audiod, on the other hand, obtains
1677 the sample rate and the number of channels from the decoder.
1678
1679 Like receivers and filters, each writer has an individual set of
1680 command line options, and for para_audiod writers can be configured
1681 per audio format separately. It is possible to activate more than
1682 one writer for the same stream simultaneously.
1683
1684 OS-dependent APIs
1685 ~~~~~~~~~~~~~~~~~
1686
1687 Unfortunately, the various flavours of Unix on which paraslash
1688 runs on have different APIs for opening a sound device and starting
1689 playback. Hence for each such API there is a paraslash writer that
1690 can play the audio stream via this API.
1691
1692 *ALSA*. The _Advanced Linux Sound Architecture_ is only available on
1693 Linux systems. Although there are several mid-layer APIs in use by
1694 the various Linux distributions (ESD, Jack, PulseAudio), paraslash
1695 currently supports only the low-level ALSA API which is not supposed
1696 to be change. ALSA is very feature-rich, in particular it supports
1697 software mixing via its DMIX plugin. ALSA is the default writer on
1698 Linux systems.
1699
1700 *OSS*. The _Open Sound System_ is the only API on *BSD Unixes and
1701 is also available on Linux systems, usually provided by ALSA as an
1702 emulation for backwards compatibility. This API is rather simple but
1703 also limited. For example only one application can open the device
1704 at any time. The OSS writer is activated by default on BSD Systems.
1705
1706 *OSX*. Mac OS X has yet another API called CoreAudio. The OSX writer
1707 for this API is only compiled in on such systems and is of course
1708 the default there.
1709
1710 *FILE*. The file writer allows to capture the audio stream and
1711 write the PCM data to a file on the file system rather than playing
1712 it through a sound device. It is supported on all platforms and is
1713 always compiled in.
1714
1715 *AO*. _Libao_ is a cross-platform audio library which supports a wide
1716 variety of platforms including PulseAudio (gnome), ESD (Enlightened
1717 Sound Daemon), AIX, Solaris and IRIX. The ao writer plays audio
1718 through an output plugin of libao.
1719
1720 Examples
1721 ~~~~~~~~
1722
1723 -> Use the OSS writer to play a wav file:
1724
1725 para_write --writer oss < file.wav
1726
1727 -> Enable ALSA software mixing for mp3 streams
1728
1729 para_audiod --writer 'mp3:alsa -d plug:swmix'
1730
1731
1732 ---
1733 Gui
1734 ---
1735
1736 para_gui executes an arbitrary command which is supposed to print
1737 status information to STDOUT. It then displays this information in
1738 a curses window. By default the command
1739
1740 para_audioc -- stat -p
1741
1742 is executed, but this can be customized via the --stat_cmd option. In
1743 particular it possible to use
1744
1745 para_client -- stat -p
1746
1747 to make para_gui work on systems on which para_audiod is not running.
1748
1749 Key bindings
1750 ~~~~~~~~~~~~
1751
1752 It is possible to bind keys to arbitrary commands via custom
1753 key-bindings. Besides the internal keys which can not be changed (help,
1754 quit, loglevel, version...), the following flavours of key-bindings
1755 are supported:
1756
1757 - external: Shutdown curses before launching the given command.
1758 Useful for starting other ncurses programs from within
1759 para_gui, e.g. aumix or dialog scripts. Or, use the mbox
1760 output format to write a mailbox containing one mail for each
1761 (admissible) file the audio file selector knows about. Then
1762 start mutt from within para_gui to browse your collection!
1763
1764 - display: Launch the command and display its stdout in
1765 para_gui's bottom window.
1766
1767 - para: Like display, but start "para_client <specified
1768 command>" instead of "<specified command>".
1769
1770 The general form of a key binding is
1771
1772 key_map k:m:c
1773
1774 which maps key k to command c using mode m. Mode may be x, d or p
1775 for external, display and paraslash commands, respectively.
1776
1777 Themes
1778 ~~~~~~
1779
1780 Currently there are only two themes for para_gui. It is easy, however,
1781 to add more themes. To create a new theme one has to define the
1782 position, color and geometry for for each status item that should be
1783 shown by this theme. See gui_theme.c for examples.
1784
1785 The "." and "," keys are used to switch between themes.
1786
1787 Examples
1788 ~~~~~~~~
1789
1790 -> Show server info:
1791
1792 key_map "i:p:si"
1793
1794 -> Jump to the middle of the current audio file by pressing F5:
1795
1796 key_map "<F5>:p:jmp 50"
1797
1798 -> vi-like bindings for jumping around:
1799
1800 key_map "l:p:ff 10"
1801 key_map "h:p:ff 10-"
1802 key_map "w:p:ff 60"
1803 key_map "b:p:ff 60-"
1804
1805 -> Print the current date and time:
1806
1807 key_map "D:d:date"
1808
1809 -> Call other curses programs:
1810
1811 key_map "U:x:aumix"
1812 key_map "!:x:/bin/bash"
1813 key_map "^E:x:/bin/sh -c 'vi ~/.paraslash/gui.conf'"
1814
1815 -----------
1816 Development
1817 -----------
1818
1819 Tools
1820 ~~~~~
1821
1822 In order to compile the sources from the git repository (rather than
1823 from tar balls) and for contributing non-trivial changes to the
1824 paraslash project, some additional tools should be installed on a
1825 developer machine.
1826
1827 http://git.or.cz/ (git). As described in more detail REFERENCE(Git
1828 branches, below), the git source code management tool is used for
1829 paraslash development. It is necessary for cloning the git repository
1830 and for getting updates.
1831
1832 ftp://ftp.gnu.org/pub/gnu/m4/ (m4). Some input files for gengetopt
1833 are generated from templates by the m4 macro processor.
1834
1835 ftp://ftp.gnu.org/pub/gnu/autoconf/ (autoconf) GNU autoconf creates
1836 the configure file which is shipped in the tarballs but has to be
1837 generated when compiling from git.
1838
1839 http://www.triptico.com/software/grutatxt.html (grutatxt). The
1840 HTML version of this manual and some of the paraslash web pages are
1841 generated by the grutatxt plain text to HTML converter. If changes
1842 are made to these text files the grutatxt package must be installed
1843 to regenerate the HTML files.
1844
1845 http://www.stack.nl/~dimitri/doxygen/ (doxygen). The documentation
1846 of paraslash's C sources uses the doxygen documentation system. The
1847 conventions for documenting the source code is described in the
1848 REFERENCE(Doxygen, Doxygen section).
1849
1850 ftp://ftp.gnu.org/pub/gnu/global (global). This is used to generate
1851 browsable HTML from the C sources. It is needed by doxygen.
1852
1853 Git branches
1854 ~~~~~~~~~~~~
1855
1856 Paraslash has been developed using the git source code management
1857 tool since 2006. Development is organized roughly in the same spirit
1858 as the git development itself, as described below.
1859
1860 The following text passage is based on "A note from the maintainer",
1861 written by Junio C Hamano, the maintainer of git.
1862
1863 There are four branches in the paraslash repository that track the
1864 source tree: "master", "maint", "next", and "pu".
1865
1866 The "master" branch is meant to contain what is well tested and
1867 ready to be used in a production setting. There could occasionally be
1868 minor breakages or brown paper bag bugs but they are not expected to
1869 be anything major, and more importantly quickly and easily fixable.
1870 Every now and then, a "feature release" is cut from the tip of this
1871 branch, named with three dotted decimal digits, like 0.4.2.
1872
1873 Whenever changes are about to be included that will eventually lead to
1874 a new major release (e.g. 0.5.0), a "maint" branch is forked off from
1875 "master" at that point. Obvious, safe and urgent fixes after the major
1876 release are applied to this branch and maintenance releases are cut
1877 from it. New features never go to this branch. This branch is also
1878 merged into "master" to propagate the fixes forward.
1879
1880 A trivial and safe enhancement goes directly on top of "master".
1881 New development does not usually happen on "master", however.
1882 Instead, a separate topic branch is forked from the tip of "master",
1883 and it first is tested in isolation; Usually there are a handful such
1884 topic branches that are running ahead of "master". The tip of these
1885 branches is not published in the public repository to keep the number
1886 of branches that downstream developers need to worry about low.
1887
1888 The quality of topic branches varies widely. Some of them start out as
1889 "good idea but obviously is broken in some areas" and then with some
1890 more work become "more or less done and can now be tested by wider
1891 audience". Luckily, most of them start out in the latter, better shape.
1892
1893 The "next" branch is to merge and test topic branches in the latter
1894 category. In general, this branch always contains the tip of "master".
1895 It might not be quite rock-solid production ready, but is expected to
1896 work more or less without major breakage. The maintainer usually uses
1897 the "next" version of paraslash for his own pleasure, so it cannot
1898 be _that_ broken. The "next" branch is where new and exciting things
1899 take place.
1900
1901 The two branches "master" and "maint" are never rewound, and "next"
1902 usually will not be either (this automatically means the topics that
1903 have been merged into "next" are usually not rebased, and you can find
1904 the tip of topic branches you are interested in from the output of
1905 "git log next"). You should be able to safely build on top of them.
1906
1907 However, at times "next" will be rebuilt from the tip of "master" to
1908 get rid of merge commits that will never be in "master". The commit
1909 that replaces "next" will usually have the identical tree, but it
1910 will have different ancestry from the tip of "master".
1911
1912 The "pu" (proposed updates) branch bundles the remainder of the
1913 topic branches. The "pu" branch, and topic branches that are only in
1914 "pu", are subject to rebasing in general. By the above definition
1915 of how "next" works, you can tell that this branch will contain quite
1916 experimental and obviously broken stuff.
1917
1918 When a topic that was in "pu" proves to be in testable shape, it
1919 graduates to "next". This is done with
1920
1921 git checkout next
1922 git merge that-topic-branch
1923
1924 Sometimes, an idea that looked promising turns out to be not so good
1925 and the topic can be dropped from "pu" in such a case.
1926
1927 A topic that is in "next" is expected to be polished to perfection
1928 before it is merged to "master". Similar to the above, this is
1929 done with
1930
1931 git checkout master
1932 git merge that-topic-branch
1933 git branch -d that-topic-branch
1934
1935 Note that being in "next" is not a guarantee to appear in the next
1936 release (being in "master" is such a guarantee, unless it is later
1937 found seriously broken and reverted), nor even in any future release.
1938
1939 Coding Style
1940 ~~~~~~~~~~~~
1941
1942 The preferred coding style for paraslash coincides more or less
1943 with the style of the Linux kernel. So rather than repeating what is
1944 written XREFERENCE(http://www.kernel.org/doc/Documentation/CodingStyle,
1945 there), here are the most important points.
1946
1947 - Burn the GNU coding standards.
1948 - Never use spaces for indentation.
1949 - Tabs are 8 characters, and thus indentations are also 8 characters.
1950 - Don't put multiple assignments on a single line.
1951 - Avoid tricky expressions.
1952 - Don't leave whitespace at the end of lines.
1953 - The limit on the length of lines is 80 columns.
1954 - Use K&R style for placing braces and spaces:
1955
1956 if (x is true) {
1957 we do y
1958 }
1959
1960 - Use a space after (most) keywords.
1961 - Do not add spaces around (inside) parenthesized expressions.
1962 - Use one space around (on each side of) most binary and ternary operators.
1963 - Do not use cute names like ThisVariableIsATemporaryCounter, call it tmp.
1964 - Mixed-case names are frowned upon.
1965 - Descriptive names for global variables are a must.
1966 - Avoid typedefs.
1967 - Functions should be short and sweet, and do just one thing.
1968 - The number of local variables shouldn't exceed 10.
1969 - Gotos are fine if they improve readability and reduce nesting.
1970 - Don't use C99-style "// ..." comments.
1971 - Names of macros defining constants and labels in enums are capitalized.
1972 - Enums are preferred when defining several related constants.
1973 - Always use the paraslash wrappers for allocating memory.
1974 - If the name of a function is an action or an imperative.
1975 command, the function should return an error-code integer
1976 (<0 means error, >=0 means success). If the name is a
1977 predicate, the function should return a "succeeded" boolean.
1978
1979
1980 Doxygen
1981 ~~~~~~~
1982
1983 Doxygen is a documentation system for various programming
1984 languages. The paraslash project uses Doxygen for generating the API
1985 reference on the web pages, but good source code documentation is
1986 also beneficial to people trying to understand the code structure
1987 and the interactions between the various source files.
1988
1989 It is more illustrative to look at the source code for examples than
1990 to describe the conventions for documenting the source in this manual,
1991 so we only describe which parts of the code need doxygen comments,
1992 but leave out details on documentation conventions.
1993
1994 As a rule, only the public part of the C source is documented with
1995 Doxygen. This includes structures, defines and enumerations in header
1996 files as well as public (non-static) C functions. These should be
1997 documented completely. For example each parameter and the return
1998 value of a public function should get a descriptive comment.
1999
2000 No doxygen comments are necessary for static functions and for
2001 structures and enumerations in C files (which are used only within
2002 this file). This does not mean, however, that those entities need
2003 no documentation at all. Instead, common sense should be applied to
2004 document what is not obvious from reading the code.
2005
2006 --------
2007 Appendix
2008 --------
2009
2010 Network protocols
2011 ~~~~~~~~~~~~~~~~~
2012
2013 *IP*. The _Internet Protocol_ is the primary networking protocol
2014 used for the Internet. All protocols described below use IP as the
2015 underlying layer. Both the prevalent IPv4 and the next-generation
2016 IPv6 variant are being deployed actively worldwide.
2017
2018 *Connection-oriented and connectionless protocols*. Connectionless
2019 protocols differ from connection-oriented ones in that state
2020 associated with the sending/receiving endpoints is treated
2021 implicitly. Connectionless protocols maintain no internal knowledge
2022 about the state of the connection. Hence they are not capable of
2023 reacting to state changes, such as sudden loss or congestion on the
2024 connection medium. Connection-oriented protocols, in contrast, make
2025 this knowledge explicit. The connection is established only after
2026 a bidirectional handshake which requires both endpoints to agree
2027 on the state of the connection, and may also involve negotiating
2028 specific parameters for the particular connection. Maintaining an
2029 up-to-date internal state of the connection also in general means
2030 that the sending endpoints perform congestion control, adapting to
2031 qualitative changes of the connection medium.
2032
2033 *Reliability*. In IP networking, packets can be lost, duplicated,
2034 or delivered out of order, and different network protocols handle
2035 these problems in different ways. We call a transport-layer protocol
2036 _reliable_, if it turns the unreliable IP delivery into an ordered,
2037 duplicate- and loss-free delivery of packets. Sequence numbers
2038 are used to discard duplicates and re-arrange packets delivered
2039 out-of-order. Retransmission is used to guarantee loss-free
2040 delivery. Unreliable protocols, in contrast, do not guarantee ordering
2041 or data integrity.
2042
2043 *Classification*. With these definitions the protocols which are used
2044 by paraslash for steaming audio data may be classified as follows.
2045
2046 - HTTP/TCP: connection-oriented, reliable,
2047 - UDP: connectionless, unreliable,
2048 - DCCP: connection-oriented, unreliable.
2049
2050 Below we give a short descriptions of these protocols.
2051
2052 *TCP*. The _Transmission Control Protocol_ provides reliable,
2053 ordered delivery of a stream and a classic window-based congestion
2054 control. In contrast to UDP and DCCP (see below), TCP does not have
2055 record-oriented or datagram-based syntax, i.e. it provides a stream
2056 which is unaware and independent of any record (packet) boundaries.
2057 TCP is used extensively by many application layers. Besides HTTP (the
2058 Hypertext Transfer Protocol), also FTP (the File Transfer protocol),
2059 SMTP (Simple Mail Transfer Protocol), SSH (Secure Shell) all sit on
2060 top of TCP.
2061
2062 *UDP*. The _User Datagram Protocol_ is the simplest transport-layer
2063 protocol, built as a thin layer directly on top of IP. For this reason,
2064 it offers the same best-effort service as IP itself, i.e. there is no
2065 detection of duplicate or reordered packets. Being a connectionless
2066 protocol, only minimal internal state about the connection is
2067 maintained, which means that there is no protection against packet
2068 loss or network congestion. Error checking and correction (if at all)
2069 are performed in the application.
2070
2071 *DCCP*. The _Datagram Congestion Control Protocol_ combines the
2072 connection-oriented state maintenance known from TCP with the
2073 unreliable, datagram-based transport of UDP. This means that it
2074 is capable of reacting to changes in the connection by performing
2075 congestion control, offering multiple alternative approaches. But it
2076 is bound to datagram boundaries (the maximum packet size supported
2077 by a medium), and like UDP it lacks retransmission to protect
2078 against loss. Due to the use of sequence numbers, it is however
2079 able to react to loss (interpreted as a congestion indication) and
2080 to ignore out-of-order and duplicate packets. Unlike TCP it allows
2081 to negotiate specific, binding features for a connection, such as
2082 the choice of congestion control: classic, window-based congestion
2083 control known from TCP is available as CCID-2, rate-based, "smooth"
2084 congestion control is offered as CCID-3.
2085
2086 *HTTP*. _The Hypertext Transfer Protocol_ is an application layer
2087 protocol on top of TCP. It is spoken by web servers and is most often
2088 used for web services. However, as can be seen by the many Internet
2089 radio stations and YouTube/Flash videos, http is by far not limited to
2090 the delivery of web pages only. Being a simple request/response based
2091 protocol, the semantics of the protocol also allow the delivery of
2092 multimedia content, such as audio over http.
2093
2094 *Multicast*. IP multicast is not really a protocol but a technique
2095 for one-to-many communication over an IP network. The challenge is to
2096 deliver information to a group of destinations simultaneously using
2097 the most efficient strategy to send the messages over each link of
2098 the network only once. This has benefits for streaming multimedia:
2099 the standard one-to-one unicast offered by TCP/DCCP means that
2100 n clients listening to the same stream also consume n-times the
2101 resources, whereas multicast requires to send the stream just once,
2102 irrespective of the number of receivers. Since it would be costly to
2103 maintain state for each listening receiver, multicast often implies
2104 connectionless transport, which is the reason that it is currently
2105 only available via UDP.
2106
2107 License
2108 ~~~~~~~
2109
2110 Paraslash is licensed under the GPL, version 2. Most of the code
2111 base has been written from scratch, and those parts are GPL V2
2112 throughout. Notable exceptions are FEC and the WMA decoder. See the
2113 corresponding source files for licencing details for these parts. Some
2114 code sniplets of several other third party software packages have
2115 been incorporated into the paraslash sources, for example log message
2116 coloring was taken from the git sources. These third party software
2117 packages are all published under the GPL or some other license
2118 compatible to the GPL.
2119
2120 Acknowledgements
2121 ~~~~~~~~~~~~~~~~
2122
2123 Many thanks to Gerrit Renker who read an early draft of this manual
2124 and contributed significant improvements.
2125
2126 ----------
2127 References
2128 ----------
2129
2130 Articles
2131 ~~~~~~~~
2132 - Reed, Irving S.; Solomon, Gustave (1960),
2133 XREFERENCE(http://kom.aau.dk/~heb/kurser/NOTER/KOFA01.PDF,
2134 Polynomial Codes over Certain Finite Fields), Journal of the
2135 Society for Industrial and Applied Mathematics (SIAM) 8 (2):
2136 300-304, doi:10.1137/0108018)
2137
2138 RFCs
2139 ~~~~
2140
2141 - XREFERENCE(http://www.ietf.org/rfc/rfc768.txt, RFC 768) (1980):
2142 User Datagram Protocol
2143 - XREFERENCE(http://www.ietf.org/rfc/rfc791.txt, RFC 791) (1981):
2144 Internet Protocol
2145 - XREFERENCE(http://www.ietf.org/rfc/rfc2437.txt, RFC 2437) (1998):
2146 RSA Cryptography Specifications
2147 - XREFERENCE(http://www.ietf.org/rfc/rfc4340.txt, RFC 4340)
2148 (2006): Datagram Congestion Control Protocol (DCCP)
2149 - XREFERENCE(http://www.ietf.org/rfc/rfc4341.txt, RFC 4341) (2006):
2150 Congestion Control ID 2: TCP-like Congestion Control
2151 - XREFERENCE(http://www.ietf.org/rfc/rfc4342.txt, RFC 4342) (2006):
2152 Congestion Control ID 3: TCP-Friendly Rate Control (TFRC)
2153
2154 Application web pages
2155 ~~~~~~~~~~~~~~~~~~~~~
2156
2157 - XREFERENCE(http://paraslash.systemlinux.org/, paraslash)
2158 - XREFERENCE(http://xmms2.org/wiki/Main_Page, xmms)
2159 - XREFERENCE(http://www.mpg123.de/, mpg123)
2160 - XREFERENCE(http://gstreamer.freedesktop.org/, gstreamer)
2161 - XREFERENCE(http://www.icecast.org/, icecast)
2162 - XREFERENCE(http://beesbuzz.biz/code/audiocompress.php, Audio Compress)
2163
2164 External documentation
2165 ~~~~~~~~~~~~~~~~~~~~~~
2166
2167 - XREFERENCE(http://kernel.org/pub/linux/kernel/people/hpa/raid6.pdf,
2168 H. Peter Anvin: The mathematics of Raid6)
2169 - XREFERENCE(http://info.iet.unipi.it/~luigi/fec_ccr.ps.gz,
2170 Luigi Rizzo: Effective Erasure Codes for reliable Computer
2171 Communication Protocols)
2172
2173 Code
2174 ~~~~
2175 - XREFERENCE(http://info.iet.unipi.it/~luigi/vdm.tar.gz,
2176 Original FEC implementation by Luigi Rizzo)
2177