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