docs/unix-phil

annotate unix-phil.ms @ 40:422679bdf384

find changesets by author, revision, files, or words in the commit message
rework in ch01
author meillo@marmaro.de
date Thu, 08 Apr 2010 23:11:28 +0200
parents c87143793d82
children 13ef7042fa28
rev   line source
meillo@2 1 .\".if n .pl 1000i
meillo@36 2 .nr PS 11
meillo@36 3 .nr VS 13
meillo@0 4 .de XX
meillo@0 5 .pl 1v
meillo@0 6 ..
meillo@0 7 .em XX
meillo@1 8 .\".nr PI 0
meillo@1 9 .\".if t .nr PD .5v
meillo@1 10 .\".if n .nr PD 1v
meillo@0 11 .nr lu 0
meillo@0 12 .de CW
meillo@0 13 .nr PQ \\n(.f
meillo@0 14 .if t .ft CW
meillo@17 15 .ie ^\\$1^^ .if n .ul 999
meillo@0 16 .el .if n .ul 1
meillo@17 17 .if t .if !^\\$1^^ \&\\$1\f\\n(PQ\\$2
meillo@0 18 .if n .if \\n(.$=1 \&\\$1
meillo@0 19 .if n .if \\n(.$>1 \&\\$1\c
meillo@0 20 .if n .if \\n(.$>1 \&\\$2
meillo@0 21 ..
meillo@0 22 .ds [. \ [
meillo@0 23 .ds .] ]
meillo@1 24 .\"----------------------------------------
meillo@0 25 .TL
meillo@6 26 Why the Unix Philosophy still matters
meillo@0 27 .AU
meillo@0 28 markus schnalke <meillo@marmaro.de>
meillo@0 29 .AB
meillo@1 30 .ti \n(.iu
meillo@39 31 This paper explains the importance of the Unix Philosophy for software design.
meillo@0 32 Today, few software designers are aware of these concepts,
meillo@39 33 and thus a lot of modern software is more limited than necessary
meillo@39 34 and makes less use of software leverage than possible.
meillo@38 35 Knowing and following the guidelines of the Unix Philosophy makes software more valuable.
meillo@0 36 .AE
meillo@0 37
meillo@10 38 .\".if t .2C
meillo@2 39
meillo@2 40 .FS
meillo@2 41 .ps -1
meillo@39 42 This paper was prepared for the ``Software Analysis'' seminar at University Ulm.
meillo@39 43 Mentor was professor Schweiggert. 2010-04-05
meillo@2 44 .br
meillo@39 45 You may retrieve this document from
meillo@39 46 .CW \s-1http://marmaro.de/docs \ .
meillo@2 47 .FE
meillo@2 48
meillo@0 49 .NH 1
meillo@0 50 Introduction
meillo@0 51 .LP
meillo@40 52 The Unix Philosophy is the essence of how the Unix operating system,
meillo@40 53 especially its toolchest, was designed.
meillo@40 54 It is no limited set of fixed rules,
meillo@40 55 but a loose set of guidelines which tell how to write software that
meillo@40 56 suites well into Unix.
meillo@40 57 Actually, the Unix Philosophy describes what is common to typical Unix software.
meillo@40 58 The Wikipedia has an accurate definition:
meillo@40 59 .[
meillo@40 60 %A Wikipedia
meillo@40 61 %T Unix philosophy
meillo@40 62 %P Wikipedia, The Free Encyclopedia
meillo@40 63 %D 2010-03-21 17:20 UTC
meillo@40 64 %O .CW \s-1http://en.wikipedia.org/w/index.php?title=Unix_philosophy&oldid=351189719
meillo@40 65 .]
meillo@40 66 .QP
meillo@40 67 The \fIUnix philosophy\fP is a set of cultural norms and philosophical
meillo@40 68 approaches to developing software based on the experience of leading
meillo@40 69 developers of the Unix operating system.
meillo@1 70 .PP
meillo@40 71 As there is no single definition of the Unix Philosophy,
meillo@40 72 several people have stated their view on what it comprises.
meillo@1 73 Best known are:
meillo@1 74 .IP \(bu
meillo@1 75 Doug McIlroy's summary: ``Write programs that do one thing and do it well.''
meillo@1 76 .[
meillo@1 77 %A M. D. McIlroy
meillo@1 78 %A E. N. Pinson
meillo@1 79 %A B. A. Taque
meillo@1 80 %T UNIX Time-Sharing System Forward
meillo@1 81 %J The Bell System Technical Journal
meillo@1 82 %D 1978
meillo@1 83 %V 57
meillo@1 84 %N 6
meillo@1 85 %P 1902
meillo@1 86 .]
meillo@1 87 .IP \(bu
meillo@1 88 Mike Gancarz' book ``The UNIX Philosophy''.
meillo@1 89 .[
meillo@1 90 %A Mike Gancarz
meillo@1 91 %T The UNIX Philosophy
meillo@1 92 %D 1995
meillo@1 93 %I Digital Press
meillo@1 94 .]
meillo@1 95 .IP \(bu
meillo@1 96 Eric S. Raymond's book ``The Art of UNIX Programming''.
meillo@1 97 .[
meillo@1 98 %A Eric S. Raymond
meillo@1 99 %T The Art of UNIX Programming
meillo@1 100 %D 2003
meillo@1 101 %I Addison-Wesley
meillo@2 102 %O .CW \s-1http://www.faqs.org/docs/artu/
meillo@1 103 .]
meillo@0 104 .LP
meillo@1 105 These different views on the Unix Philosophy have much in common.
meillo@40 106 Especially, the main concepts are similar in all of them.
meillo@40 107 McIlroy's definition can surely be called the core of the Unix Philosophy,
meillo@40 108 but the fundamental idea behind it all, is ``small is beautiful''.
meillo@40 109
meillo@40 110 .PP
meillo@40 111 The Unix Philosophy tells how to design and write good software for Unix.
meillo@40 112 Many concepts described here base on facilities of Unix.
meillo@40 113 Other operating systems may not offer such facilities,
meillo@40 114 hence it may not be possible to design software is the way of the Unix Philosophy.
meillo@40 115
meillo@40 116 FIXME
meillo@40 117
meillo@40 118 This paper discusses the recommendations of the Unix Philosophy about software design.
meillo@40 119 .PP
meillo@40 120 The here discussed ideas can get applied by any development process.
meillo@40 121 The Unix Philosophy does recommend how the software development process should look like,
meillo@40 122 but this shall not be of matter here.
meillo@40 123 Similar, the question of how to write the code is out of focus.
meillo@1 124 .PP
meillo@1 125 Before we will have a look at concrete concepts,
meillo@1 126 we discuss why software design is important
meillo@1 127 and what problems bad design introduces.
meillo@0 128
meillo@0 129
meillo@0 130 .NH 1
meillo@6 131 Importance of software design in general
meillo@0 132 .LP
meillo@40 133 Software design is the planning of how the internal structure
meillo@40 134 and external interfaces of a software should look like.
meillo@39 135 It has nothing to do with visual appearance.
meillo@39 136 If we take a program as a car, then its color is of no matter.
meillo@39 137 Its design would be the car's size, its shape, the locations of doors,
meillo@39 138 the passenger/space ratio, the luggage capacity, and so forth.
meillo@39 139 .PP
meillo@39 140 Why should software get designed at all?
meillo@6 141 It is general knowledge, that even a bad plan is better than no plan.
meillo@39 142 Not designing software means programming without plan.
meillo@39 143 This will pretty sure lead to horrible results.
meillo@39 144 Horrible to use and horrible to maintain.
meillo@39 145 These two aspects are the visible ones.
meillo@39 146 Often invisible are the wasted possible gains.
meillo@39 147 Good software design can make these gains available.
meillo@2 148 .PP
meillo@39 149 A software's design deals with quality properties.
meillo@39 150 Good design leads to good quality, and quality is important.
meillo@39 151 Any car may be able to drive from A to B,
meillo@39 152 but it depends on the car's properties whether it is a good choice
meillo@39 153 for passenger transport or not.
meillo@39 154 It depends on its properties if it is a good choice
meillo@39 155 for a rough mountain area.
meillo@39 156 And it depends on its properties if the ride will be fun.
meillo@39 157
meillo@2 158 .PP
meillo@39 159 Requirements for a software are twofold:
meillo@39 160 functional and non-functional.
meillo@39 161 .IP \(bu
meillo@39 162 Functional requirements define directly the software's functions.
meillo@39 163 They are the reason why software gets written.
meillo@39 164 Someone has a problem and needs a tool to solve it.
meillo@39 165 Being able to solve the problem is the main functional goal.
meillo@39 166 It is the driving force behind all programming effort.
meillo@39 167 Functional requirements are easier to define and to verify.
meillo@39 168 .IP \(bu
meillo@39 169 Non-functional requirements are also called \fIquality\fP requirements.
meillo@39 170 The quality of a software are the properties that are not directly related to
meillo@39 171 the software's basic functions.
meillo@39 172 Tools of bad quality often solve the problems they were written for,
meillo@39 173 but introduce problems and difficulties for usage and development, later on.
meillo@39 174 Quality aspects are often overlooked at first sight,
meillo@39 175 and they are often difficult to define clearly and to verify.
meillo@2 176 .PP
meillo@39 177 Quality is of few matter when the software gets built initially,
meillo@39 178 but it is of matter for usage and maintenance of the software.
meillo@6 179 A short-sighted might see in developing a software mainly building something up.
meillo@39 180 But experience shows, that building the software the first time is
meillo@39 181 only a small amount of the overall work.
meillo@39 182 Bug fixing, extending, rebuilding of parts
meillo@39 183 \(en maintenance work, for short \(en
meillo@6 184 does soon take over the major part of the time spent on a software.
meillo@6 185 Not to forget the usage of the software.
meillo@6 186 These processes are highly influenced by the software's quality.
meillo@39 187 Thus, quality must not be neglected.
meillo@39 188 The problem with quality is that you hardly ``stumble over''
meillo@39 189 bad quality during the first build,
meillo@6 190 but this is the time when you should care about good quality most.
meillo@6 191 .PP
meillo@39 192 Software design is less the basic function of a software \(en
meillo@39 193 this requirement will get satisfied anyway.
meillo@39 194 Software design is more about quality aspects of the software.
meillo@39 195 Good design leads to good quality, bad design to bad quality.
meillo@6 196 The primary functions of the software will be affected modestly by bad quality,
meillo@39 197 but good quality can provide a lot of additional gain,
meillo@6 198 even at places where one never expected it.
meillo@6 199 .PP
meillo@6 200 The ISO/IEC 9126-1 standard, part 1,
meillo@6 201 .[
meillo@9 202 %I International Organization for Standardization
meillo@6 203 %T ISO Standard 9126: Software Engineering \(en Product Quality, part 1
meillo@6 204 %C Geneve
meillo@6 205 %D 2001
meillo@6 206 .]
meillo@6 207 defines the quality model as consisting out of:
meillo@6 208 .IP \(bu
meillo@6 209 .I Functionality
meillo@6 210 (suitability, accuracy, inter\%operability, security)
meillo@6 211 .IP \(bu
meillo@6 212 .I Reliability
meillo@6 213 (maturity, fault tolerance, recoverability)
meillo@6 214 .IP \(bu
meillo@6 215 .I Usability
meillo@6 216 (understandability, learnability, operability, attractiveness)
meillo@6 217 .IP \(bu
meillo@6 218 .I Efficiency
meillo@9 219 (time behavior, resource utilization)
meillo@6 220 .IP \(bu
meillo@6 221 .I Maintainability
meillo@23 222 (analyzability, changeability, stability, testability)
meillo@6 223 .IP \(bu
meillo@6 224 .I Portability
meillo@6 225 (adaptability, installability, co-existence, replaceability)
meillo@6 226 .LP
meillo@39 227 Good design can improve these properties of a software,
meillo@39 228 bad designed software probably suffers from not having them.
meillo@7 229 .PP
meillo@7 230 One further goal of software design is consistency.
meillo@7 231 Consistency eases understanding, working on, and using things.
meillo@39 232 Consistent internal structure and consistent interfaces to the outside
meillo@39 233 can be provided by good design.
meillo@7 234 .PP
meillo@39 235 Software should be well designed because good design avoids many
meillo@39 236 problems during the software's lifetime.
meillo@39 237 And software should be well designed because good design can offer
meillo@39 238 much additional gain.
meillo@39 239 Indeed, much effort should be spent into good design to make software more valuable.
meillo@39 240 The Unix Philosophy shows a way of how to design software well.
meillo@7 241 It offers guidelines to achieve good quality and high gain for the effort spent.
meillo@0 242
meillo@0 243
meillo@0 244 .NH 1
meillo@0 245 The Unix Philosophy
meillo@4 246 .LP
meillo@4 247 The origins of the Unix Philosophy were already introduced.
meillo@8 248 This chapter explains the philosophy, oriented on Gancarz,
meillo@8 249 and shows concrete examples of its application.
meillo@5 250
meillo@16 251 .NH 2
meillo@14 252 Pipes
meillo@4 253 .LP
meillo@4 254 Following are some examples to demonstrate how applied Unix Philosophy feels like.
meillo@4 255 Knowledge of using the Unix shell is assumed.
meillo@4 256 .PP
meillo@4 257 Counting the number of files in the current directory:
meillo@9 258 .DS I 2n
meillo@4 259 .CW
meillo@9 260 .ps -1
meillo@4 261 ls | wc -l
meillo@4 262 .DE
meillo@4 263 The
meillo@4 264 .CW ls
meillo@4 265 command lists all files in the current directory, one per line,
meillo@4 266 and
meillo@4 267 .CW "wc -l
meillo@8 268 counts the number of lines.
meillo@4 269 .PP
meillo@8 270 Counting the number of files that do not contain ``foo'' in their name:
meillo@9 271 .DS I 2n
meillo@4 272 .CW
meillo@9 273 .ps -1
meillo@4 274 ls | grep -v foo | wc -l
meillo@4 275 .DE
meillo@4 276 Here, the list of files is filtered by
meillo@4 277 .CW grep
meillo@4 278 to remove all that contain ``foo''.
meillo@4 279 The rest is the same as in the previous example.
meillo@4 280 .PP
meillo@4 281 Finding the five largest entries in the current directory.
meillo@9 282 .DS I 2n
meillo@4 283 .CW
meillo@9 284 .ps -1
meillo@4 285 du -s * | sort -nr | sed 5q
meillo@4 286 .DE
meillo@4 287 .CW "du -s *
meillo@4 288 returns the recursively summed sizes of all files
meillo@8 289 \(en no matter if they are regular files or directories.
meillo@4 290 .CW "sort -nr
meillo@4 291 sorts the list numerically in reverse order.
meillo@4 292 Finally,
meillo@4 293 .CW "sed 5q
meillo@4 294 quits after it has printed the fifth line.
meillo@4 295 .PP
meillo@4 296 The presented command lines are examples of what Unix people would use
meillo@4 297 to get the desired output.
meillo@4 298 There are also other ways to get the same output.
meillo@4 299 It's a user's decision which way to go.
meillo@14 300 .PP
meillo@8 301 The examples show that many tasks on a Unix system
meillo@4 302 are accomplished by combining several small programs.
meillo@4 303 The connection between the single programs is denoted by the pipe operator `|'.
meillo@4 304 .PP
meillo@4 305 Pipes, and their extensive and easy use, are one of the great
meillo@4 306 achievements of the Unix system.
meillo@4 307 Pipes between programs have been possible in earlier operating systems,
meillo@4 308 but it has never been a so central part of the concept.
meillo@4 309 When, in the early seventies, Doug McIlroy introduced pipes for the
meillo@4 310 Unix system,
meillo@4 311 ``it was this concept and notation for linking several programs together
meillo@4 312 that transformed Unix from a basic file-sharing system to an entirely new way of computing.''
meillo@4 313 .[
meillo@4 314 %T Unix: An Oral History
meillo@5 315 %O .CW \s-1http://www.princeton.edu/~hos/frs122/unixhist/finalhis.htm
meillo@4 316 .]
meillo@4 317 .PP
meillo@4 318 Being able to specify pipelines in an easy way is,
meillo@4 319 however, not enough by itself.
meillo@5 320 It is only one half.
meillo@4 321 The other is the design of the programs that are used in the pipeline.
meillo@8 322 They have to interfaces that allows them to be used in such a way.
meillo@5 323
meillo@16 324 .NH 2
meillo@14 325 Interface design
meillo@5 326 .LP
meillo@11 327 Unix is, first of all, simple \(en Everything is a file.
meillo@5 328 Files are sequences of bytes, without any special structure.
meillo@5 329 Programs should be filters, which read a stream of bytes from ``standard input'' (stdin)
meillo@5 330 and write a stream of bytes to ``standard output'' (stdout).
meillo@5 331 .PP
meillo@8 332 If the files \fIare\fP sequences of bytes,
meillo@8 333 and the programs \fIare\fP filters on byte streams,
meillo@11 334 then there is exactly one standardized data interface.
meillo@5 335 Thus it is possible to combine them in any desired way.
meillo@5 336 .PP
meillo@5 337 Even a handful of small programs will yield a large set of combinations,
meillo@5 338 and thus a large set of different functions.
meillo@5 339 This is leverage!
meillo@5 340 If the programs are orthogonal to each other \(en the best case \(en
meillo@5 341 then the set of different functions is greatest.
meillo@5 342 .PP
meillo@11 343 Programs might also have a separate control interface,
meillo@11 344 besides their data interface.
meillo@11 345 The control interface is often called ``user interface'',
meillo@11 346 because it is usually designed to be used by humans.
meillo@11 347 The Unix Philosophy discourages to assume the user to be human.
meillo@11 348 Interactive use of software is slow use of software,
meillo@11 349 because the program waits for user input most of the time.
meillo@11 350 Interactive software requires the user to be in front of the computer
meillo@11 351 all the time.
meillo@11 352 Interactive software occupy the user's attention while they are running.
meillo@11 353 .PP
meillo@11 354 Now we come back to the idea of using several small programs, combined,
meillo@11 355 to have a more specific function.
meillo@11 356 If these single tools would all be interactive,
meillo@11 357 how would the user control them?
meillo@11 358 It is not only a problem to control several programs at once if they run at the same time,
meillo@11 359 it also very inefficient to have to control each of the single programs
meillo@11 360 that are intended to work as one large program.
meillo@11 361 Hence, the Unix Philosophy discourages programs to demand interactive use.
meillo@11 362 The behavior of programs should be defined at invocation.
meillo@11 363 This is done by specifying arguments (``command line switches'') to the program call.
meillo@11 364 Gancarz discusses this topic as ``avoid captive user interfaces''.
meillo@11 365 .[
meillo@11 366 %A Mike Gancarz
meillo@11 367 %T The UNIX Philosophy
meillo@11 368 %I Digital Press
meillo@11 369 %D 1995
meillo@11 370 %P 88 ff.
meillo@11 371 .]
meillo@11 372 .PP
meillo@11 373 Non-interactive use is, during development, also an advantage for testing.
meillo@11 374 Testing of interactive programs is much more complicated,
meillo@11 375 than testing of non-interactive programs.
meillo@5 376
meillo@16 377 .NH 2
meillo@8 378 The toolchest approach
meillo@5 379 .LP
meillo@5 380 A toolchest is a set of tools.
meillo@5 381 Instead of having one big tool for all tasks, one has many small tools,
meillo@5 382 each for one task.
meillo@5 383 Difficult tasks are solved by combining several of the small, simple tools.
meillo@5 384 .PP
meillo@11 385 The Unix toolchest \fIis\fP a set of small, (mostly) non-interactive programs
meillo@11 386 that are filters on byte streams.
meillo@11 387 They are, to a large extend, unrelated in their function.
meillo@11 388 Hence, the Unix toolchest provides a large set of functions
meillo@11 389 that can be accessed by combining the programs in the desired way.
meillo@11 390 .PP
meillo@11 391 There are also advantages for developing small toolchest programs.
meillo@5 392 It is easier and less error-prone to write small programs.
meillo@5 393 It is also easier and less error-prone to write a large set of small programs,
meillo@5 394 than to write one large program with all the functionality included.
meillo@5 395 If the small programs are combinable, then they offer even a larger set
meillo@5 396 of functions than the single large program.
meillo@5 397 Hence, one gets two advantages out of writing small, combinable programs.
meillo@5 398 .PP
meillo@5 399 There are two drawbacks of the toolchest approach.
meillo@8 400 First, one simple, standardized, unidirectional interface has to be sufficient.
meillo@5 401 If one feels the need for more ``logic'' than a stream of bytes,
meillo@8 402 then a different approach might be of need.
meillo@13 403 But it is also possible, that he just can not imagine a design where
meillo@8 404 a stream of bytes is sufficient.
meillo@8 405 By becoming more familiar with the ``Unix style of thinking'',
meillo@8 406 developers will more often and easier find simple designs where
meillo@8 407 a stream of bytes is a sufficient interface.
meillo@8 408 .PP
meillo@8 409 The second drawback of a toolchest affects the users.
meillo@5 410 A toolchest is often more difficult to use for novices.
meillo@9 411 It is necessary to become familiar with each of the tools,
meillo@5 412 to be able to use the right one in a given situation.
meillo@9 413 Additionally, one needs to combine the tools in a senseful way on its own.
meillo@9 414 This is like a sharp knife \(en it is a powerful tool in the hand of a master,
meillo@5 415 but of no good value in the hand of an unskilled.
meillo@5 416 .PP
meillo@8 417 However, learning single, small tool of the toolchest is easier than
meillo@8 418 learning a complex tool.
meillo@8 419 The user will have a basic understanding of a yet unknown tool,
meillo@8 420 if the several tools of the toolchest have a common style.
meillo@8 421 He will be able to transfer knowledge over one tool to another.
meillo@8 422 .PP
meillo@8 423 Moreover, the second drawback can be removed easily by adding wrappers
meillo@8 424 around the single tools.
meillo@5 425 Novice users do not need to learn several tools if a professional wraps
meillo@8 426 the single commands into a more high-level script.
meillo@5 427 Note that the wrapper script still calls the small tools;
meillo@5 428 the wrapper script is just like a skin around.
meillo@8 429 No complexity is added this way,
meillo@8 430 but new programs can get created out of existing one with very low effort.
meillo@5 431 .PP
meillo@5 432 A wrapper script for finding the five largest entries in the current directory
meillo@5 433 could look like this:
meillo@9 434 .DS I 2n
meillo@5 435 .CW
meillo@9 436 .ps -1
meillo@5 437 #!/bin/sh
meillo@5 438 du -s * | sort -nr | sed 5q
meillo@5 439 .DE
meillo@5 440 The script itself is just a text file that calls the command line
meillo@5 441 a professional user would type in directly.
meillo@8 442 Making the program flexible on the number of entries it prints,
meillo@8 443 is easily possible:
meillo@9 444 .DS I 2n
meillo@8 445 .CW
meillo@9 446 .ps -1
meillo@8 447 #!/bin/sh
meillo@8 448 num=5
meillo@8 449 [ $# -eq 1 ] && num="$1"
meillo@8 450 du -sh * | sort -nr | sed "${num}q"
meillo@8 451 .DE
meillo@8 452 This script acts like the one before, when called without an argument.
meillo@8 453 But one can also specify a numerical argument to define the number of lines to print.
meillo@5 454
meillo@16 455 .NH 2
meillo@8 456 A powerful shell
meillo@8 457 .LP
meillo@10 458 It was already said, that the Unix shell provides the possibility to
meillo@10 459 combine small programs into large ones easily.
meillo@10 460 A powerful shell is a great feature in other ways, too.
meillo@8 461 .PP
meillo@10 462 For instance by including a scripting language.
meillo@10 463 The control statements are build into the shell.
meillo@8 464 The functions, however, are the normal programs, everyone can use on the system.
meillo@10 465 Thus, the programs are known, so learning to program in the shell is easy.
meillo@8 466 Using normal programs as functions in the shell programming language
meillo@10 467 is only possible because they are small and combinable tools in a toolchest style.
meillo@8 468 .PP
meillo@8 469 The Unix shell encourages to write small scripts out of other programs,
meillo@8 470 because it is so easy to do.
meillo@8 471 This is a great step towards automation.
meillo@8 472 It is wonderful if the effort to automate a task equals the effort
meillo@8 473 it takes to do it the second time by hand.
meillo@8 474 If it is so, then the user will be happy to automate everything he does more than once.
meillo@8 475 .PP
meillo@8 476 Small programs that do one job well, standardized interfaces between them,
meillo@8 477 a mechanism to combine parts to larger parts, and an easy way to automate tasks,
meillo@8 478 this will inevitably produce software leverage.
meillo@8 479 Getting multiple times the benefit of an investment is a great offer.
meillo@10 480 .PP
meillo@10 481 The shell also encourages rapid prototyping.
meillo@10 482 Many well known programs started as quickly hacked shell scripts,
meillo@10 483 and turned into ``real'' programs, written in C, later.
meillo@10 484 Building a prototype first is a way to avoid the biggest problems
meillo@10 485 in application development.
meillo@10 486 Fred Brooks writes in ``No Silver Bullet'':
meillo@10 487 .[
meillo@10 488 %A Frederick P. Brooks, Jr.
meillo@10 489 %T No Silver Bullet: Essence and Accidents of Software Engineering
meillo@10 490 %B Information Processing 1986, the Proceedings of the IFIP Tenth World Computing Conference
meillo@10 491 %E H.-J. Kugler
meillo@10 492 %D 1986
meillo@10 493 %P 1069\(en1076
meillo@10 494 %I Elsevier Science B.V.
meillo@10 495 %C Amsterdam, The Netherlands
meillo@10 496 .]
meillo@10 497 .QP
meillo@10 498 The hardest single part of building a software system is deciding precisely what to build.
meillo@10 499 No other part of the conceptual work is so difficult as establishing the detailed
meillo@10 500 technical requirements, [...].
meillo@10 501 No other part of the work so cripples the resulting system if done wrong.
meillo@10 502 No other part is more difficult to rectify later.
meillo@10 503 .PP
meillo@10 504 Writing a prototype is a great method to become familiar with the requirements
meillo@10 505 and to actually run into real problems.
meillo@10 506 Today, prototyping is often seen as a first step in building a software.
meillo@10 507 This is, of course, good.
meillo@10 508 However, the Unix Philosophy has an \fIadditional\fP perspective on prototyping:
meillo@10 509 After having built the prototype, one might notice, that the prototype is already
meillo@10 510 \fIgood enough\fP.
meillo@10 511 Hence, no reimplementation, in a more sophisticated programming language, might be of need,
meillo@10 512 for the moment.
meillo@23 513 Maybe later, it might be necessary to rewrite the software, but not now.
meillo@10 514 .PP
meillo@10 515 By delaying further work, one keeps the flexibility to react easily on
meillo@10 516 changing requirements.
meillo@10 517 Software parts that are not written will not miss the requirements.
meillo@10 518
meillo@16 519 .NH 2
meillo@10 520 Worse is better
meillo@10 521 .LP
meillo@10 522 The Unix Philosophy aims for the 80% solution;
meillo@10 523 others call it the ``Worse is better'' approach.
meillo@10 524 .PP
meillo@10 525 First, practical experience shows, that it is almost never possible to define the
meillo@10 526 requirements completely and correctly the first time.
meillo@10 527 Hence one should not try to; it will fail anyway.
meillo@10 528 Second, practical experience shows, that requirements change during time.
meillo@10 529 Hence it is best to delay requirement-based design decisions as long as possible.
meillo@10 530 Also, the software should be small and flexible as long as possible
meillo@10 531 to react on changing requirements.
meillo@10 532 Shell scripts, for example, are more easily adjusted as C programs.
meillo@10 533 Third, practical experience shows, that maintenance is hard work.
meillo@10 534 Hence, one should keep the amount of software as small as possible;
meillo@10 535 it should just fulfill the \fIcurrent\fP requirements.
meillo@10 536 Software parts that will be written later, do not need maintenance now.
meillo@10 537 .PP
meillo@10 538 Starting with a prototype in a scripting language has several advantages:
meillo@10 539 .IP \(bu
meillo@10 540 As the initial effort is low, one will likely start right away.
meillo@10 541 .IP \(bu
meillo@10 542 As working parts are available soon, the real requirements can get identified soon.
meillo@10 543 .IP \(bu
meillo@10 544 When a software is usable, it gets used, and thus tested.
meillo@10 545 Hence problems will be found at early stages of the development.
meillo@10 546 .IP \(bu
meillo@10 547 The prototype might be enough for the moment,
meillo@10 548 thus further work on the software can be delayed to a time
meillo@10 549 when one knows better about the requirements and problems,
meillo@10 550 than now.
meillo@10 551 .IP \(bu
meillo@10 552 Implementing now only the parts that are actually needed now,
meillo@10 553 requires fewer maintenance work.
meillo@10 554 .IP \(bu
meillo@10 555 If the global situation changes so that the software is not needed anymore,
meillo@10 556 then less effort was spent into the project, than it would have be
meillo@10 557 when a different approach had been used.
meillo@10 558
meillo@16 559 .NH 2
meillo@11 560 Upgrowth and survival of software
meillo@11 561 .LP
meillo@12 562 So far it was talked about \fIwriting\fP or \fIbuilding\fP software.
meillo@13 563 Although these are just verbs, they do imply a specific view on the work process
meillo@13 564 they describe.
meillo@12 565 The better verb, however, is to \fIgrow\fP.
meillo@12 566 .PP
meillo@12 567 Creating software in the sense of the Unix Philosophy is an incremental process.
meillo@12 568 It starts with a first prototype, which evolves as requirements change.
meillo@12 569 A quickly hacked shell script might become a large, sophisticated,
meillo@13 570 compiled program this way.
meillo@13 571 Its lifetime begins with the initial prototype and ends when the software is not used anymore.
meillo@13 572 While being alive it will get extended, rearranged, rebuilt (from scratch).
meillo@12 573 Growing software matches the view that ``software is never finished. It is only released.''
meillo@12 574 .[
meillo@13 575 %O FIXME
meillo@13 576 %A Mike Gancarz
meillo@13 577 %T The UNIX Philosophy
meillo@13 578 %P 26
meillo@12 579 .]
meillo@12 580 .PP
meillo@13 581 Software can be seen as being controlled by evolutionary processes.
meillo@13 582 Successful software is software that is used by many for a long time.
meillo@12 583 This implies that the software is needed, useful, and better than alternatives.
meillo@12 584 Darwin talks about: ``The survival of the fittest.''
meillo@12 585 .[
meillo@13 586 %O FIXME
meillo@13 587 %A Charles Darwin
meillo@12 588 .]
meillo@12 589 Transferred to software: The most successful software, is the fittest,
meillo@12 590 is the one that survives.
meillo@13 591 (This may be at the level of one creature, or at the level of one species.)
meillo@13 592 The fitness of software is affected mainly by four properties:
meillo@15 593 portability of code, portability of data, range of usability, and reusability of parts.
meillo@15 594 .\" .IP \(bu
meillo@15 595 .\" portability of code
meillo@15 596 .\" .IP \(bu
meillo@15 597 .\" portability of data
meillo@15 598 .\" .IP \(bu
meillo@15 599 .\" range of usability
meillo@15 600 .\" .IP \(bu
meillo@15 601 .\" reuseability of parts
meillo@13 602 .PP
meillo@15 603 (1)
meillo@15 604 .I "Portability of code
meillo@15 605 means, using high-level programming languages,
meillo@13 606 sticking to the standard,
meillo@13 607 and avoiding optimizations that introduce dependencies on specific hardware.
meillo@13 608 Hardware has a much lower lifetime than software.
meillo@13 609 By chaining software to a specific hardware,
meillo@13 610 the software's lifetime gets shortened to that of this hardware.
meillo@13 611 In contrast, software should be easy to port \(en
meillo@23 612 adaptation is the key to success.
meillo@13 613 .\" cf. practice of prog: ch08
meillo@13 614 .PP
meillo@15 615 (2)
meillo@15 616 .I "Portability of data
meillo@15 617 is best achieved by avoiding binary representations
meillo@13 618 to store data, because binary representations differ from machine to machine.
meillo@23 619 Textual representation is favored.
meillo@13 620 Historically, ASCII was the charset of choice.
meillo@13 621 In the future, UTF-8 might be the better choice, however.
meillo@13 622 Important is that it is a plain text representation in a
meillo@13 623 very common charset encoding.
meillo@13 624 Apart from being able to transfer data between machines,
meillo@13 625 readable data has the great advantage, that humans are able
meillo@13 626 to directly edit it with text editors and other tools from the Unix toolchest.
meillo@13 627 .\" gancarz tenet 5
meillo@13 628 .PP
meillo@15 629 (3)
meillo@15 630 A large
meillo@15 631 .I "range of usability
meillo@23 632 ensures good adaptation, and thus good survival.
meillo@13 633 It is a special distinction if a software becomes used in fields of action,
meillo@13 634 the original authors did never imagine.
meillo@13 635 Software that solves problems in a general way will likely be used
meillo@13 636 for all kinds of similar problems.
meillo@13 637 Being too specific limits the range of uses.
meillo@13 638 Requirements change through time, thus use cases change or even vanish.
meillo@13 639 A good example in this point is Allman's sendmail.
meillo@13 640 Allman identifies flexibility to be one major reason for sendmail's success:
meillo@13 641 .[
meillo@13 642 %O FIXME
meillo@13 643 %A Allman
meillo@13 644 %T sendmail
meillo@13 645 .]
meillo@13 646 .QP
meillo@13 647 Second, I limited myself to the routing function [...].
meillo@13 648 This was a departure from the dominant thought of the time, [...].
meillo@13 649 .QP
meillo@13 650 Third, the sendmail configuration file was flexible enough to adopt
meillo@13 651 to a rapidly changing world [...].
meillo@12 652 .LP
meillo@13 653 Successful software adopts itself to the changing world.
meillo@13 654 .PP
meillo@15 655 (4)
meillo@15 656 .I "Reuse of parts
meillo@15 657 is even one step further.
meillo@13 658 A software may completely lose its field of action,
meillo@13 659 but parts of which the software is build may be general and independent enough
meillo@13 660 to survive this death.
meillo@13 661 If software is build by combining small independent programs,
meillo@13 662 then there are parts readily available for reuse.
meillo@13 663 Who cares if the large program is a failure,
meillo@13 664 but parts of it become successful instead?
meillo@10 665
meillo@16 666 .NH 2
meillo@14 667 Summary
meillo@0 668 .LP
meillo@14 669 This chapter explained the central ideas of the Unix Philosophy.
meillo@14 670 For each of the ideas, it was exposed what advantages they introduce.
meillo@14 671 The Unix Philosophy are guidelines that help to write valuable software.
meillo@14 672 From the view point of a software developer or software designer,
meillo@14 673 the Unix Philosophy provides answers to many software design problem.
meillo@14 674 .PP
meillo@14 675 The various ideas of the Unix Philosophy are very interweaved
meillo@14 676 and can hardly be applied independently.
meillo@14 677 However, the probably most important messages are:
meillo@14 678 .I "``Do one thing well!''" ,
meillo@14 679 .I "``Keep it simple!''" ,
meillo@14 680 and
meillo@14 681 .I "``Use software leverage!''
meillo@0 682
meillo@8 683
meillo@8 684
meillo@0 685 .NH 1
meillo@19 686 Case study: \s-1MH\s0
meillo@18 687 .LP
meillo@30 688 The previous chapter introduced and explained the Unix Philosophy
meillo@18 689 from a general point of view.
meillo@30 690 The driving force were the guidelines; references to
meillo@18 691 existing software were given only sparsely.
meillo@18 692 In this and the next chapter, concrete software will be
meillo@18 693 the driving force in the discussion.
meillo@18 694 .PP
meillo@23 695 This first case study is about the mail user agents (\s-1MUA\s0)
meillo@23 696 \s-1MH\s0 (``mail handler'') and its descendent \fInmh\fP
meillo@23 697 (``new mail handler'').
meillo@23 698 \s-1MUA\s0s provide functions to read, compose, and organize mail,
meillo@23 699 but (ideally) not to transfer.
meillo@19 700 In this document, the name \s-1MH\s0 will be used for both of them.
meillo@19 701 A distinction will only be made if differences between
meillo@19 702 them are described.
meillo@18 703
meillo@0 704
meillo@0 705 .NH 2
meillo@19 706 Historical background
meillo@0 707 .LP
meillo@19 708 Electronic mail was available in Unix very early.
meillo@30 709 The first \s-1MUA\s0 on Unix was \f(CWmail\fP,
meillo@30 710 which was already present in the First Edition.
meillo@30 711 .[
meillo@30 712 %A Peter H. Salus
meillo@30 713 %T A Quarter Century of UNIX
meillo@30 714 %D 1994
meillo@30 715 %I Addison-Wesley
meillo@30 716 %P 41 f.
meillo@30 717 .]
meillo@30 718 It was a small program that either prints the user's mailbox file
meillo@19 719 or appends text to someone elses mailbox file,
meillo@19 720 depending on the command line arguments.
meillo@19 721 .[
meillo@19 722 %O http://cm.bell-labs.com/cm/cs/who/dmr/pdfs/man12.pdf
meillo@19 723 .]
meillo@19 724 It was a program that did one job well.
meillo@23 725 This job was emailing, which was very simple then.
meillo@19 726 .PP
meillo@23 727 Later, emailing became more powerful, and thus more complex.
meillo@19 728 The simple \f(CWmail\fP, which knew nothing of subjects,
meillo@19 729 independent handling of single messages,
meillo@19 730 and long-time storage of them, was not powerful enough anymore.
meillo@19 731 At Berkeley, Kurt Shoens wrote \fIMail\fP (with capital `M')
meillo@19 732 in 1978 to provide additional functions for emailing.
meillo@19 733 Mail was still one program, but now it was large and did
meillo@19 734 several jobs.
meillo@23 735 Its user interface is modeled after the one of \fIed\fP.
meillo@19 736 It is designed for humans, but is still scriptable.
meillo@23 737 \fImailx\fP is the adaptation of Berkeley Mail into System V.
meillo@19 738 .[
meillo@19 739 %A Gunnar Ritter
meillo@19 740 %O http://heirloom.sourceforge.net/mailx_history.html
meillo@19 741 .]
meillo@30 742 Elm, pine, mutt, and a whole bunch of graphical \s-1MUA\s0s
meillo@19 743 followed Mail's direction.
meillo@19 744 They are large, monolithic programs which include all emailing functions.
meillo@19 745 .PP
meillo@23 746 A different way was taken by the people of \s-1RAND\s0 Corporation.
meillo@38 747 In the beginning, they also had used a monolithic mail system,
meillo@30 748 called \s-1MS\s0 (for ``mail system'').
meillo@19 749 But in 1977, Stockton Gaines and Norman Shapiro
meillo@19 750 came up with a proposal of a new email system concept \(en
meillo@19 751 one that honors the Unix Philosophy.
meillo@19 752 The concept was implemented by Bruce Borden in 1978 and 1979.
meillo@19 753 This was the birth of \s-1MH\s0 \(en the ``mail handler''.
meillo@18 754 .PP
meillo@18 755 Since then, \s-1RAND\s0, the University of California at Irvine and
meillo@19 756 at Berkeley, and several others have contributed to the software.
meillo@18 757 However, it's core concepts remained the same.
meillo@23 758 In the late 90s, when development of \s-1MH\s0 slowed down,
meillo@19 759 Richard Coleman started with \fInmh\fP, the new mail handler.
meillo@19 760 His goal was to improve \s-1MH\s0, especially in regard of
meillo@23 761 the requirements of modern emailing.
meillo@19 762 Today, nmh is developed by various people on the Internet.
meillo@18 763 .[
meillo@18 764 %T RAND and the Information Evolution: A History in Essays and Vignettes
meillo@18 765 %A Willis H. Ware
meillo@18 766 %D 2008
meillo@18 767 %I The RAND Corporation
meillo@18 768 %P 128\(en137
meillo@18 769 %O .CW \s-1http://www.rand.org/pubs/corporate_pubs/CP537/
meillo@18 770 .]
meillo@18 771 .[
meillo@18 772 %T MH & xmh: Email for Users & Programmers
meillo@18 773 %A Jerry Peek
meillo@18 774 %D 1991, 1992, 1995
meillo@18 775 %I O'Reilly & Associates, Inc.
meillo@18 776 %P Appendix B
meillo@18 777 %O Also available online: \f(CW\s-2http://rand-mh.sourceforge.net/book/\fP
meillo@18 778 .]
meillo@0 779
meillo@0 780 .NH 2
meillo@20 781 Contrasts to monolithic mail systems
meillo@0 782 .LP
meillo@19 783 All \s-1MUA\s0s are monolithic, except \s-1MH\s0.
meillo@38 784 Although there might actually exist further, very little known,
meillo@30 785 toolchest \s-1MUA\s0s, this statement reflects the situation pretty well.
meillo@19 786 .PP
meillo@30 787 Monolithic \s-1MUA\s0s gather all their functions in one program.
meillo@30 788 In contrast, \s-1MH\s0 is a toolchest of many small tools \(en one for each job.
meillo@23 789 Following is a list of important programs of \s-1MH\s0's toolchest
meillo@30 790 and their function.
meillo@30 791 It gives a feeling of how the toolchest looks like.
meillo@19 792 .IP \(bu
meillo@19 793 .CW inc :
meillo@30 794 incorporate new mail (this is how mail enters the system)
meillo@19 795 .IP \(bu
meillo@19 796 .CW scan :
meillo@19 797 list messages in folder
meillo@19 798 .IP \(bu
meillo@19 799 .CW show :
meillo@19 800 show message
meillo@19 801 .IP \(bu
meillo@19 802 .CW next\fR/\fPprev :
meillo@19 803 show next/previous message
meillo@19 804 .IP \(bu
meillo@19 805 .CW folder :
meillo@19 806 change current folder
meillo@19 807 .IP \(bu
meillo@19 808 .CW refile :
meillo@19 809 refile message into folder
meillo@19 810 .IP \(bu
meillo@19 811 .CW rmm :
meillo@19 812 remove message
meillo@19 813 .IP \(bu
meillo@19 814 .CW comp :
meillo@19 815 compose a new message
meillo@19 816 .IP \(bu
meillo@19 817 .CW repl :
meillo@19 818 reply to a message
meillo@19 819 .IP \(bu
meillo@19 820 .CW forw :
meillo@19 821 forward a message
meillo@19 822 .IP \(bu
meillo@19 823 .CW send :
meillo@30 824 send a prepared message (this is how mail leaves the system)
meillo@0 825 .LP
meillo@19 826 \s-1MH\s0 has no special user interface like monolithic \s-1MUA\s0s have.
meillo@19 827 The user does not leave the shell to run \s-1MH\s0,
meillo@30 828 but he uses the various \s-1MH\s0 programs within the shell.
meillo@23 829 Using a monolithic program with a captive user interface
meillo@23 830 means ``entering'' the program, using it, and ``exiting'' the program.
meillo@23 831 Using toolchests like \s-1MH\s0 means running programs,
meillo@38 832 alone or in combination with others, even from other toolchests,
meillo@23 833 without leaving the shell.
meillo@30 834
meillo@30 835 .NH 2
meillo@30 836 Data storage
meillo@30 837 .LP
meillo@34 838 \s-1MH\s0's mail storage is a directory tree under the user's
meillo@34 839 \s-1MH\s0 directory (usually \f(CW$HOME/Mail\fP),
meillo@34 840 where mail folders are directories and mail messages are text files
meillo@34 841 within them.
meillo@34 842 Each mail folder contains a file \f(CW.mh_sequences\fP which lists
meillo@34 843 the public message sequences of that folder, for instance new messages.
meillo@34 844 Mail messages are text files located in a mail folder.
meillo@34 845 The files contain the messages as they were received.
meillo@34 846 They are numbered in ascending order in each folder.
meillo@19 847 .PP
meillo@30 848 This mailbox format is called ``\s-1MH\s0'' after the \s-1MUA\s0.
meillo@30 849 Alternatives are \fImbox\fP and \fImaildir\fP.
meillo@30 850 In the mbox format all messages are stored within one file.
meillo@30 851 This was a good solution in the early days, when messages
meillo@30 852 were only a few lines of text and were deleted soon.
meillo@30 853 Today, when single messages often include several megabytes
meillo@30 854 of attachments, it is a bad solution.
meillo@30 855 Another disadvantage of the mbox format is that it is
meillo@30 856 more difficult to write tools that work on mail messages,
meillo@30 857 because it is always necessary to first find and extract
meillo@30 858 the relevant message in the mbox file.
meillo@30 859 With the \s-1MH\s0 mailbox format,
meillo@30 860 each message is a self-standing item, by definition.
meillo@30 861 Also, the problem of concurrent access to one mailbox is
meillo@30 862 reduced to the problem of concurrent access to one message.
meillo@30 863 Maildir is generally similar to \s-1MH\s0's format,
meillo@30 864 but modified towards guaranteed reliability.
meillo@30 865 This involves some complexity, unfortunately.
meillo@34 866 .PP
meillo@34 867 Working with \s-1MH\s0's toolchest on mailboxes is much like
meillo@34 868 working with Unix' toolchest on directory trees:
meillo@34 869 \f(CWscan\fP is like \f(CWls\fP,
meillo@34 870 \f(CWshow\fP is like \f(CWcat\fP,
meillo@34 871 \f(CWfolder\fP is like \f(CWcd\fP and \f(CWpwd\fP,
meillo@34 872 \f(CWrefile\fP is like \f(CWmv\fP,
meillo@34 873 and \f(CWrmm\fP is like \f(CWrm\fP.
meillo@34 874 .PP
meillo@34 875 The context of tools in Unix consists mainly the current working directory,
meillo@34 876 the user identification, and the environment variables.
meillo@34 877 \s-1MH\s0 extends this context by two more items:
meillo@34 878 .IP \(bu
meillo@34 879 The current mail folder, which is similar to the current working directory.
meillo@34 880 For mail folders, \f(CWfolder\fP provides the corresponding functionality
meillo@34 881 of \f(CWcd\fP and \f(CWpwd\fP for directories.
meillo@34 882 .IP \(bu
meillo@34 883 Sequences, which are named sets of messages in a mail folder.
meillo@34 884 The current message, relative to a mail folder, is a special sequence.
meillo@34 885 It enables commands like \f(CWnext\fP and \f(CWprev\fP.
meillo@34 886 .LP
meillo@34 887 In contrast to Unix' context, which is chained to the shell session,
meillo@34 888 \s-1MH\s0's context is independent.
meillo@34 889 Usually there is one context for each user, but a user can have many
meillo@34 890 contexts.
meillo@34 891 Public sequences are an exception, as they belong to the mail folder.
meillo@34 892 .[
meillo@34 893 %O mh-profile(5) and mh-sequence(5)
meillo@34 894 .]
meillo@20 895
meillo@0 896 .NH 2
meillo@20 897 Discussion of the design
meillo@0 898 .LP
meillo@20 899 The following paragraphs discuss \s-1MH\s0 in regard to the tenets
meillo@23 900 of the Unix Philosophy which Gancarz identified.
meillo@20 901
meillo@20 902 .PP
meillo@33 903 .B "Small is beautiful
meillo@20 904 and
meillo@33 905 .B "do one thing well
meillo@20 906 are two design goals that are directly visible in \s-1MH\s0.
meillo@20 907 Gancarz actually presents \s-1MH\s0 as example under the headline
meillo@20 908 ``Making UNIX Do One Thing Well'':
meillo@20 909 .QP
meillo@20 910 [\s-1MH\s0] consists of a series of programs which
meillo@20 911 when combined give the user an enormous ability
meillo@20 912 to manipulate electronic mail messages.
meillo@20 913 A complex application, it shows that not only is it
meillo@20 914 possible to build large applications from smaller
meillo@20 915 components, but also that such designs are actually preferable.
meillo@20 916 .[
meillo@20 917 %A Mike Gancarz
meillo@20 918 %T unix-phil
meillo@20 919 %P 125
meillo@20 920 .]
meillo@20 921 .LP
meillo@20 922 The various small programs of \s-1MH\s0 were relatively easy
meillo@23 923 to write, because each of them is small, limited to one function,
meillo@23 924 and has clear boundaries.
meillo@20 925 For the same reasons, they are also good to maintain.
meillo@20 926 Further more, the system can easily get extended.
meillo@20 927 One only needs to put a new program into the toolchest.
meillo@23 928 This was done, for instance, when \s-1MIME\s0 support was added
meillo@20 929 (e.g. \f(CWmhbuild\fP).
meillo@20 930 Also, different programs can exist to do the basically same job
meillo@20 931 in different ways (e.g. in nmh: \f(CWshow\fP and \f(CWmhshow\fP).
meillo@20 932 If someone needs a mail system with some additionally
meillo@23 933 functions that are available nowhere yet, he best takes a
meillo@20 934 toolchest system like \s-1MH\s0 where he can add the
meillo@20 935 functionality with little work.
meillo@20 936
meillo@20 937 .PP
meillo@34 938 .B "Store data in flat text files
meillo@34 939 is followed by \s-1MH\s0.
meillo@34 940 This is not surprising, because email messages are already plain text.
meillo@34 941 \s-1MH\s0 stores the messages as it receives them,
meillo@34 942 thus any other tool that works on RFC 2822 mail messages can operate
meillo@34 943 on the messages in an \s-1MH\s0 mailbox.
meillo@34 944 All other files \s-1MH\s0 uses are plain text too.
meillo@34 945 It is therefore possible and encouraged to use the text processing
meillo@34 946 tools of Unix' toolchest to extend \s-1MH\s0's toolchest.
meillo@20 947
meillo@20 948 .PP
meillo@33 949 .B "Avoid captive user interfaces" .
meillo@19 950 \s-1MH\s0 is perfectly suited for non-interactive use.
meillo@19 951 It offers all functions directly and without captive user interfaces.
meillo@30 952 If, nonetheless, users want a graphical user interface,
meillo@20 953 they can have it with \fIxmh\fP or \fIexmh\fP, too.
meillo@19 954 These are graphical frontends for the \s-1MH\s0 toolchest.
meillo@19 955 This means, all email-related work is still done by \s-1MH\s0 tools,
meillo@20 956 but the frontend issues the appropriate calls when the user
meillo@30 957 clicks on buttons.
meillo@20 958 Providing easy-to-use user interfaces in form of frontends is a good
meillo@19 959 approach, because it does not limit the power of the backend itself.
meillo@20 960 The frontend will anyway only be able to make a subset of the
meillo@23 961 backend's power and flexibility available to the user.
meillo@20 962 But if it is a separate program,
meillo@20 963 then the missing parts can still be accessed at the backend directly.
meillo@19 964 If it is integrated, then this will hardly be possible.
meillo@30 965 Further more, it is possible to have different frontends to the same
meillo@30 966 backend.
meillo@19 967
meillo@19 968 .PP
meillo@33 969 .B "Choose portability over efficiency
meillo@20 970 and
meillo@33 971 .B "use shell scripts to increase leverage and portability" .
meillo@20 972 These two tenets are indirectly, but nicely, demonstrated by
meillo@30 973 Bolsky and Korn in their book about the Korn Shell.
meillo@20 974 .[
meillo@20 975 %T The KornShell: command and programming language
meillo@20 976 %A Morris I. Bolsky
meillo@20 977 %A David G. Korn
meillo@20 978 %I Prentice Hall
meillo@20 979 %D 1989
meillo@30 980 %P 254\(en290
meillo@20 981 %O \s-1ISBN\s0: 0-13-516972-0
meillo@20 982 .]
meillo@30 983 They demonstrated, in chapter 18 of the book, a basic implementation
meillo@20 984 of a subset of \s-1MH\s0 in ksh scripts.
meillo@20 985 Of course, this was just a demonstration, but a brilliant one.
meillo@20 986 It shows how quickly one can implement such a prototype with shell scripts,
meillo@20 987 and how readable they are.
meillo@20 988 The implementation in the scripting language may not be very fast,
meillo@20 989 but it can be fast enough though, and this is all that matters.
meillo@20 990 By having the code in an interpreted language, like the shell,
meillo@20 991 portability becomes a minor issue, if we assume the interpreter
meillo@20 992 to be widespread.
meillo@20 993 This demonstration also shows how easy it is to create single programs
meillo@20 994 of a toolchest software.
meillo@30 995 There are eight tools (two of them have multiple names) and 16 functions
meillo@30 996 with supporting code.
meillo@30 997 Each tool comprises between 12 and 38 lines of ksh,
meillo@30 998 in total about 200 lines.
meillo@30 999 The functions comprise between 3 and 78 lines of ksh,
meillo@30 1000 in total about 450 lines.
meillo@20 1001 Such small software is easy to write, easy to understand,
meillo@20 1002 and thus easy to maintain.
meillo@23 1003 A toolchest improves the possibility to only write some parts
meillo@20 1004 and though create a working result.
meillo@20 1005 Expanding the toolchest without global changes will likely be
meillo@20 1006 possible, too.
meillo@20 1007
meillo@20 1008 .PP
meillo@33 1009 .B "Use software leverage to your advantage
meillo@20 1010 and the lesser tenet
meillo@33 1011 .B "allow the user to tailor the environment
meillo@20 1012 are ideally followed in the design of \s-1MH\s0.
meillo@21 1013 Tailoring the environment is heavily encouraged by the ability to
meillo@30 1014 directly define default options to programs.
meillo@30 1015 It is even possible to define different default options
meillo@21 1016 depending on the name under which the program was called.
meillo@21 1017 Software leverage is heavily encouraged by the ease it is to
meillo@21 1018 create shell scripts that run a specific command line,
meillo@30 1019 built of several \s-1MH\s0 programs.
meillo@21 1020 There is few software that so much wants users to tailor their
meillo@21 1021 environment and to leverage the use of the software, like \s-1MH\s0.
meillo@21 1022 Just to make one example:
meillo@23 1023 One might prefer a different listing format for the \f(CWscan\fP
meillo@21 1024 program.
meillo@30 1025 It is possible to take one of the distributed format files
meillo@21 1026 or to write one yourself.
meillo@21 1027 To use the format as default for \f(CWscan\fP, a single line,
meillo@21 1028 reading
meillo@21 1029 .DS
meillo@21 1030 .CW
meillo@21 1031 scan: -form FORMATFILE
meillo@21 1032 .DE
meillo@21 1033 must be added to \f(CW.mh_profile\fP.
meillo@21 1034 If one wants this different format as an additional command,
meillo@23 1035 instead of changing the default, he needs to create a link to
meillo@23 1036 \f(CWscan\fP, for instance titled \f(CWscan2\fP.
meillo@21 1037 The line in \f(CW.mh_profile\fP would then start with \f(CWscan2\fP,
meillo@30 1038 as the option should only be in effect when scan is called as
meillo@21 1039 \f(CWscan2\fP.
meillo@20 1040
meillo@20 1041 .PP
meillo@33 1042 .B "Make every program a filter
meillo@21 1043 is hard to find in \s-1MH\s0.
meillo@21 1044 The reason therefore is that most of \s-1MH\s0's tools provide
meillo@21 1045 basic file system operations for the mailboxes.
meillo@30 1046 The reason is the same because of which
meillo@21 1047 \f(CWls\fP, \f(CWcp\fP, \f(CWmv\fP, and \f(CWrm\fP
meillo@21 1048 aren't filters neither.
meillo@23 1049 However, they build a basis on which filters can operate.
meillo@23 1050 \s-1MH\s0 does not provide many filters itself, but it is a basis
meillo@23 1051 to write filters for.
meillo@30 1052 An example would be a mail message text highlighter,
meillo@30 1053 that means a program that makes use of a color terminal to display
meillo@30 1054 header lines, quotations, and signatures in distinct colors.
meillo@30 1055 The author's version of this program, for instance,
meillo@30 1056 is a 25 line awk script.
meillo@21 1057
meillo@21 1058 .PP
meillo@33 1059 .B "Build a prototype as soon as possible
meillo@21 1060 was again well followed by \s-1MH\s0.
meillo@21 1061 This tenet, of course, focuses on early development, which is
meillo@21 1062 long time ago for \s-1MH\s0.
meillo@21 1063 But without following this guideline at the very beginning,
meillo@23 1064 Bruce Borden may have not convinced the management of \s-1RAND\s0
meillo@23 1065 to ever create \s-1MH\s0.
meillo@23 1066 In Bruce' own words:
meillo@21 1067 .QP
meillo@30 1068 [...] but they [Stockton Gaines and Norm Shapiro] were not able
meillo@23 1069 to convince anyone that such a system would be fast enough to be usable.
meillo@21 1070 I proposed a very short project to prove the basic concepts,
meillo@21 1071 and my management agreed.
meillo@21 1072 Looking back, I realize that I had been very lucky with my first design.
meillo@21 1073 Without nearly enough design work,
meillo@21 1074 I built a working environment and some header files
meillo@21 1075 with key structures and wrote the first few \s-1MH\s0 commands:
meillo@21 1076 inc, show/next/prev, and comp.
meillo@21 1077 [...]
meillo@21 1078 With these three, I was able to convince people that the structure was viable.
meillo@21 1079 This took about three weeks.
meillo@21 1080 .[
meillo@21 1081 %O FIXME
meillo@21 1082 .]
meillo@0 1083
meillo@0 1084 .NH 2
meillo@0 1085 Problems
meillo@0 1086 .LP
meillo@22 1087 \s-1MH\s0, for sure is not without problems.
meillo@30 1088 There are two main problems: one is technical, the other is about human behavior.
meillo@22 1089 .PP
meillo@22 1090 \s-1MH\s0 is old and email today is very different to email in the time
meillo@22 1091 when \s-1MH\s0 was designed.
meillo@22 1092 \s-1MH\s0 adopted to the changes pretty well, but it is limited.
meillo@22 1093 For example in development resources.
meillo@22 1094 \s-1MIME\s0 support and support for different character encodings
meillo@22 1095 is available, but only on a moderate level.
meillo@22 1096 More active developers could quickly improve there.
meillo@22 1097 It is also limited by design, which is the larger problem.
meillo@22 1098 \s-1IMAP\s0, for example, conflicts with \s-1MH\s0's design to a large extend.
meillo@22 1099 These design conflicts are not easily solvable.
meillo@22 1100 Possibly, they require a redesign.
meillo@30 1101 Maybe \s-1IMAP\s0 is too different to the classic mail model which \s-1MH\s0 covers,
meillo@30 1102 hence \s-1MH\s0 may never work well with \s-1IMAP\s0.
meillo@22 1103 .PP
meillo@22 1104 The other kind of problem is human habits.
meillo@22 1105 When in this world almost all \s-1MUA\s0s are monolithic,
meillo@22 1106 it is very difficult to convince people to use a toolbox style \s-1MUA\s0
meillo@22 1107 like \s-1MH\s0.
meillo@22 1108 The habits are so strong, that even people who understood the concept
meillo@30 1109 and advantages of \s-1MH\s0 do not like to switch,
meillo@30 1110 simply because \s-1MH\s0 is different.
meillo@30 1111 Unfortunately, the frontends to \s-1MH\s0, which could provide familiar look'n'feel,
meillo@30 1112 are quite outdated and thus not very appealing compared to the modern interfaces
meillo@30 1113 which monolithic \s-1MUA\s0s offer.
meillo@20 1114
meillo@20 1115 .NH 2
meillo@20 1116 Summary \s-1MH\s0
meillo@20 1117 .LP
meillo@31 1118 \s-1MH\s0 is an \s-1MUA\s0 that follows the Unix Philosophy in its design
meillo@31 1119 and implementation.
meillo@31 1120 It consists of a toolchest of small tools, each of them does one job well.
meillo@31 1121 The tools are orthogonal to each other, to a large extend.
meillo@31 1122 However, for historical reasons, there also exist distinct tools
meillo@31 1123 that cover the same task.
meillo@31 1124 .PP
meillo@31 1125 The toolchest approach offers great flexibility to the user.
meillo@31 1126 He can use the complete power of the Unix shell with \s-1MH\s0.
meillo@31 1127 This makes \s-1MH\s0 a very powerful mail system.
meillo@31 1128 Extending and customizing \s-1MH\s0 is easy and encouraged, too.
meillo@31 1129 .PP
meillo@31 1130 Apart from the user's perspective, \s-1MH\s0 is development-friendly.
meillo@31 1131 Its overall design follows clear rules.
meillo@31 1132 The single tools do only one job, thus they are easy to understand,
meillo@31 1133 easy to write, and good to maintain.
meillo@31 1134 They are all independent and do not interfere with the others.
meillo@31 1135 Automated testing of their function is a straight forward task.
meillo@31 1136 .PP
meillo@31 1137 It is sad, that \s-1MH\s0's differentness is its largest problem,
meillo@31 1138 as its differentness is also its largest advantage.
meillo@31 1139 Unfortunately, for most people their habits are stronger
meillo@31 1140 than the attraction of the clear design and the power, \s-1MH\s0 offers.
meillo@0 1141
meillo@8 1142
meillo@8 1143
meillo@0 1144 .NH 1
meillo@0 1145 Case study: uzbl
meillo@32 1146 .LP
meillo@32 1147 The last chapter took a look on the \s-1MUA\s0 \s-1MH\s0,
meillo@32 1148 this chapter is about uzbl, a web browser that adheres to the Unix Philosophy.
meillo@32 1149 ``uzbl'' is the \fIlolcat\fP's word for the English adjective ``usable''.
meillo@32 1150 It is pronounced the identical.
meillo@0 1151
meillo@0 1152 .NH 2
meillo@32 1153 Historical background
meillo@0 1154 .LP
meillo@32 1155 Uzbl was started by Dieter Plaetinck in April 2009.
meillo@32 1156 The idea was born in a thread in the Arch Linux forum.
meillo@32 1157 .[
meillo@32 1158 %O http://bbs.archlinux.org/viewtopic.php?id=67463
meillo@32 1159 .]
meillo@32 1160 After some discussion about failures of well known web browsers,
meillo@32 1161 Plaetinck (alias Dieter@be) came up with a very sketchy proposal
meillo@32 1162 of how a better web browser could look like.
meillo@32 1163 To the question of another member, if Plaetinck would write that program,
meillo@32 1164 because it would sound fantastic, Plaetinck replied:
meillo@32 1165 ``Maybe, if I find the time ;-)''.
meillo@32 1166 .PP
meillo@32 1167 Fortunately, he found the time.
meillo@32 1168 One day later, the first prototype was out.
meillo@32 1169 One week later, uzbl had an own website.
meillo@32 1170 One month after the first code showed up,
meillo@32 1171 a mailing list was installed to coordinate and discuss further development.
meillo@32 1172 A wiki was set up to store documentation and scripts that showed up on the
meillo@32 1173 mailing list and elsewhere.
meillo@32 1174 .PP
meillo@38 1175 In the, now, one year of uzbl's existence, it was heavily developed in various branches.
meillo@32 1176 Plaetinck's task became more and more to only merge the best code from the
meillo@32 1177 different branches into his main branch, and to apply patches.
meillo@32 1178 About once a month, Plaetinck released a new version.
meillo@32 1179 In September 2009, he presented several forks of uzbl.
meillo@38 1180 Uzbl, actually, opened the field for a whole family of web browsers with similar shape.
meillo@32 1181 .PP
meillo@32 1182 In July 2009, \fILinux Weekly News\fP published an interview with Plaetinck about uzbl.
meillo@32 1183 In September 2009, the uzbl web browser was on \fISlashdot\fP.
meillo@0 1184
meillo@0 1185 .NH 2
meillo@32 1186 Contrasts to other web browsers
meillo@0 1187 .LP
meillo@32 1188 Like most \s-1MUA\s0s are monolithic, but \s-1MH\s0 is a toolchest,
meillo@32 1189 most web browsers are monolithic, but uzbl is a frontend to a toolchest.
meillo@32 1190 .PP
meillo@32 1191 Today, uzbl is divided into uzbl-core and uzbl-browser.
meillo@32 1192 Uzbl-core is, how its name already indicates, the core of uzbl.
meillo@32 1193 It handles commands and events to interface other programs,
meillo@32 1194 and also displays webpages by using webkit as render engine.
meillo@32 1195 Uzbl-browser combines uzbl-core with a bunch of handler scripts, a status bar,
meillo@32 1196 an event manager, yanking, pasting, page searching, zooming, and more stuff,
meillo@32 1197 to form a ``complete'' web browser.
meillo@32 1198 In the following text, the term ``uzbl'' usually stands for uzbl-browser,
meillo@32 1199 so uzbl-core is included.
meillo@32 1200 .PP
meillo@32 1201 Unlike most other web browsers, uzbl is mainly the mediator between the
meillo@32 1202 various tools that cover single jobs of web browsing.
meillo@35 1203 Therefore, uzbl listens for commands on a named pipe (fifo), a Unix socket,
meillo@35 1204 and on stdin, and it writes events to a Unix socket and to stdout.
meillo@35 1205 The graphical rendering of the webpage is done by webkit, a web content engine.
meillo@35 1206 Uzbl-core is build around this library.
meillo@35 1207 Loading a webpage in a running uzbl instance requires only:
meillo@32 1208 .DS
meillo@32 1209 .CW
meillo@32 1210 echo 'uri http://example.org' >/path/to/uzbl-fifo
meillo@32 1211 .DE
meillo@32 1212 .PP
meillo@32 1213 Downloads, browsing history, bookmarks, and thelike are not provided
meillo@32 1214 by uzbl-core itself, as they are in other web browsers.
meillo@35 1215 Uzbl-browser also only provides, so called, handler scripts that wrap
meillo@35 1216 external applications which provide the actual functionality.
meillo@32 1217 For instance, \fIwget\fP is used to download files and uzbl-browser
meillo@32 1218 includes a script that calls wget with appropriate options in
meillo@32 1219 a prepared environment.
meillo@32 1220 .PP
meillo@32 1221 Modern web browsers are proud to have addons, plugins, and modules, instead.
meillo@32 1222 This is their effort to achieve similar goals.
meillo@35 1223 But instead of using existing, external programs, modern web browsers
meillo@35 1224 include these functions, although they might be loaded at runtime.
meillo@0 1225
meillo@0 1226 .NH 2
meillo@32 1227 Discussion of the design
meillo@0 1228 .LP
meillo@32 1229 This section discusses uzbl in regard of the Unix Philosophy,
meillo@32 1230 as identified by Gancarz.
meillo@32 1231
meillo@32 1232 .PP
meillo@35 1233 .B "Make each program do one thing well" .
meillo@35 1234 Uzbl tries to be a web browser and nothing else.
meillo@36 1235 The common definition of a web browser is, of course, highly influenced by
meillo@36 1236 existing implementations of web browsers, although they are degenerated.
meillo@35 1237 Web browsers should be programs to browse the web, and nothing more.
meillo@35 1238 This is the one thing they should do, as demanded by the Unix Philosophy.
meillo@36 1239 .PP
meillo@36 1240 Web browsers should, for instance, not manage downloads.
meillo@35 1241 This is the job download managers exist for.
meillo@35 1242 Download managers do primary care about being good in downloading files.
meillo@35 1243 Modern web browsers provide download management only as a secondary feature.
meillo@35 1244 How could they perform this job better, than programs that exist only for
meillo@35 1245 this very job?
meillo@35 1246 And how could anyone want less than the best download manager available?
meillo@32 1247 .PP
meillo@35 1248 A web browser's job is to let the user browse the web.
meillo@35 1249 This means, navigating through websites by following links.
meillo@36 1250 Rendering the \s-1HTML\s0 sources is a different job, too.
meillo@36 1251 It is covered by the webkit render engine, in uzbl's case.
meillo@35 1252 Audio and video content and files like PostScript, \s-1PDF\s0, and the like,
meillo@36 1253 are also not the job of a web browser.
meillo@36 1254 They should be handled by external applications \(en
meillo@36 1255 ones which's job is to handle such data.
meillo@35 1256 Uzbl strives to do it this way.
meillo@36 1257 .PP
meillo@36 1258 Remember Doug McIlroy:
meillo@35 1259 .I
meillo@35 1260 ``Write programs that do one thing and do it well.
meillo@35 1261 Write programs to work together.''
meillo@35 1262 .R
meillo@35 1263 .PP
meillo@35 1264 The lesser tenet
meillo@35 1265 .B "allow the user to tailor the environment
meillo@35 1266 matches good here.
meillo@35 1267 There was the question, how anyone could want anything less than the
meillo@35 1268 best program for the job.
meillo@36 1269 But as personal preferences matter much,
meillo@36 1270 it is probably more important to ask:
meillo@35 1271 How could anyone want something else than his preferred program for the job?
meillo@36 1272 .PP
meillo@35 1273 Usually users want one program for one job.
meillo@35 1274 Hence, whenever the task is, for instance, downloading,
meillo@36 1275 exactly one download manager should be used.
meillo@35 1276 More advanced users might want to have this download manager in this
meillo@35 1277 situation and that one in that situation.
meillo@35 1278 They should be able to configure it this way.
meillo@35 1279 With uzbl, one can use any download manager the user wants.
meillo@36 1280 To switch to a different one, only one line in a small handler script
meillo@35 1281 needs to be changed.
meillo@36 1282 Alternatively it would be possible to query an entry in a global file
meillo@36 1283 or an environment variable, which specifies the download manager to use,
meillo@35 1284 in the handler script.
meillo@36 1285 .PP
meillo@35 1286 As uzbl does neither have its own download manager nor depends on a
meillo@35 1287 specific one, thus uzbl's browsing abilities will not be lowered by having
meillo@35 1288 a bad download manager.
meillo@36 1289 Uzbl's download capabilities will just as good as the ones of the best
meillo@36 1290 download manager available on the system.
meillo@38 1291 Of course, this applies to all of the other supplementary tools, too.
meillo@32 1292
meillo@32 1293 .PP
meillo@36 1294 .B "Use software leverage to your advantage" .
meillo@36 1295 Shell scripts are a good choice to extend uzbl.
meillo@36 1296 Uzbl is designed to be extended by external tools.
meillo@36 1297 These external tools are usually wrapped by small handler shell scripts.
meillo@36 1298 Shell scripts are the glue in this approach.
meillo@36 1299 They make the various parts fit together.
meillo@36 1300 .PP
meillo@36 1301 As an example, the history mechanism of uzbl shall be presented.
meillo@36 1302 Uzbl is configured to spawn a script to append an entry to the history
meillo@36 1303 whenever the event of a fully loaded page occurs.
meillo@36 1304 The script to append the entry to the history not much more than:
meillo@36 1305 .DS
meillo@36 1306 .CW
meillo@36 1307 #!/bin/sh
meillo@36 1308 file=/path/to/uzbl-history
meillo@36 1309 echo `date +'%Y-%m-%d %H:%M:%S'`" $6 $7" >> $file
meillo@36 1310 .DE
meillo@36 1311 \f(CW$6\fP and \f(CW$7\fP expand to the \s-1URL\s0 and the page title.
meillo@36 1312 For loading an entry, a key is bound to spawn a load from history script.
meillo@36 1313 The script reverses the history to have newer entries first,
meillo@36 1314 then displays \fIdmenu\fP to select an item,
meillo@36 1315 and afterwards writes the selected \s-1URL\s0 into uzbl's command input pipe.
meillo@36 1316 With error checking and corner cases removed, the script looks like this:
meillo@36 1317 .DS
meillo@36 1318 .CW
meillo@36 1319 #!/bin/sh
meillo@36 1320 file=/path/to/uzbl-history
meillo@36 1321 goto=`tac $file | dmenu | cut -d' ' -f 3`
meillo@36 1322 echo "uri $goto" > $4
meillo@36 1323 .DE
meillo@36 1324 \f(CW$4\fP expands to the path of the command input pipe of the current
meillo@36 1325 uzbl instance.
meillo@32 1326
meillo@32 1327 .PP
meillo@33 1328 .B "Avoid captive user interfaces" .
meillo@36 1329 One could say, that uzbl, to a large extend, actually \fIis\fP
meillo@36 1330 a captive user interface.
meillo@37 1331 But the difference to most other web browsers is, that uzbl is only
meillo@37 1332 the captive user interface frontend and the core of the backend.
meillo@38 1333 Many parts of the backend are independent of uzbl.
meillo@37 1334 Some are distributed with uzbl, for some external programs, handler scripts
meillo@37 1335 are distributed, arbitrary additional functionality can be added if desired.
meillo@37 1336 .PP
meillo@37 1337 The frontend is captive \(en that is true.
meillo@37 1338 This is okay for the task of browsing the web, as this task is only relevant
meillo@37 1339 for humans.
meillo@37 1340 Automated programs would \fIcrawl\fP the web.
meillo@37 1341 That means, they read the source directly.
meillo@37 1342 The source includes all the semantics.
meillo@37 1343 The graphical representation is just for humans to transfer the semantics
meillo@37 1344 more intuitively.
meillo@32 1345
meillo@32 1346 .PP
meillo@33 1347 .B "Make every program a filter" .
meillo@37 1348 Graphical web browsers are almost dead ends in the chain of information flow.
meillo@37 1349 Thus it is difficult to see what graphical web browsers should filter.
meillo@37 1350 Graphical web browsers exist almost only for interactive use by humans.
meillo@37 1351 The only case when one might want to automate the rendering function is
meillo@37 1352 to generate images of rendered webpages.
meillo@37 1353
meillo@37 1354 .PP
meillo@37 1355 .B "Small is beautiful"
meillo@38 1356 is not easy to apply to a web browser, primary because modern web technology
meillo@38 1357 is very complex; hence the rendering task is very complex.
meillo@37 1358 Modern web browsers will always consist of many thousand lines of code,
meillo@37 1359 unfortunately.
meillo@37 1360 Using the toolchest approach and wrappers can split the browser into
meillo@37 1361 several small parts, tough.
meillo@37 1362 .PP
meillo@37 1363 Uzbl-core consists of about 3\,500 lines of C code.
meillo@37 1364 The distribution includes another 3\,500 lines of Shell and Python code,
meillo@37 1365 which are the handler scripts and plugins like a modal interface.
meillo@38 1366 Further more, uzbl uses functionality of external tools like
meillo@38 1367 \fIwget\fP and \fInetcat\fP.
meillo@37 1368 Up to this point, uzbl looks pretty neat and small.
meillo@38 1369 The ugly part of uzbl is the web content renderer, webkit.
meillo@37 1370 Webkit consists of roughly 400\,000 (!) lines of code.
meillo@38 1371 Unfortunately, small web render engines are not possible anymore
meillo@38 1372 because of the modern web.
meillo@38 1373 The problems section will explain this in more detail.
meillo@35 1374
meillo@35 1375 .PP
meillo@35 1376 .B "Build a prototype as soon as possible" .
meillo@35 1377 Plaetinck made his code public, right from the beginning.
meillo@38 1378 Discussion and development was, and still is, open to everyone interested.
meillo@38 1379 Development versions of uzbl can be obtained very simply from the code
meillo@38 1380 repository.
meillo@38 1381 Within the first year of uzbl's existence, a new version was released
meillo@35 1382 more often than once a month.
meillo@38 1383 Different forks and branches arose.
meillo@38 1384 They introduced new features, which were tested for suitability.
meillo@35 1385 The experiences of using prototypes influenced further development.
meillo@35 1386 Actually, all development was community driven.
meillo@38 1387 Plaetinck says, three months after uzbl's birth:
meillo@35 1388 ``Right now I hardly code anything myself for Uzbl.
meillo@35 1389 I just merge in other people's code, ponder a lot, and lead the discussions.''
meillo@35 1390 .[
meillo@36 1391 %A FIXME
meillo@35 1392 %O http://lwn.net/Articles/341245/
meillo@35 1393 .]
meillo@32 1394
meillo@0 1395
meillo@0 1396 .NH 2
meillo@0 1397 Problems
meillo@0 1398 .LP
meillo@38 1399 Similar to \s-1MH\s0, uzbl, too suffers from being different.
meillo@38 1400 It is sad, but people use what they know.
meillo@38 1401 Fortunately, uzbl's user interface can look and feel very much the
meillo@38 1402 same as the one of the well known web browsers,
meillo@38 1403 hiding the internal differences.
meillo@38 1404 But uzbl has to provide this similar look and feel to be accepted
meillo@38 1405 as a ``normal'' browser by ``normal'' users.
meillo@37 1406 .PP
meillo@38 1407 The more important problem is the modern web.
meillo@38 1408 The modern web is simply broken.
meillo@38 1409 It has state in a state-less protocol,
meillo@38 1410 it misuses technologies,
meillo@38 1411 and it is helplessly overloaded.
meillo@38 1412 The result are web content render engines that must consist
meillo@38 1413 of hundreds of thousands lines of code.
meillo@38 1414 They also must combine and integrate many different technologies,
meillo@38 1415 only to make our modern web usable.
meillo@38 1416 Website to image converter are hardly possible to run without
meillo@38 1417 human interaction because of state in sessions, impossible
meillo@38 1418 deep-linking, and unautomatable technologies.
meillo@37 1419 .PP
meillo@38 1420 The web was misused to provide all kinds of imaginable wishes.
meillo@38 1421 Now web browsers, and eventually the users, suffer from it.
meillo@37 1422
meillo@8 1423
meillo@32 1424 .NH 2
meillo@32 1425 Summary uzbl
meillo@32 1426 .LP
meillo@38 1427 ``Uzbl is a browser that adheres to the Unix Philosophy'',
meillo@38 1428 that is how uzbl is seen by its authors.
meillo@38 1429 Indeed, uzbl follows the Unix Philosophy in many ways.
meillo@38 1430 It consists of independent parts that work together,
meillo@38 1431 its core is mainly a mediator which glues the parts together.
meillo@38 1432 .PP
meillo@38 1433 Software leverage can excellently be seen in uzbl.
meillo@38 1434 It makes use of external tools, separates independent tasks
meillo@38 1435 in independent parts, and glues them together with small
meillo@38 1436 handler scripts, around uzbl-core.
meillo@38 1437 .PP
meillo@38 1438 As uzbl, more or less, consists of a set of tools and a bit
meillo@38 1439 of glue, anyone can put the parts together and expand it
meillo@38 1440 in any desired way.
meillo@38 1441 Uzbl is very flexible and customizable.
meillo@38 1442 These properties make it valuable for advanced users,
meillo@38 1443 but may keep novice users from using it.
meillo@38 1444 .PP
meillo@38 1445 Uzbl's main problem is the modern web, that makes it hard
meillo@38 1446 to design a sane web browser.
meillo@38 1447 Despite this bad situation, uzbl does a fairly good job.
meillo@32 1448
meillo@8 1449
meillo@0 1450 .NH 1
meillo@0 1451 Final thoughts
meillo@0 1452
meillo@0 1453 .NH 2
meillo@0 1454 Quick summary
meillo@0 1455 .LP
meillo@0 1456 good design
meillo@0 1457 .LP
meillo@0 1458 unix phil
meillo@0 1459 .LP
meillo@0 1460 case studies
meillo@0 1461
meillo@0 1462 .NH 2
meillo@0 1463 Why people should choose
meillo@0 1464 .LP
meillo@0 1465 Make the right choice!
meillo@0 1466
meillo@0 1467 .nr PI .5i
meillo@0 1468 .rm ]<
meillo@0 1469 .de ]<
meillo@0 1470 .LP
meillo@0 1471 .de FP
meillo@0 1472 .IP \\\\$1.
meillo@0 1473 \\..
meillo@0 1474 .rm FS FE
meillo@0 1475 ..
meillo@0 1476 .SH
meillo@0 1477 References
meillo@0 1478 .[
meillo@0 1479 $LIST$
meillo@0 1480 .]
meillo@0 1481 .wh -1p