The UNIX Time-Sharing System* D. M. Ritchie and K. Thompson _A_B_S_T_R_A_C_T UNIX* is a general-purpose, multi-user, interactive operating system for the larger Digi- tal Equipment Corporation PDP-11 and the Interdata 8/32 computers. It offers a number of features seldom found even in larger operating systems, including i A hierarchical file system incorporating demount- able volumes, ii Compatible file, device, and inter-process I/O, iii The ability to initiate asynchronous processes, iv System command language selectable on a per-user basis, v Over 100 subsystems including a dozen languages, vi High degree of portability. This paper discusses the nature and implementation of the file system and of the user command interface. _1. _I_N_T_R_O_D_U_C_T_I_O_N There have been four versions of the UNIX time-sharing system. The earliest (circa 1969-70) ran on the Digital Equipment Corporation PDP-7 and -9 computers. The second __________________________ * Copyright 1974, Association for Computing Machinery, Inc., reprinted by permission. This is a revised ver- sion of an article that appeared in Communications of the ACM, _1_7, No. 7 (July 1974), pp. 365-375. That ar- ticle was a revised version of a paper presented at the Fourth ACM Symposium on Operating Systems Principles, IBM Thomas J. Watson Research Center, Yorktown Heights, New York, October 15-17, 1973. * UNIX is a Trademark of Bell Laboratories. September 27, 1987 - 2 - version ran on the unprotected PDP-11/20 computer. The third incorporated multiprogramming and ran on the PDP- 11/34, /40, /45, /60, and /70 computers; it is the one described in the previously published version of this paper, and is also the most widely used today. This paper describes only the fourth, current system that runs on the PDP-11/70 and the Interdata 8/32 computers. In fact, the differences among the various systems is rather small; most of the revisions made to the originally published version of this paper, aside from those concerned with style, had to do with details of the implementation of the file system. Since PDP-11 UNIX became operational in February, 1971, over 600 installations have been put into service. Most of them are engaged in applications such as computer science education, the preparation and formatting of documents and other textual material, the collection and processing of trouble data from various switching machines within the Bell System, and recording and checking telephone service orders. Our own installation is used mainly for research in operat- ing systems, languages, computer networks, and other topics in computer science, and also for document preparation. Perhaps the most important achievement of UNIX is to demonstrate that a powerful operating system for interactive use need not be expensive either in equipment or in human effort: it can run on hardware costing as little as $40,000, and less than two man-years were spent on the main system software. We hope, however, that users find that the most important characteristics of the system are its simplicity, elegance, and ease of use. Besides the operating system proper, some major pro- grams available under UNIX are C compiler Text editor based on QED [ qed lampson ] Assembler, linking loader, symbolic debugger Phototypesetting and equation setting programs [ cherry kernighan typesetting mathematics cacm ] [ kernighan lesk ossanna document preparation bstj %Q This issue ] Dozens of languages including Fortran 77, Basic, Snobol, APL, Algol 68, M6, TMG, Pas- cal There is a host of maintenance, utility, recreation and September 27, 1987 - 3 - novelty programs, all written locally. The UNIX user com- munity, which numbers in the thousands, has contributed many more programs and languages. It is worth noting that the system is totally self-supporting. All UNIX software is maintained on the system; likewise, this paper and all other documents in this issue were generated and formatted by the UNIX editor and text formatting programs. _I_I. _H_A_R_D_W_A_R_E _A_N_D _S_O_F_T_W_A_R_E _E_N_V_I_R_O_N_M_E_N_T The PDP-11/70 on which the Research UNIX system is installed is a 16-bit word (8-bit byte) computer with 768K bytes of core memory; the system kernel occupies 90K bytes about equally divided between code and data tables. This system, however, includes a very large number of device drivers and enjoys a generous allotment of space for I/O buffers and system tables; a minimal system capable of run- ning the software mentioned above can require as little as 96K bytes of core altogether. There are even larger instal- lations; see the description of the PWB/UNIX systems, [ dolotta mashey workbench software engineering ] [ dolotta haight mashey workbench bstj %Q This issue ] for example. There are also much smaller, though somewhat restricted, versions of the system. [ lycklama microprocessor bstj %Q This issue ] Our own PDP-11 has two 200-Mb moving-head disks for file system storage and swapping. There are 20 variable- speed communications interfaces attached to 300- and 1200- baud data sets, and an additional 12 communication lines hard-wired to 9600-baud terminals and satellite computers. There are also several 2400- and 4800-baud synchronous com- munication interfaces used for machine-to-machine file transfer. Finally, there is a variety of miscellaneous dev- ices including nine-track magnetic tape, a line printer, a voice synthesizer, a phototypesetter, a digital switching network, and a chess machine. The preponderance of UNIX software is written in the abovementioned C language. [ c programming language ker- nighan ritchie prentice-hall ] Early versions of the operat- ing system were written in assembly language, but during the summer of 1973, it was rewritten in C. The size of the new system was about one-third greater than that of the old. Since the new system not only became much easier to under- stand and to modify but also included many functional improvements, including multiprogramming and the ability to share reentrant code among several user programs, we con- sider this increase in size quite acceptable. _I_I_I. _T_H_E _F_I_L_E _S_Y_S_T_E_M The most important role of the system is to provide a file system. From the point of view of the user, there are September 27, 1987 - 4 - three kinds of files: ordinary disk files, directories, and special files. _3._1 _O_r_d_i_n_a_r_y _f_i_l_e_s A file contains whatever information the user places on it, for example, symbolic or binary (object) programs. No particular structuring is expected by the system. A file of text consists simply of a string of characters, with lines demarcated by the newline character. Binary programs are sequences of words as they will appear in core memory when the program starts executing. A few user programs manipu- late files with more structure; for example, the assembler generates, and the loader expects, an object file in a par- ticular format. However, the structure of files is con- trolled by the programs that use them, not by the system. _3._2 _D_i_r_e_c_t_o_r_i_e_s Directories provide the mapping between the names of files and the files themselves, and thus induce a structure on the file system as a whole. Each user has a directory of his own files; he may also create subdirectories to contain groups of files conveniently treated together. A directory behaves exactly like an ordinary file except that it cannot be written on by unprivileged programs, so that the system controls the contents of directories. However, anyone with appropriate permission may read a directory just like any other file. The system maintains several directories for its own use. One of these is the root directory. All files in the system can be found by tracing a path through a chain of directories until the desired file is reached. The starting point for such searches is often the root. Other system directories contain all the programs provided for general use; that is, all the _c_o_m_m_a_n_d_s. As will be seen, however, it is by no means necessary that a program reside in one of these directories for it to be executed. Files are named by sequences of 14 or fewer characters. When the name of a file is specified to the system, it may be in the form of a _p_a_t_h _n_a_m_e, which is a sequence of direc- tory names separated by slashes, ``/'', and ending in a file name. If the sequence begins with a slash, the search begins in the root directory. The name /alpha/beta/gamma causes the system to search the root for directory alpha, then to search alpha for beta, finally to find gamma in beta. gamma may be an ordinary file, a directory, or a spe- cial file. As a limiting case, the name ``/'' refers to the root itself. A path name not starting with ``/'' causes the system to begin the search in the user's current directory. Thus, September 27, 1987 - 5 - the name alpha/beta specifies the file named beta in sub- directory alpha of the current directory. The simplest kind of name, for example, alpha, refers to a file that itself is found in the current directory. As another limiting case, the null file name refers to the current directory. The same non-directory file may appear in several directories under possibly different names. This feature is called _l_i_n_k_i_n_g; a directory entry for a file is sometimes called a link. The UNIX system differs from other systems in which linking is permitted in that all links to a file have equal status. That is, a file does not exist within a particular directory; the directory entry for a file con- sists merely of its name and a pointer to the information actually describing the file. Thus a file exists indepen- dently of any directory entry, although in practice a file is made to disappear along with the last link to it. Each directory always has at least two entries. The name ``.'' in each directory refers to the directory itself. Thus a program may read the current directory under the name ``.'' without knowing its complete path name. The name ``..'' by convention refers to the parent of the directory in which it appears, that is, to the directory in which it was created. The directory structure is constrained to have the form of a rooted tree. Except for the special entries ``.'' and ``..'', each directory must appear as an entry in exactly one other directory, which is its parent. The reason for this is to simplify the writing of programs that visit sub- trees of the directory structure, and more important, to avoid the separation of portions of the hierarchy. If arbi- trary links to directories were permitted, it would be quite difficult to detect when the last connection from the root to a directory was severed. _3._3 _S_p_e_c_i_a_l _f_i_l_e_s Special files constitute the most unusual feature of the UNIX file system. Each supported I/O device is associ- ated with at least one such file. Special files are read and written just like ordinary disk files, but requests to read or write result in activation of the associated device. An entry for each special file resides in directory /dev, although a link may be made to one of these files just as it may to an ordinary file. Thus, for example, to write on a magnetic tape one may write on the file /dev/mt. Special files exist for each communication line, each disk, each tape drive, and for physical main memory. Of course, the active disks and the memory special file are protected from indiscriminate access. There is a threefold advantage in treating I/O devices September 27, 1987 - 6 - this way: file and device I/O are as similar as possible; file and device names have the same syntax and meaning, so that a program expecting a file name as a parameter can be passed a device name; finally, special files are subject to the same protection mechanism as regular files. _3._4 _R_e_m_o_v_a_b_l_e _f_i_l_e _s_y_s_t_e_m_s Although the root of the file system is always stored on the same device, it is not necessary that the entire file system hierarchy reside on this device. There is a mount system request with two arguments: the name of an existing ordinary file, and the name of a special file whose associ- ated storage volume (e.g., a disk pack) should have the structure of an independent file system containing its own directory hierarchy. The effect of mount is to cause refer- ences to the heretofore ordinary file to refer instead to the root directory of the file system on the removable volume. In effect, mount replaces a leaf of the hierarchy tree (the ordinary file) by a whole new subtree (the hierar- chy stored on the removable volume). After the mount, there is virtually no distinction between files on the removable volume and those in the permanent file system. In our installation, for example, the root directory resides on a small partition of one of our disk drives, while the other drive, which contains the user's files, is mounted by the system initialization sequence. A mountable file system is generated by writing on its corresponding special file. A utility program is available to create an empty file system, or one may simply copy an existing file system. There is only one exception to the rule of identical treatment of files on different devices: no link may exist between one file system hierarchy and another. This res- triction is enforced so as to avoid the elaborate bookkeep- ing that would otherwise be required to assure removal of the links whenever the removable volume is dismounted. _3._5 _P_r_o_t_e_c_t_i_o_n Although the access control scheme is quite simple, it has some unusual features. Each user of the system is assigned a unique user identification number. When a file is created, it is marked with the user ID of its owner. Also given for new files is a set of ten protection bits. Nine of these specify independently read, write, and execute permission for the owner of the file, for other members of his group, and for all remaining users. If the tenth bit is on, the system will temporarily change the user identification (hereafter, user ID) of the current user to that of the creator of the file whenever the file is executed as a program. This change in user ID is effective only during the execution of the program that September 27, 1987 - 7 - calls for it. The set-user-ID feature provides for privileged programs that may use files inaccessible to other users. For example, a program may keep an accounting file that should neither be read nor changed except by the pro- gram itself. If the set-user-ID bit is on for the program, it may access the file although this access might be forbid- den to other programs invoked by the given program's user. Since the actual user ID of the invoker of any program is always available, set-user-ID programs may take any measures desired to satisfy themselves as to their invoker's creden- tials. This mechanism is used to allow users to execute the carefully written commands that call privileged system entries. For example, there is a system entry invokable only by the ``super-user'' (below) that creates an empty directory. As indicated above, directories are expected to have entries for ``.'' and ``..''. The command which creates a directory is owned by the super-user and has the set-user-ID bit set. After it checks its invoker's authori- zation to create the specified directory, it creates it and makes the entries for ``.'' and ``..''. Because anyone may set the set-user-ID bit on one of his own files, this mechanism is generally available without administrative intervention. For example, this protection scheme easily solves the MOO accounting problem posed by ``Aleph-null.'' [ aleph null software practice ] The system recognizes one particular user ID (that of the ``super-user'') as exempt from the usual constraints on file access; thus (for example), programs may be written to dump and reload the file system without unwanted interfer- ence from the protection system. _3._6 _I/_O _c_a_l_l_s The system calls to do I/O are designed to eliminate the differences between the various devices and styles of access. There is no distinction between ``random'' and ``sequential'' I/O, nor is any logical record size imposed by the system. The size of an ordinary file is determined by the number of bytes written on it; no predetermination of the size of a file is necessary or possible. To illustrate the essentials of I/O, some of the basic calls are summarized below in an anonymous language that will indicate the required parameters without getting into the underlying complexities. Each call to the system may potentially result in an error return, which for simplicity is not represented in the calling sequence. To read or write a file assumed to exist already, it must be opened by the following call: filep = open(name, flag) September 27, 1987 - 8 - where name indicates the name of the file. An arbitrary path name may be given. The flag argument indicates whether the file is to be read, written, or ``updated,'' that is, read and written simultaneously. The returned value filep is called a _f_i_l_e _d_e_s_c_r_i_p_t_o_r. It is a small integer used to identify the file in subse- quent calls to read, write, or otherwise manipulate the file. To create a new file or completely rewrite an old one, there is a create system call that creates the given file if it does not exist, or truncates it to zero length if it does exist; create also opens the new file for writing and, like open, returns a file descriptor. The file system maintains no locks visible to the user, nor is there any restriction on the number of users who may have a file open for reading or writing. Although it is possible for the contents of a file to become scrambled when two users write on it simultaneously, in practice difficul- ties do not arise. We take the view that locks are neither necessary nor sufficient, in our environment, to prevent interference between users of the same file. They are unnecessary because we are not faced with large, single-file data bases maintained by independent processes. They are insufficient because locks in the ordinary sense, whereby one user is prevented from writing on a file that another user is reading, cannot prevent confusion when, for example, both users are editing a file with an editor that makes a copy of the file being edited. There are, however, sufficient internal interlocks to maintain the logical consistency of the file system when two users engage simultaneously in activities such as writing on the same file, creating files in the same directory, or deleting each other's open files. Except as indicated below, reading and writing are sequential. This means that if a particular byte in the file was the last byte written (or read), the next I/O call implicitly refers to the immediately following byte. For each open file there is a pointer, maintained inside the system, that indicates the next byte to be read or written. If _n bytes are read or written, the pointer advances by _n bytes. Once a file is open, the following calls may be used: n = read(filep, buffer, count) n = write(filep, buffer, count) Up to count bytes are transmitted between the file specified by filep and the byte array specified by buffer. The September 27, 1987 - 9 - returned value n is the number of bytes actually transmit- ted. In the write case, n is the same as count except under exceptional conditions, such as I/O errors or end of physi- cal medium on special files; in a read, however, n may without error be less than count. If the read pointer is so near the end of the file that reading count characters would cause reading beyond the end, only sufficient bytes are transmitted to reach the end of the file; also, typewriter- like terminals never return more than one line of input. When a read call returns with n equal to zero, the end of the file has been reached. For disk files this occurs when the read pointer becomes equal to the current size of the file. It is possible to generate an end-of-file from a ter- minal by use of an escape sequence that depends on the dev- ice used. Bytes written affect only those parts of a file implied by the position of the write pointer and the count; no other part of the file is changed. If the last byte lies beyond the end of the file, the file is made to grow as needed. To do random (direct-access) I/O it is only necessary to move the read or write pointer to the appropriate loca- tion in the file. location = lseek(filep, offset, base) The pointer associated with filep is moved to a position offset bytes from the beginning of the file, from the current position of the pointer, or from the end of the file, depending on base. offset may be negative. For some devices (e.g., paper tape and terminals) seek calls are ignored. The actual offset from the beginning of the file to which the pointer was moved is returned in location. There are several additional system entries having to do with I/O and with the file system that will not be dis- cussed. For example: close a file, get the status of a file, change the protection mode or the owner of a file, create a directory, make a link to an existing file, delete a file. _I_V. _I_M_P_L_E_M_E_N_T_A_T_I_O_N _O_F _T_H_E _F_I_L_E _S_Y_S_T_E_M As mentioned in Section 3.2 above, a directory entry contains only a name for the associated file and a pointer to the file itself. This pointer is an integer called the _i-_n_u_m_b_e_r (for index number) of the file. When the file is accessed, its i-number is used as an index into a system table (the _i-_l_i_s_t) stored in a known part of the device on which the directory resides. The entry found thereby (the file's _i-_n_o_d_e) contains the description of the file: i the user and group-ID of its owner September 27, 1987 - 10 - ii its protection bits iii the physical disk or tape addresses for the file con- tents iv its size v time of creation, last use, and last modification vi the number of links to the file, that is, the number of times it appears in a directory vii a code indicating whether the file is a directory, an ordinary file, or a special file. The purpose of an open or create system call is to turn the path name given by the user into an i-number by searching the explicitly or implicitly named directories. Once a file is open, its device, i-number, and read/write pointer are stored in a system table indexed by the file descriptor returned by the open or create. Thus, during a subsequent call to read or write the file, the descriptor may be easily related to the information necessary to access the file. When a new file is created, an i-node is allocated for it and a directory entry is made that contains the name of the file and the i-node number. Making a link to an exist- ing file involves creating a directory entry with the new name, copying the i-number from the original file entry, and incrementing the link-count field of the i-node. Removing (deleting) a file is done by decrementing the link-count of the i-node specified by its directory entry and erasing the directory entry. If the link-count drops to 0, any disk blocks in the file are freed and the i-node is de-allocated. The space on all disks that contain a file system is divided into a number of 512-byte blocks logically addressed from 0 up to a limit that depends on the device. There is space in the i-node of each file for 13 device addresses. For nonspecial files, the first 10 device addresses point at the first 10 blocks of the file. If the file is larger than 10 blocks, the 11 device address points to an indirect block containing up to 128 addresses of additional blocks in the file. Still larger files use the twelfth device address of the i-node to point to a double-indirect block naming 128 indirect blocks, each pointing to 128 blocks of the file. If required, the thirteenth device address is a triple- indirect block. Thus files may conceptually grow to [(10+128+128829+128839)8.9512] bytes. Once opened, bytes numbered below 5120 can be read with a single disk access; bytes in the range 5120 to 70,656 require two accesses; bytes in the range 70,656 to 8,459,264 require three accesses; bytes from there to the largest file (1,082,201,088) require four accesses. In practice, a device cache mechanism (see below) September 27, 1987 - 11 - proves effective in eliminating most of the indirect fetches. The foregoing discussion applies to ordinary files. When an I/O request is made to a file whose i-node indicates that it is special, the last 12 device address words are immaterial, and the first specifies an internal _d_e_v_i_c_e _n_a_m_e, which is interpreted as a pair of numbers representing, respectively, a device type and subdevice number. The dev- ice type indicates which system routine will deal with I/O on that device; the subdevice number selects, for example, a disk drive attached to a particular controller or one of several similar terminal interfaces. In this environment, the implementation of the mount system call (Section 3.4) is quite straightforward. mount maintains a system table whose argument is the i-number and device name of the ordinary file specified during the mount, and whose corresponding value is the device name of the indicated special file. This table is searched for each i- number/device pair that turns up while a path name is being scanned during an open or create; if a match is found, the i-number is replaced by the i-number of the root directory and the device name is replaced by the table value. To the user, both reading and writing of files appear to be synchronous and unbuffered. That is, immediately after return from a read call the data are available; con- versely, after a write the user's workspace may be reused. In fact, the system maintains a rather complicated buffering mechanism that reduces greatly the number of I/O operations required to access a file. Suppose a write call is made specifying transmission of a single byte. The system will search its buffers to see whether the affected disk block currently resides in main memory; if not, it will be read in from the device. Then the affected byte is replaced in the buffer and an entry is made in a list of blocks to be writ- ten. The return from the write call may then take place, although the actual I/O may not be completed until a later time. Conversely, if a single byte is read, the system determines whether the secondary storage block in which the byte is located is already in one of the system's buffers; if so, the byte can be returned immediately. If not, the block is read into a buffer and the byte picked out. The system recognizes when a program has made accesses to sequential blocks of a file, and asynchronously pre-reads the next block. This significantly reduces the running time of most programs while adding little to system overhead. A program that reads or writes files in units of 512 bytes has an advantage over a program that reads or writes a single byte at a time, but the gain is not immense; it comes mainly from the avoidance of system overhead. If a program September 27, 1987 - 12 - is used rarely or does no great volume of I/O, it may quite reasonably read and write in units as small as it wishes. The notion of the i-list is an unusual feature of UNIX. In practice, this method of organizing the file system has proved quite reliable and easy to deal with. To the system itself, one of its strengths is the fact that each file has a short, unambiguous name related in a simple way to the protection, addressing, and other information needed to access the file. It also permits a quite simple and rapid algorithm for checking the consistency of a file system, for example, verification that the portions of each device con- taining useful information and those free to be allocated are disjoint and together exhaust the space on the device. This algorithm is independent of the directory hierarchy, because it need only scan the linearly organized i-list. At the same time the notion of the i-list induces certain pecu- liarities not found in other file system organizations. For example, there is the question of who is to be charged for the space a file occupies, because all directory entries for a file have equal status. Charging the owner of a file is unfair in general, for one user may create a file, another may link to it, and the first user may delete the file. The first user is still the owner of the file, but it should be charged to the second user. The simplest reasonably fair algorithm seems to be to spread the charges equally among users who have links to a file. Many installations avoid the issue by not charging any fees at all. _V. _P_R_O_C_E_S_S_E_S _A_N_D _I_M_A_G_E_S An _i_m_a_g_e is a computer execution environment. It includes a memory image, general register values, status of open files, current directory and the like. An image is the current state of a pseudo-computer. A _p_r_o_c_e_s_s is the execution of an image. While the pro- cessor is executing on behalf of a process, the image must reside in main memory; during the execution of other processes it remains in main memory unless the appearance of an active, higher-priority process forces it to be swapped out to the disk. The user-memory part of an image is divided into three logical segments. The program text segment begins at loca- tion 0 in the virtual address space. During execution, this segment is write-protected and a single copy of it is shared among all processes executing the same program. At the first hardware protection byte boundary above the program text segment in the virtual address space begins a non- shared, writable data segment, the size of which may be extended by a system call. Starting at the highest address in the virtual address space is a stack segment, which automatically grows downward as the stack pointer September 27, 1987 - 13 - fluctuates. _5._1 _P_r_o_c_e_s_s_e_s Except while the system is bootstrapping itself into operation, a new process can come into existence only by use of the fork system call: processid = fork() When fork is executed, the process splits into two indepen- dently executing processes. The two processes have indepen- dent copies of the original memory image, and share all open files. The new processes differ only in that one is con- sidered the parent process: in the parent, the returned processid actually identifies the child process and is never 0, while in the child, the returned value is always 0. Because the values returned by fork in the parent and child process are distinguishable, each process may deter- mine whether it is the parent or child. _5._2 _P_i_p_e_s Processes may communicate with related processes using the same system read and write calls that are used for file-system I/O. The call: filep = pipe() returns a file descriptor filep and creates an inter-process channel called a _p_i_p_e. This channel, like other open files, is passed from parent to child process in the image by the fork call. A read using a pipe file descriptor waits until another process writes using the file descriptor for the same pipe. At this point, data are passed between the images of the two processes. Neither process need know that a pipe, rather than an ordinary file, is involved. Although inter-process communication via pipes is a quite valuable tool (see Section 6.2), it is not a com- pletely general mechanism, because the pipe must be set up by a common ancestor of the processes involved. _5._3 _E_x_e_c_u_t_i_o_n _o_f _p_r_o_g_r_a_m_s Another major system primitive is invoked by execute(file, arg918, arg928, ... , arg9n8) which requests the system to read in and execute the program named by file, passing it string arguments arg918, arg928, ..., arg9n8. All the code and data in the process invoking execute is replaced from the file, but open files, current September 27, 1987 - 14 - directory, and inter-process relationships are unaltered. Only if the call fails, for example because file could not be found or because its execute-permission bit was not set, does a return take place from the execute primitive; it resembles a ``jump'' machine instruction rather than a sub- routine call. _5._4 _P_r_o_c_e_s_s _s_y_n_c_h_r_o_n_i_z_a_t_i_o_n Another process control system call: processid = wait(status) causes its caller to suspend execution until one of its children has completed execution. Then wait returns the processid of the terminated process. An error return is taken if the calling process has no descendants. Certain status from the child process is also available. _5._5 _T_e_r_m_i_n_a_t_i_o_n Lastly: exit(status) terminates a process, destroys its image, closes its open files, and generally obliterates it. The parent is notified through the wait primitive, and status is made available to it. Processes may also terminate as a result of various illegal actions or user-generated signals (Section VII below). _V_I. _T_H_E _S_H_E_L_L For most users, communication with the system is car- ried on with the aid of a program called the shell. The shell is a command-line interpreter: it reads lines typed by the user and interprets them as requests to execute other programs. (The shell is described fully elsewhere, [ bourne shell bstj %Q This issue ] so this section will discuss only the theory of its operation.) In simplest form, a command line consists of the command name followed by arguments to the command, all separated by spaces: command arg918 arg928 ... arg9n 9The shell splits up the command name and the arguments into separate strings. Then a file with name command is sought; command may be a path name including the ``/'' character to specify any file in the system. If command is found, it is brought into memory and executed. The arguments collected by the shell are accessible to the command. When the com- mand is finished, the shell resumes its own execution, and indicates its readiness to accept another command by typing September 27, 1987 - 15 - a prompt character. If file command cannot be found, the shell generally prefixes a string such as /bin/ to command and attempts again to find the file. Directory /bin contains commands intended to be generally used. (The sequence of directories to be searched may be changed by user request.) _6._1 _S_t_a_n_d_a_r_d _I/_O The discussion of I/O in Section III above seems to imply that every file used by a program must be opened or created by the program in order to get a file descriptor for the file. Programs executed by the shell, however, start off with three open files with file descriptors 0, 1, and 2. As such a program begins execution, file 1 is open for writ- ing, and is best understood as the standard output file. Except under circumstances indicated below, this file is the user's terminal. Thus programs that wish to write informa- tive information ordinarily use file descriptor 1. Con- versely, file 0 starts off open for reading, and programs that wish to read messages typed by the user read this file. The shell is able to change the standard assignments of these file descriptors from the user's terminal printer and keyboard. If one of the arguments to a command is prefixed by ``>'', file descriptor 1 will, for the duration of the command, refer to the file named after the ``>''. For exam- ple: ls ordinarily lists, on the typewriter, the names of the files in the current directory. The command: ls >there creates a file called there and places the listing there. Thus the argument >there means ``place output on there.'' On the other hand: ed ordinarily enters the editor, which takes requests from the user via his keyboard. The command ed