PC Online 1996 October

home *** CD-ROM | disk | FTP | other *** search

/ PC Online 1996 October / PCO_10.ISO / filesbbs / flstw136.arj / FASTLST.DOC < prev next >

Wrap

Text File | 1996-07-18 | 135.1 KB | 3,655 lines

************************************************************** * * * * * ******* ** **** ****** **** **** ****** * * ** * **** ** ** * ** * ** ** ** * ** * * * ** * ** ** ** ** ** ** ** * * **** ** ** **** ** ** **** ** * * ** * ****** ** ** ** * ** ** * * ** ** ** ** ** ** ** ** ** ** ** * * **** ** ** **** **** ******* **** **** * * * * * * Version 1.36 * * * * The ultimate V7 nodelist processor * * * * * ************************************************************** * * * (C) Copyright 1992-1996 by Alberto Pasquale * * * * A L L R I G H T S R E S E R V E D * * * ************************************************************** FastLst 1.36 User's Manual, by Alberto Pasquale INTRODUCTION -> For licensing information, please see License.Doc. Thanks for evaluating FastLst: the ultimate "Version 7" nodelist processor. Version 7 is a common format for binary nodelists to be used by mailers, message editors and all the programs that need access to a compiled nodelist. Main Features - Compiles to Version 7 format nodelist. - Support for old "Fidouser.Lst" sysop list. - Multiple output nodelist (V7) compilation from one config file. - The complete maintenance of archived lists and diffs is achieved through internal flexible configuration, with no need for clumsy batch files. - Uses "Squish Style" Compress.Cfg. - Can be invoked from a batch file at predefined events: the compilation will take place only if some new list/diff is found. - The 32 bit versions can use "in memory" mode to accomplish blazing speed on systems with enough memory or "temporary file" mode to avoid very slow operations in compiling large nodelists on systems with little physical memory. - Multitasking friendly - The OS/2 version allows for priority setting in the configuration file. - Compilation reports can be posted to Fido or Squish format message areas. - Full 4D (point) support, both via the "Point,..." and "Boss,..." keywords. - Internal support for "German type" pointlists. - Easy addition of nodes via the "Node,<address>,..." keyword in a private list. - Easy specification of phone strings not to be translated for internet addresses and script names. - In the case of SysOps of multiple nodes, keeps in the sysop lists (fidouser.lst and sysop.ndx) all the name/address entries. Allows to keep one address only for specified ones. - User Cost (Msg Fee) can be set different from Call Cost. CREDITS This program uses the Squish "MsgAPI" code, Copyright 1991-1994 by Lanius Corporation. "Squish" and "Maximus" are trademarks of Lanius Corporation. "BinkleyTerm" is a trademark of Bit Bucket Software Co. "4OS2" is a trademark of JP Software Inc. The archivers referred-to throughout this documentation are Copyright and/or trademarks of the respective owners. VMODEM and SIO are Copyright by Raymond L. Gwinn. OVERALL OPERATION FastLst has been designed to be invoked regularly from one of your main batch files, after mail has been received or at pre-arranged times at your pleasure: if any new (compressed or not) nodelist/nodediff is detected, Fastlst processes them as required (exiting with Errorlevel 0), otherwise it immediately exits (errorlevel 100) with no further delay. When FastLst detects a changed config or password file, it compiles all the affected nodelists even if they are not new. If you want FastLst to compile all of your nodelists even if no new ones are present, you need to use the -f or -i command line switch. For each "output block" in the config file: - New compressed lists or diffs are detected, unarchived and optionally rearchived in supplementary formats. - New diffs are detected and applied: the resulting new nodelist is archived. Before applying a diff, the day number and CRC of the old nodelist are compared against the ones requested by the diff; after application, the CRC of the new nodelist is checked. - New lists are detected and the pertinent output nodelists are rebuilt. If no new list is found for a specific "output block", that output nodelist is not compiled, unless the -f or -i command line switch is specified. OS commands can be issued before or after each operation: for example you can hatch the just created archive file. ATTENTION: - Every time the config file is changed, FastLst rebuilds all the output nodelists, as if the -f command line switch were specified. - Every time a PasswordFile is changed, FastLst rebuilds the nodelists that use it. - The date of the archive files handled by FastLst is set to the same value of the contained file (see ArcDate). INPUT NODELIST FORMAT The source nodelists and nodediffs must be in standard "St. Louis" format, as described in FTS-0005. Anyway, FastLst allows some extensions in order to support 4D points, "German style" pointlists, easy single node specifications and "verbatim" phone strings. 4D Point Support: POINT and BOSS keywords First method: Points are entered in the nodelist directly following their bossnode. Each one starts with the "Point,<point>" keyword. Example: ... ... ,504,Videl_3,Modena_I,Roberto_Zanasi,39-59-450600,9600,CM,XA,V32B,V42B Point,1,Pasquale,Modena_I,Alberto_Pasquale,-!Unpublished-,9600, Point,2,SysOp,Modena_I,Roberto_Zanasi,-!Unpublished-,9600, Point,3,Carta,Modena,Francesco_Carta,-!Unpublished-,9600, ... ... Second method: Points are entered in the nodelist following the "Boss,<address>" keyword. Example: ... ... Boss,2:332/504 ,1,Pasquale,Modena_I,Alberto_Pasquale,-!Unpublished-,9600, ,2,SysOp,Modena_I,Roberto_Zanasi,-!Unpublished-,9600, ,3,Carta,Modena,Francesco_Carta,-!Unpublished-,9600, ... ... German Point List This is a "normal" 3D nodelist that lists each Boss as a "fakenet" HOST, with the real address as the system name, followed by its points listed as nodes. Example: The following nodelist segment lists points 2:2400/1.1 .2 .3: Host,20000,2400/1,City,Sysop_Name,49-951-999999,9600,CM,V34 ,1,System_Name_1,City_1,Sysop_Name_1,49-951-999999,9600, ,2,System_Name_2,City_2,Sysop_Name_2,49-951-999999,9600, ,3,System_Name_3,City_3,Sysop_Name_3,49-951-999999,9600, The NODE Keyword Another extension over FTS-0005 is provided to allow easy addition of nodes in small private lists. When you need to add a node to your nodelist to call it or to enforce a session password with it, you can use the "Node,<address>,..." keyword to avoid the necessity of adding its Zone and Host coordinators. You should specify a full 4D address (point optional). Any subsequent entry will take the current address as a starting point. E.g.: You want to add 9:8/7.3 With "Node,...": ... ... Node,9:8/7.3,Mickey,DisneyLand,Mickey_Mouse,1-800-111,9600,CM ... ... With "Boss,...": ... ... Boss,9:8/7 ,3,Mickey,DisneyLand,Mickey_Mouse,1-800-111,9600,CM ... ... With the traditional method: ... ... Zone,9,... Host,8,... ,7,... Point,3,Mickey,DisneyLand,Mickey_Mouse,1-800-111,9600,CM ... ... Now let's add 8:7/6 and 8:7/7: With "Node,...": ... ... Node,8:7/6,Mickey,DisneyLand,Mickey_Mouse,1-800-111,9600,CM ,7,Duck,DisneyLand,Donald_Duck,1-800-112,9600,CM ... ... With the traditional method: ... ... Zone,8,... Host,7,... ,6,Mickey,DisneyLand,Mickey_Mouse,1-800-111,9600,CM ,7,Duck,DisneyLand,Donald_Duck,1-800-112,9600,CM ... ... Verbatim Phones When you have to specify a phone string not to be translated, it's handy to just inform FastLst by prefixing the string with the '#' character. The number is considered already in the correct form for dialing, neither dial translation nor cost lookup are possible. Call and User cost defaults are defined by the CostVerbatimPhone statement. This type of phone specification is legal everywhere (any type of entry in nodelist, Phone overrides). Note: you might need to change the dots '.' in internet addresses to asterisks '*', in order to avoid that the mailer (e.g. Binkley) translates them. Examples: Internet address: ,6,Mickey,DisneyLand,Mickey_Mouse,#Fantasy*Com,9600,CM,VMP IP address: ,6,Mickey,DisneyLand,Mickey_Mouse,#123*456*789*012,9600,CM,VMP Script name: ,6,Mickey,DisneyLand,Mickey_Mouse,#"Fantasy.Scr",9600,CM,V34 See also the Phone statement. MISCELLANEOUS INFO Multiple Sysops In the case of SysOps of more than one system, all the name/address couples are put in the SysOp Name Lists (Fidouser.lst and Sysop.ndx). If you want to keep one only name/address couple for some multiple SysOps, you can use the "SysDup <AddrLst>" option in the config file: the SysOps who have one of the listed addresses will be present in the SysOp lists with that entry only, even if they have other addresses. Example: Let's suppose that Robert Everywhere has 4 nodes: EveryWhere, Robert 1:456/987.0 EveryWhere, Robert 2:123/457.0 EveryWhere, Robert 2:123/457.8 EveryWhere, Robert 6:321/567.0 By default all the 4 addresses are available in the sysop lists (and some programs allow to choose among them, e.g. the excellent TimEd message editor by Gerard van Essen). If you prefer to keep one address only, you can specify the desired address in a SysDup line in the config file: e.g. "SysDup 2:123/457.8". Redirected Systems Systems that have no valid phone number (Unpublished, on Hold), are redirected, provided you do not exclude redirection using the "NoRedir" config keyword. A redirected system is given the phone number, baud rate, modem type, cost and flags of the preceding coordinator, the Board name is prepended with '-R-'. If you have a session password with the system to be redirected or with the system it should be redirected to, no redirection is done in order to prevent password-mismatch errors in the case the Unpublished/Hold system calls you. Points are never redirected. The systems that have no valid phone number and cannot be redirected take an EMPTY phone number string, so that your mailer does not send unexpected strings to your modem attempting to call these systems, should something appear in your outbound addressed to them. Pay attention: if you want to directly call these null_phone-systems or their coordinators, you have to give them a phone number using the "Phone" statement in the configuration file. V7 Semaphore To avoid collisions between FastLst and V7-able programs, a semaphore for each output nodelist is used. When FastLst needs exclusive access to the V7 files, it creates and keeps open in SH_DENYRW mode a "<NODEX>.BSY" file, where <NODEX> is the base name of the nodelist binary data file. For example, if you use NODEX.DAT, a NODEX.BSY will be created. When a program must access the V7 nodelist, it should create (if not existent) and keep open for reading in SH_DENYWR mode the "<NODEX>.BSY" file. When the access to the V7 files is completed, the program should close and delete "<NODEX>.BSY". This method allows for concurrent access by multiple programs in "read" mode, while granting exclusive access to FastLst. Please note: - When the "KillAfter" statement is used, FastLst needs exclusive access only while deleting/renaming the files. - The <NODEX>.BSY does NOT need to be deleted in case of abnormal termination or power failure since it's considered busy only when open. Example for a program that must read V7: bsyname is the "<NODEX>.BSY" file name; timeout is the timeout in seconds; the file handle is returned on success, -1 on timeout int waitopen (const char *bsyname, int timeout) // -1 on timeout { int ret = -1; int i = 0; do { if (i > 0) sleep (1); if (access (bsyname, F_OK)) { // file not existent int handle = open (bsyname, O_WRONLY | O_CREAT, S_IWRITE); if (handle != -1) close (handle); } ret = sopen (bsyname, O_RDONLY, SH_DENYWR); i ++; } while ((ret == -1) && (i < timeout) && ((errno == EACCES) || (errno == ENOENT))); // sharing violation or file not found return ret; } INSTALLATION 1) There are 4 versions of FastLst: FASTLST.EXE OS/2 FASTLST.EXE NT FASTLST.EXE DOS 32 bit, with DOS4GW DOS Extender FASTL16.EXE DOS 16 bit The OS/2, NT and Dos versions are distributed in separate archives (see Readme.1st). OS/2 and NT: You have only to decide whether to use the "FileMode" statement in the configuration file. If you have enough physical memory available you should use the default "in memory" mode. On the other hand, if the execution of FastLst is excessively slow, you probably are low on memory and you could benefit by the use of the temporary file. You can use the "-t" command line switch to experiment, before enabling the "FileMode" statement in the configuration file, if really necessary. DOS 32 bit: to be used with Dos on a 386sx or better with free extended memory. If you have enough free extended memory, you can use the default and fast "in memory" mode (4MB allow the compilation of approximately 64,000 nodes). If you do not have enough free memory, you can use the "temporary file" mode, provided you have 800KB of free extended memory (see the "-t" command line switch and the "FileMode" configuration statement). If you have no free extended memory at all, you have to use the 16 bit version. DOS 16 bit: to be used with Dos on 8088 or better. The only requirement is 530KB (or better 600KB) of free conventional memory. If you can't free enough conventional memory, you must use a 386sx or better with at least 800KB of free extended memory and use the 32 bit version, which has less stringent conventional memory requirements. 2) Write your FastLst.Cfg. You can find useful examples in the Fast_*.Cfg files and detailed information in the "CFG REFERENCE" section of this documentation. 3) Edit your batch file in order to call FastLst whenever you would like to test for the presence of new list/diff files and process them. If you do not pass a different pathname as a command line parameter, FastLst.Cfg must reside in the current directory. 4) (OS/2): Make sure you have the MSGAPI32.DLL in a directory contained in your LIBPATH. MSGAPI32.DLL can be found in the Squish 1.11 archive (SQSHP111.LZH). If you have little memory and FastLst runs excessively slow, try using the "temporary file" mode (see the "-t" command line switch and the "FileMode" configuration statement). (NT): Make sure you have the MSGAPINT.DLL in a directory contained in your PATH. MSGAPINT.DLL can be found in the Max 3.01 for Windows archive (MAX301N.ZIP). If you have little memory and FastLst runs excessively slow, try using the "temporary file" mode (see the "-t" command line switch and the "FileMode" configuration statement). (DOS32): Make sure you have the DOS4GW.EXE Dos extender (from Rational System Inc.) in your path. The DOS4GW extender requires an XMS or DPMI memory driver installed in your config.sys: e.g. HIMEM.SYS, QEMM (by QuarterDeck Office Systems Inc.). FastLst requires lots of memory to compile long nodelists: you must make sure that the Dos extender can make it available to FastLst. If you need more memory than you physically have, you must use the "temporary file" mode (see the "-t" command line switch and the "FileMode" configuration statement). You could also use the "in memory" mode and activate the DOS4GW virtual memory mode, but it is not recommended since it could unacceptably slow down the compilation process. Anyway, if you really want to test it on your own, the virtual mode is activated using the DOS4GVM environment variable (e.g. for 16MB virtual allocation size: SET DOS4GVM=VirtualSize#16384). (DOS16): Make sure you have at least 530KB of free (conventional) memory. You should use small message buffers (MsgSize statement) in order to avoid out of memory conditions during the merge-sort process. (DOS): Please note that FastLst tells you (on screen, in the logs, in the report message) how much memory remains after compilation, so that you can realize when you are running in marginal conditions and consequently adjust your configuration before you run out of memory. During the merge-sort process (16 bit and 32 bit in "temporary file" mode) FastLst automatically uses the maximum number of ways (up to 8) allowed by the available memory. The minimum number of ways is obviously 2. So, if FastLst reports a "2 Way Merging", you know that the final "MemFree <bytes>" is the actual margin for the "out of memory" error. On the other hand, if FastLst reports "3 Way Merging" or more, you know that there is an additional margin for the "out of memory", approximately 60KB for each way above the minimum of 2. COMMAND LINE SWITCHES -c<cfg> Use <cfg> configuration file instead of FASTLST.CFG. -f Force compilation even if no new list/diff has been detected. -i Ignore FastLst.Dat: run as if it were the first time. All nodelists compiled, all exports executed. -p Prepare: Unarc new lists and diffs, Apply diffs and Arc new nodelists, do not compile nodelists. -r When applying a diff, FastLst usually deletes the newly generated source nodelist file if a CRC error is detected. With this switch the new nodelist is _not_ deleted, so that it will be processed anyway. When compiling a list, FastLst usually aborts the compilation of the current output nodelist if a CRC error is detected. With this switch the current output nodelist will be entirely compiled anyway. -t Toggles between "temporary file" mode and "in memory" mode. If "FileMode" is enabled in the configuration file, it switches to "in memory" mode and vice versa. -h or -? for help ERRORLEVELS 0 - Normal termination, something compiled 1 - Help requested 2 - File Open error 3 - Abnormal termination 4 - Disk Full 5 - Can't find config file 6 - Configuration error 7 - Out of memory 8 - Read error while applying diff 9 - CRC error (applying diff) 10 - CRC error (compiling list) 11 - User Break 12 - Cannot rename temporary output nodelist files 13 - Cannot open source nodelist file 14 - Timeout waiting on V7 semaphore 15 - Too many nodelists in inbound directories 16 - Nothing found after unarchiving a fixed-name nodelist 100 - Normal termination, nothing compiled 250 - MsgApi: Init Error 251 - MsgApi: Area Open Error 252 - MsgApi: Area Lock Error 253 - MsgApi: Area Close Error CONFIGURATION Before analyzing the cfg keywords in detail, let's introduce the overall mechanism that is at the basis of FastLst's configuration. If you are converting from a different nodelist compiler, please forget the old configuration and start from scratch. FastLst.Cfg is divided into several logical blocks and the sequence of the various statements is essential: you cannot just put keywords somewhere in the config file; they must be arranged in the correct order. At first, this characteristic of FastLst's configuration might appear complex to understand, but, as soon as you will grasp its logic, you will understand that it's really easy to write a correct configuration file and you will appreciate its extraordinary flexibility. The first block of configuration is the "Global" one. The verbs in this block refer to the compilation of all the nodelists. Then there are one or more "Output Blocks": each output block refers to the compilation of a single V7 nodelist (e.g. NODEX.*). Each "Output Block" has a "Output section" (with statements regarding the compilation of the whole V7 list) and one or more "Input blocks" containing the verbs that describe how to compile each of the source nodelists. The first "Output Block" can optionally be of a special kind: a "NoCompile" block, whose "Input Blocks" describe nodelists that must be maintained (e.g. diffs applied) but not compiled to any V7 list. Some statements can be used in blocks of a particular type only, others can be used in many different places depending on what nodelists you want to be affected. As a rule of thumb, you can use each statement anywhere it seems to be logically acceptable. If you feel frightened by such abstract considerations, please take a look at the example Fast*.Cfg files, so that you can quickly realize it's not that difficult. To write your own configuration file you should start modifying the example one that is more adequate to your needs. Now, let's consider all the verbs that are allowed in FastLst's configuration. CFG REFERENCE - Items between square brackets (e.g. [<item>]) are optional. - The names of the various Keywords are NOT case sensitive. - When a directory path is required, the trailing backslash '\' is optional. - The ';' character starts comments: any character following the ';' is ignored, unless inside quoted strings. - The maximum length of configuration lines is 254 characters, so don't go further (you can always split address lists into smaller ones). - In the OS/2 version, any file specification can be a legal OS/2 name, between double quotes if necessary. Please, note that the order of the configuration statements follows some logical rule. In order not to create confusion in the .cfg file and not to break some _necessary_ order relation, please follow the scheme proposed in the example Fast*.CFG files and in this reference documentation. Please, be aware that the generation of text files (FidoTxt, FidoPrn, FidoUserLst verbs) and the use of lots of options and overrides can slow down the compilation process: use only the options/overrides that you really need if you mind compilation time. G L O B A L The following verbs can be used in the Global section of FastLst.cfg. Some of them can be used in other places also, so they are divided into separate sections. G L O B A L Section A The following configuration verbs can be used in the GLOBAL section of FastLst.Cfg. RegKey <RegKey> Registered Users only: <RegKey> is the registration key and it is NOT case sensitive. Example: RegKey dfhyuwru6274623 Priority <type> [<level>] Changes the execution priority of the FastLst process (OS/2 only). Ignored by NT and DOS versions. <type> is one of: Idle Regular High <level> is an integer in the range 0...31 and defaults to 0. If you do not use this statement, FastLst will run at the default priority, which normally is "Regular 0". Examples: Priority High 31 Gives Fastlst the highest priority for "non time-critical" processes. It will run fast even if it is in the background and other processes are active. Priority Idle Gives FastLst the lowest priority, so that it loads the system as minimally as possible. It will run significantly slower, especially if in the background or when other CPU intensive processes are in execution. FileMode Enables the temporary file mode (32 bit only). Ignored by the 16 bit versions. Makes Fastlst use temporary files, so that the memory requirements remain low and independent of the size of the compiled nodelists. If you do not use this statement, FastLst (32 bit) uses memory for storing the data it needs for building the V7 indexes. If you have not enough free memory, the DOS 32 version will return an "out of memory" error while the OS/2 one will run slowly. StatusLog <LogFile> <LogFile> is the name of the file where all the operations performed by FastLst will be logged, following the "Binkley Style". In multitasking environments, please be sure to use a file that cannot be used by other processes at the same time. For example: if (in your system) FastLst can be executed while Binkley is running, please use different log files. Should you not want the log file, you can comment this keyword out. Example: StatusLog d:\bbs\log\FastLst.log CompressCfg <compress_cfg> This is a "Squish style" compress definition file. In the case you are using a case-sensitive de/compression program (e.g. OS/2 ZIP/UNZIP), please make sure to use the correct switches in <compress_cfg> and/or the correct case (Lower/Upper) in <NodeList> and <NodeDiff> specifications. You can find the suggested <compress_cfg> in the example Compress.Cfg file included in the FastLst pack. If you are already using Squish and/or Maximus, you can just specify the name of their compress.cfg (but do check that they indicate the necessary switches to avoid case sensitiveness during extraction). Refer to the "Compress Definition File" section at the end of this document for the syntax of <compress_cfg>. TmpPath <TmpDir> Specifies a directory to be used for temporary files. If TmpPath is not used, FastLst will use the path specified in the TEMP or TMP environment variables or (if they are not defined) the current directory. FastLst will not damage any file already present in the temporary directory. The 16 bit versions always use temporary files, while the 32 bit ones work in memory unless the "FileMode" statement or -t command line switch are used. InputPath <NodeDir> Specifies the default path for input files (source nodelists/nodediffs). You can override it by using a full pathname in input-file specifications. Created if not existing. Example: InputPath d:\bbs\nodelist\ ArcPath <ArcNodeDir> Specifies the default path for Archived nodelist files. It usually points to the file area where your TIC processor moves the inbound nodelist archives. You can override it by using a full pathname in Archived-file specifications. Example: ArcPath d:\bbs\file\nodelist\ ArcDate Write|Creation (OS/2) Selects the date to be used for computing the age of fixed-name archived nodelist files. This setting is useful for HPFS (which has separate Write and Creation dates) and ignored for FAT. If not specified, "Creation" is assumed. Attention: in order to avoid problems in the case the date has been corrupted during the transfer of the file, it is best to choose the same date that your mailer sets as "receive/upload" date or that is touched by your TIC processor. Examples: ArcDate Write ; Use the Write date ArcDate Creation ; same as default MultiLineDesc <nnn> [<c>] By default, files.bbs description must be on a single line; this statement enables Multi-Line support. <nnn> is the number of spaces that must precede the continuation lines. <c> is the continuation character. If <c> is NOT specified, it is assumed that the continuation lines must be preceded by <nnn> spaces. If <c> IS specified, it is assumed that the continuation lines must be preceded by <nnn> spaces, the <c> character and one more space. For example, to have the 2nd and following description lines in files.bbs start at the 32nd column, use: MultiLineDesc 31 A description in files.bbs would be like: Test.Zip This is the first description line this is the 2nd line this is the 3rd line ^ ^^ 1 31 32 To have the continuation lines preceded by a '|' character, use: MultiLineDesc 29 | A description in files.bbs would be like: Test.Zip This is the first description line | this is the 2nd line | this is the 3rd line ^ ^ ^ 1 29 32 KillAfter Old V7 files are killed after the new ones have been successfully written. The new V7 files are written to temporary names, then the old ones are killed and the new ones renamed (and FastLst retries for 15s in case of error, to be multitasking smart). Thus you will always have a valid V7 nodelist, even in the case of a compilation error and consequent compile abortion. Besides, your multitasking system can continue operations while FastLst is working. On the other hand you might need some more spare disk space to hold the old and new files during compilation. KillSource Tells FastLst to kill all uncompressed nodelists (that are also available in archived format) before terminating. Please note that FastLst deletes a source nodelist only if the ArcList statement is defined. Besides, when the NodeDiff statement is used, an ArcMethod must be defined to allow the deletion of the nodelist. When no NodeDiff is defined, FastLst assumes that the uncompressed NodeList has been extracted from a corresponding archive. BeforeKillSource <command> This statement is used to invoke a command of your choice before the source nodelists are killed, upon FastLst completion. <command> is executed even if "KillSource" is not used. It is a means of invoking a command before FastLst ends. The "NeededBeforeKill" verb must be used to specify the NodeLists needed by this command: if one of these nodelists is found new, then this command is invoked after decompressing all the nodelists that have the "NeededBeforeKill" attribute (and have not been decompressed yet). IMPORTANT: <command> is executed ONLY if some nodelist affected by a "NeededBeforeKill" verb has been detected as new. <command> can be any command that is valid for the command interpreter specified in your COMSPEC environment variable. If <command> invokes an executable file, it is loaded and executed directly; otherwise your command interpreter is invoked, so that you can execute a batch file or any other valid command. No variable parameters are supported. Dash2Comma Change dashes to commas in the phone numbers. Useful for people that are still connected to ancient "rotary pulse" electromechanic telephone exchanges. NoDash Remove dashes from the phone numbers. NoReport Do not output nodelist statistics NoRedir Nodes that do not have a valid phone number (Hold, Unpublished) are usually redirected to their coordinators. When this verb is used, redirection does not take place and the node is given an empty phone number, so that you never call a system different from that you think you are calling. Please note that (even with no NoRedir verb): - points are never redirected; - if you have a password with a system or its coordinator, this node will never be redirected. V7BugFix Circumvents a bug with V7 nodelist in Binkley 2.50 (and perhaps in many other programs whose V7 search function was inspired by Binkley's sources) that can sometimes hide segments of V7 nodelist. Binkley 2.60 is OK, but some other programs may not: if you are unsure, keep this keyword active. MsgLogArea <path> [-$] Some information about the compilation can be reported to a fido or squish message area. <path> indicates a message area for reporting compilation logs. -$ specifies that the area is in Squish format; otherwise it is assumed to be *.MSG. The "MsgLog" statement (see Global Section C) can be used to add some information that is not reported by default. Examples: MsgLogArea \bbs\mail\net -$ MsgLogArea \bbs\mail\net\ MsgRemArea <path> [-$] The comments found in the compiled nodelists can be selectively reported to a fido or squish message area. <path> indicates a message area for reporting compilation logs. -$ specifies that the area is in Squish format; otherwise it is assumed to be *.MSG. The "MsgRem" statement (see Global Section C) MUST be used to specify which types of comments you want to be reported. Please note that no comments are reported if no "MsgRem" statement is used. Examples: MsgRemArea \bbs\mail\net -$ MsgRemArea \bbs\mail\net\ MsgFromNode <address> MsgToNode <address> Specify the addresses for the created messages. <address> is a 4D address. Example: MsgFromNode 2:332/504 MsgToNode 2:332/504.1 MsgTo <name> Specifies the name of the addressee of the created messages. Example: MsgTo Alberto Pasquale MsgAttr <attributes> Specifies the attributes for the created messages. <attributes> can be a (not case sensitive) combination of: P : Private K : Kill/Sent C : Crash H : Hold D : Direct (crash + hold in squish messages) Examples: MsgAttr P MsgAttr pk MsgSize <bytes> Specifies the maximum size of a single message: after this length, the message is split. Defaults to 7K, greater values are recommended, so that the message is not divided into too many parts. 16 bit: maximum 65535 32 bit: no maximum (4294967295) WARNING: Dos 16 version has stringent problems of memory, so you should be happy with the 7K default. Example: MsgSize 60000 CostNullPhone <Cost> [<UCost>] Specifies the costs to be assigned to nodes with empty (unpublished, etc.) phone number. <Cost> is the "call cost" . <UCost> is the "user cost" (fee for netmail messages). <UCost> defaults to <Cost>. If CostNullPhone is not used, <Cost> defaults to 65535 and <UCost> to 0. Example: CostNullPhone 1000 0 Note: Some programs might have bugs that cause problems dealing with high costs (such as the default 65535). Should you experience problems with entries that have a "NullPhone", try setting a lower cost e.g. "CostNullPhone 900 0". CostVerbatimPhone <Cost> [<UCost>] Specifies the costs to be assigned to nodes that have a verbatim phone specification. <Cost> is the "call cost" . <UCost> is the "user cost" (fee for netmail messages). <UCost> defaults to <Cost>. If the statement is not used, both costs default to 0. Example: CostVerbatimPhone 10 0 Cost and Dial translation Tables You can specify costs and dial translations in separate tables or in a combined "Dial and Cost" table. Note: These tables are not used against "Verbatim Phones". To Europeans: you usually will prefer separate tables, since the dial translations are quite simple. This way you can keep the (very short) dial table always fixed, while adjusting the cost table to your needs. Anyway, if you do not use long cost tables, you could find the combined dial & cost table convenient. To Americans: you have quite long dial tables since you need to list all the "local exchange" codes. So you usually don't like duplicating this list for assigning the costs: you will appreciate the combined table. The custom definition for the list of local exchanges should also speed up the compilation of all the nodelist entries that are out of your Area Code. Cost Table This is the Cost Table to be used in association with a separate Dial table, that must follow. It begins with "Cost" and ends with the "End" keyword. Each entry in the cost table has the following format: <PartPhone> <Cost> [<UCost> [<StripFlags>]] <PartPhone> is a partial phone number to be matched with the initial part of nodelist entries. The dashes are ignored. The <PartPhone> of the last entry must be a single dash "-", to mean that all the remaining numbers will take the costs specified there. <Cost> is the cost field of the compiled nodelist (0->65535). It represents the "Call Cost" for any node whose phone number begins with <PartPhone> in the source nodelist (before dialing translations). <UCost> is an optional "User Cost" (fee for a netmail message on BBS). If it is not used, it's taken equal to <Cost>. If you would like your users to be able to send netmail messages from the BBS with no need for "credits", you should set <UCost> to 0. <StripFlags> is an optional list of nodelist flags (not case sensitive) to be stripped from entries of nodes whose phone number begins with <PartPhone>. Thus you can strip some modem-type flags (V32, HST, ZYX, ...) when calling into critical areas. This parameter can only be used after <UCost>. When searching for "PartPhone", the first matching entry is applied: in the case of entries with an initial part in common, you have to specify them in sequence from the longest to the shortest. If no match is possible, the last line specifies the default (thereby international) <Cost> and optional <UCost> and <StripFlags>. WARNING: This table CANNOT be left totally empty: it must contain at least the default entry "- <Cost>". Example 1: The following example assigns <UCost> = 0 to all, <Cost> = 0 to local systems, 60 to domestic and 300 to international ones. Cost 39-59 0 0 ; local 39- 60 0 ; domestic long distance - 300 0 ; international End Example 2: Now let's strip some flags from some areas. Cost 1-703 300 0 HST,H14,H16 ; strip HST 45-123 300 0 V32,V32b ; strip V32 and V32b 39-59 0 0 ; local 39- 60 0 ; long distance - 300 0 ; international End Example 3: Now let's differentiate between urban and district calls and between Continental and Intercontinental calls; call and user cost considered equal. Cost 39-59-2 5 ; urban 39-59-3 5 ; urban 39-59-4 5 ; urban 39-59-81 5 ; urban 39-59-82 5 ; urban 39-59 30 ; district 39 60 ; domestic long distance 43 100 ; Austria 32 100 ; Belgium 45 100 ; Denmark 33 100 ; France 49 100 ; Germany 44 100 ; UK 31 100 ; Netherlands 34 100 ; Spain 46 100 ; Sweden 41 100 ; Switzerland 1 200 ; USA/Canada - 300 ; others End Example 4: Minimal cost table for sysops that do not want to use the nodelist cost fields. Cost - 0 End Dial Table This table can be used either as a simple Dial Table in association with a separate preceding Cost Table or as a combined Cost/Dial Table (in which case the Cost Table must NOT exist). It begins with "Dial" and ends with the "End" keyword. Each entry has the following format: <PartPhone> [<PreSuf>] or <PartPhone> <PreSuf> [<Cost> [<UCost> [<StripFlags>]]] The first form is for the separate Dial Table, the second one is for the combined Cost/Dial Table. The following two keywords allow to specify groups of exchanges that must be handled by a certain dial table entry: LocalValues <PartPhone> <PreSuf> [<Cost> [<UCost> [<StripFlags>]]] This keyword is provided for clarity only: it is taken just as a normal specification with no heading "LocalValues". LocalExchanges <num> ... Lists all the exchanges that must be handled as per the preceding dial translation entry (which may be preceded by the "LocalValues" keyword for clarity). Please remember that the line length is limited to 254 characters. You can use multiple "LocalExchanges <num> ..." statements if you like short lines or need more than 254 characters. Please note that all numbers that (after <PartPhone> stripping) begin with <num> are considered local. For example, if 220, 221, 222, 223, 224, 225, 226, 227, 228, 229 are all local exchanges, you can indicate 22 to include them all. The use of these two statements in the place of a long list of normal table lines (one for each local exchange) should also speed up the processing of all the nodelist entries that are not in your area code. <PartPhone> is a partial phone number to be matched with the initial part of nodelist entries. The dashes are ignored. The <PartPhone> of the last entry must be a single dash "-", to mean that all the remaining numbers will take the parameters specified there. <PreSuf> can be one of: a: <Prefix> b: /<Suffix> c: <Prefix>/<Suffix> d: / <PartPhone> is stripped from numbers beginning with it, then <PreSuf> is used to prepend/append the specified strings to the remainder. Case a: <Prefix> is prepended. e.g.: "39- 0" strips "39-" and adds "0" at the beginning of the number. Case b: <Suffix> is appended. e.g.: "39-59- /!" strips "39-59-" and adds "!" at the end of the number. Case c: <Prefix> is prepended and <Suffix> appended. e.g.: "39- 0/!" strips "39-", adds "0" at the beginning and "!" at the end. Case d: Nothing is prepended nor appended. e.g.: "/" The slash is needed to allow the correct interpretation of the subsequent fields, in the case of combined Dial/Cost table. No spaces are allowed between prefix, suffix and the separating slash. The remaining fields are used as described in the "Cost Table" section. If the <Cost> field is omitted, it is taken equal to 65535. When searching for "PartPhone", the first matching entry is applied: in the case of entries with an initial part in common, you have to specify them in sequence from the longest to the shortest. If no match is possible, the last line specifies the default (thereby international) parameters. WARNING: This table CANNOT be left totally empty: it must contain at least the default entry "- [<PreSuf>]". Example 1 (separate tables, European viewpoint): Dial 39-59- ; local 39- 0 ; domestic - 00 ; international End "39-59": Country and District codes are stripped for local calls. "39-" : Country code is stripped and replaced by "0" (domestic long distance code) for domestic calls. Others: "00" (International access code) is prepended. Example 2a/b/c (combined tables, North American viewpoint): Since this one is a bit more complex, let's make clear some points. There are some groups of phone numbers: 1 - Local numbers. These are usually "toll free" numbers. As a rule of thumb, you must only dial the local number, stripping the Country Code "1" and the Area Code. In the case you live at the crossroads of two or more area codes, it is possible that you have local numbers in area codes that you must _not_ strip. 2 - Area Code numbers. They have your same Area Code but they are long distance. As far as I know, there are many different situations regarding the need to dial the long distance access code "1" and the Area Code. In any case you usually want to differentiate costs. The Country "1" and Area Codes must be stripped and replaced by: (a) the long distance access code "1" (b) the long distance access code "1" + the Area Code (c) nothing In case (b) the number really remains untouched. Even if the Country Code for USA/Canada "1" is numerically equal to your long distance acces code, they are conceptually quite different, and so they will be treated. 3 - Domestic numbers. USA/Canada numbers, with a leading "1", that is the international Country Code for USA and Canada. They must be left untouched, since the american long distance access code is equal to the international Country Code for North America. 4 - International numbers. These are numbers out of USA and Canada. They must be prefixed by "011", that is the international access code. And now let's see how to achieve our goal using FastLst's Dial Table. Example 2a: Let's suppose: - we are in Area Code 414 - the 414 must be stripped for LD calls - the local exchanges are 231, 232, 233, 235, 236, 424 Dial ; strip 1-414- from local numbers, do not add ; a prefix, set call and user costs to 0. LocalValues 1-414- / 0 LocalExchanges 231 232 233 235 236 424 ; Remaining "1-414-" numbers are long distance: ; strip the 414 Area Code and assign costs = 25. 1-414- 1- 25 ; Remaining "1-" numbers are Domestic Long Distance. ; Set costs to 50 1- 1- 50 ; Remaining numbers are international. ; Prepend 011 and set call cost to 250 and ; user cost to 500 - 011 250 500 End Example 2b: Let's suppose: - we are in Area Code 604 - the 604 must NOT be stripped for LD calls - the local exchanges are 220 221 222 224 228 230 231 240 241 244 250 251 252 253 254 255 257 258 261 263 264 266 Dial ; strip 1-604- from local numbers, do not add ; a prefix, set call and user costs to 0. LocalValues 1-604- / 0 LocalExchanges 220 221 222 224 228 230 231 240 LocalExchanges 241 244 250 251 252 253 254 255 LocalExchanges 257 258 261 263 264 266 ; Remaining "1-604-" numbers are long distance: ; assign costs = 25. 1-604- 1-604- 25 ; Remaining "1-" numbers are Domestic Long Distance. ; Set costs to 50 1- 1- 50 ; Remaining numbers are international. ; Prepend 011 and set call cost to 250 and ; user cost to 500 - 011 250 500 End Example 2c: Let's suppose: - we are at the crossroads of Area Codes 510, 408, 415 - "1-510" must always be stripped - "1-408" and "1-415" must never be stripped - some exchanges are toll-free, others are not Dial ; strip 1-510- from local numbers, do not add ; a prefix, set call and user costs to 0. LocalValues 1-510- / 0 LocalExchanges 224 225 226 227 247 249 252 264 276 LocalExchanges 278 293 317 353 354 416 417 ; Specify local exchanges for area code 408, ; keep "1-408". LocalValues 1-408- 1-408- 0 LocalExchanges 232 251 254 258 259 262 263 272 276 LocalExchanges 321 324 325 383 428 432 433 434 ; Specify local exchanges for area code 415, ; keep "1-415". LocalValues 1-415- 1-415- 0 LocalExchanges 233 234 321 322 323 324 325 326 327 LocalExchanges 328 329 354 424 462 473 493 ; Remaining "1-510-" numbers are not free, ; the country and area codes are stripped. 1-510- / 25 ; Remaining numbers in area codes 408 and ; 415 are not free, nothing is stripped. 1-408- 1-408- 25 1-415- 1-415- 25 ; Remaining numbers in country code "1" ; are domestic: nothing changed, cost 100. 1- 1- 100 ; Remaining numbers are international. ; Prepend 011 and set cost to 2000. - 011 2000 End Example 3 (combined table, European viewpoint): The following combined table is equivalent to the separate "Example 3" for the Cost Table and Example 1 for the Dial Table. Dial LocalValues 39-59- / 5 LocalExchanges 2 3 4 81 82 ; urban 39-59- / 30 ; district 39- 0 60 ; domestic long distance 43 0043- 100 ; Austria 32 0032- 100 ; Belgium 45 0045- 100 ; Denmark 33 0033- 100 ; France 49 0049- 100 ; Germany 44 0044- 100 ; UK 31 0031- 100 ; Netherlands 34 0034- 100 ; Spain 46 0046- 100 ; Sweden 41 0041- 100 ; Switzerland 1 001- 200 ; USA/Canada - 00 300 ; others End Example 4: Minimal table for sysops that make dial translations somewhere else. Dial - End Modem Type Table If you have a modem that does not need different dial strings for different protocol connections, you can skip this section. For Example a Zyxel modem usually needs one only dial string for any type of connection (unless you do not use "Multi-Auto" mode). Instead, if you need different dial strings, you can use the Modem_Type field in conjunction with some front-end feature that allows to specify different dial strings for different modem types ("ModemTrans" in Binkley). Each entry in the TypeDef table has the following format: <Flag> <Value> <Flag> is a Nodelist flag (max 9 chars), <Value> is a number 0->255. The nodelist flags of each node are searched for <Flag>. The <Flag> must match completely a nodelist flag: if <Flag> is V32 and the nodelist flag is V32B, it's not a match. The search is not case sensitive. If <Flag> is found, the corresponding ModemType field is set to <Value>, otherwise the next <Flag> is searched for. The ModemType field of the compiled nodelist will be determined by the first match only: If you define HST before V32, a node with both V32 and HST will have a HST modem type. The following examples are valid for Binkley's new "exact match" modemtrans style only (for use of old bitwise style, see below). WARNING: Please note that when you do not want Binkley to dial out for certain modem types, you need to specify an empty dial string in the modemtrans statement. Due to a bug of Binkley, you must also avoid comments on those lines ! Example 1: for USR Courier Dual Standard V.Everything: TypeDef V34 1 ; first choice VFC 2 V32T 3 H16 4 V32B 5 ZYX 5 ; ZYX implies V32B Z19 5 Z16 5 H14 6 V32 7 HST 8 End In Binkley.Cfg you can use: ModemTrans 0 ATB0D/ ; default ModemTrans 1 ATB0D/ ; V34 ModemTrans 2 ATB0D/ ; VFC ModemTrans 3 ATB0D/ ; V32T ModemTrans 4 ATB1D/ ; H16 ModemTrans 5 ATB0D/ ; V32B, ZYX ModemTrans 6 ATB1D/ ; H14 ModemTrans 7 ATB0D/ ; V32 ModemTrans 8 ATB1D/ ; HST Example 2: for 2 lines, ISDN and USR DS V.Everything: The UISDNA, UISDNB, UISDNC entries are required only until the incorrect ",UISDNA," type entries are changed to the correct ",U,ISDNA," style or the ISDN flags become standard flags (not user defined as they currently are). TypeDef ISDNC 1 UISDNC 1 ISDNB 2 UISDNB 2 ISDNA 3 UISDNA 3 V34 4 VFC 5 V32T 6 H16 7 V32B 8 ZYX 8 ; ZYX implies V32B Z19 8 Z16 8 H14 9 V32 10 HST 11 End In Binkley.Cfg, ISDN line, you can use: ModemTrans 0 ModemTrans 1 <ISDNC_Dial_String> ; ISDNC ModemTrans 2 <ISDNB_Dial_String> ; ISDNB ModemTrans 3 <ISDNA_Dial_String> ; ISDNA ModemTrans 4 ModemTrans 5 ModemTrans 6 ModemTrans 7 ModemTrans 8 ModemTrans 9 ModemTrans 10 ModemTrans 11 In Binkley.Cfg, USR line, you can use: ModemTrans 0 ATB0D/ ; default (2400) ModemTrans 1 ModemTrans 2 ModemTrans 3 ModemTrans 4 ATB0D/ ; V34 ModemTrans 5 ATB0D/ ; VFC ModemTrans 6 ATB0D/ ; V32T ModemTrans 7 ATB1D/ ; H16 ModemTrans 8 ATB0D/ ; V32B, ZYX ModemTrans 9 ATB1D/ ; H14 ModemTrans 10 ATB0D/ ; V32 ModemTrans 11 ATB1D/ ; HST Example 3: for 2 lines, ISDN and analog, when you want both lines enabled to dial systems that have both ISDN and analog access. You usually should not want such a behaviour (why calling a ISDN system the analog way ?); however, if you really know what you are doing, here is a viable trick. You have to use the "BitType" mode (see below): BitType Typedef ISDNA 1 ISDNB 1 ISDNC 1 V34 2 VFC 2 V32T 2 V32B 2 V32 2 End In Binkley.Cfg, ISDN line, you can use: ModemTrans 0 ModemTrans 1 <ISDN_Dial_String> ; ISDN only ModemTrans 2 ModemTrans 3 <ISDN_Dial_String> ; ISDN + analog In Binkley.Cfg, analog line, you can use: ModemTrans 0 ATDT ; 2400 ModemTrans 1 ModemTrans 2 ATDT ; analog only ModemTrans 3 ATDT ; analog + ISDN For compatibility with old Binkleys and derived systems, the following statement is provided: BitType If you need old-style bit-oriented modem type, you must enable this verb. In this case the "TypeDef" works differently: - <Value> should be a power of 2 (1, 2, 4, 8, 16, 32, 64, 128). - The ModemType will be determined by ORing all the <Value>s corresponding to <Flag>s that found a match in the nodelist flags. - The dial string used by your front-end will be determined by the order of their specifications (the first ModemTrans that has some bit in common with the modem type will be used). Example for USR Courier Dual Standard V.Everything: BitType TypeDef V34 1 ; first choice VFC 2 V32T 4 H16 8 V32B 16 ZYX 16 ; ZYX implies V32B Z19 16 Z16 16 H14 32 V32 64 HST 128 End In Binkley.Cfg you can use: ModemTrans 0 ATB0D/ ; default ModemTrans 1 ATB0D/ ; V34 ModemTrans 2 ATB0D/ ; VFC ModemTrans 4 ATB0D/ ; V32T ModemTrans 8 ATB1D/ ; H16 ModemTrans 16 ATB0D/ ; V32B, ZYX ModemTrans 32 ATB1D/ ; H14 ModemTrans 64 ATB0D/ ; V32 ModemTrans 128 ATB1D/ ; HST User Flags Table This is an optional table used to handle the "user defined" bits in the Flags word of the compiled nodelist entry. These bits are named 5,6,7,8,9,A,B,D,E,F where bit 5 is the 6th bit and F is the 16th bit of the word. Using this table, you can associate (source) nodelist flags to user defined bits in the output binary nodelist. The table is delimited by the "FlagDef" and "End" keywords; each line is in the form "<sFlag> <bFlags>" where <sFlag> is a flag (max 9 chars) to be looked for in the source nodelists while <bFlags> represents one or more user-defined bits in the output Flags word. Example: FlagDef V42B AB ; V42B -> user bits A and B V32B DE ; V32B -> user bits D and E End To add further flags on a node by node basis, please use the "Flags <Addr> <Flags>" statement. To remove flags, please specify the source flags via the "NodeFlags <Addr> <NodeFlags>" statement. G L O B A L Section B The statements in this section affect the processing of all the output blocks and thereby of all the input nodelists. These statements can also be used in the "OUTPUT" section of an OUTPUT block or inside an INPUT block, in which case they affect the compilation of the relevant block only. In the case you use a verb that has already been used in a "higher level" block, it will behave as a local override. NeededBeforeKill Tells FastLst that the affected NodeList(s) are needed by the command run via the "BeforeKillSource" statement. The "BeforeKillSource" verb allows you to run a command (executable or batch file) after the compilation has completed, just before FastLst ends and (if "KillSource" is used) deletes the source files that are also present in archived form. The lists affected by "NeededBeforeKill" are extracted, if not already present, before the "BeforeKillSource" command is executed. ArcMethod <meth>[,<f>] ... Tells FastLst that it must make sure that all new nodelists are archived using the specified methods. The original archive is NOT killed. Obviously, a new nodelist is not rearchived to its original method. <meth> is the name of an archiver defined in compress.cfg. <f> is the optional specification of the letter to be used for the variable archive extension. If not specified, it is assumed equal to the first letter of the defaults extension for this archiver. Multiple ArcMethod statements are allowed. Example 1: ArcMethod ZIP LH,H NodeList.Z48 arrives: it is archived to NodeList.H48 also, using the LH archiver. Example 2: ArcMethod ZIP LH NodeDiff.Z48 arrives: the resulting nodelist is archived to NodeList.Z48 using the ZIP archiver and to NodeList.L48 using LH. ArcDiffMethod <meth>[,<f>] ... Tells FastLst that it must make sure that all new nodediffs are archived using the specified methods. The original archive is NOT killed. Obviously, a new nodediff is not rearchived to its original method. <meth> is the name of an archiver defined in compress.cfg. <f> is the optional specification of the letter to be used for the variable archive extension. If not specified, it is assumed equal to the first letter of the defaults extension for this archiver. Multiple ArcDiffMethod statements are allowed. Example: ArcDiffMethod ZIP LH,H NodeDiff.Z48 arrives: it is archived to NodeDiff.H48 also, using the LH archiver. EXTERNAL COMMANDS The following verbs allow to invoke external commands. <command> can be any legal command-line command: it can be the name of an executable file, a batch file or any command that can be understood by your command-line interpreter (OS/2's CMD, 4OS2, etc.). If <command> does not directly invoke an executable file, FastLst automatically invokes your default command line interpreter (as specified by the COMSPEC environment variable). Archive Related Commands The following verbs share the same syntax: Two parameters are allowed in <command>: %a is translated to the full pathname of the archive file. %f is translated to the name of the file to be added or extracted (no path). <command> is run from the path where %f belongs. BeforeArcList <command> Command to be run before archiving a nodelist. AfterArcList <command> Command to be run after archiving a nodelist. BeforeUnArcList <command> Command to be run before extracting a nodelist. AfterUnArcList <command> Command to be run after extracting a nodelist. BeforeArcDiff <command> Command to be run before archiving a nodediff. AfterArcDiff <command> Command to be run after archiving a nodediff. BeforeUnArcDiff <command> Command to be run before extracting a nodediff. AfterUnArcDiff <command> Command to be run after extracting a nodediff. Example To hatch the new nodelist (note that you probably need to specify the location of the config file since the command is executed from the directory where %f resides): AfterArcList Hatch %a NODELIST "New NodeList" NodeDiff Related Commands The following verbs accept different parameters: %l is translated to the full pathname of the nodelist. %d is translated to the full pathname of the nodediff. <command> is run from the current directory. BeforeEdit <command> Command to be run before applying a nodediff. AfterEdit <command> Command to be run after applying a nodediff. Only %l can be used. G L O B A L Section C The statements in this section affect the processing of all the output blocks and thereby of all the input nodelists. These statements can also be used in the "OUTPUT" section of an OUTPUT block (except for the "NoCompile" one) or inside an INPUT block, in which case they affect the compilation of the relevant block only. In the case you use a verb that has already been used in a "higher level" block, it will behave as a local override. MsgRem [<string>] If MsgRemArea is used, FastLst reports the following comments: No MsgRem statement: none; MsgRem with no <string>: all; MsgRem with <string>: only the comments that begin with ";<l> " where <l> is one of the characters in <string>. The ";" character in <string> means that the comments beginning with "; " or ";<word>" can be reported. Common types of comment lines: ;S This is a comment for SysOps ;U This is a comment for users ;F This comment should appear in formatted Fido lists ;A This is a comment of general interest ;E This comment is an error message Example: "MsgRem SE" Only comments destined to SysOps and Error messages are reported (lines beginning with ";S " and ";E "). MsgLog [NullPhone] [Redirected] [Points] Some common situations (not really errors) are not reported to MsgLogArea by default: if you want FastLst to report them anyway, you can use this statement, but be aware that very long reports could come out. "NullPhone": systems with empty phone string are logged. "Redirected": systems redirected to their coordinators are logged (Hold, unpublished). "Points": points with empty phone string are logged; be aware that most pointlists contain unpublished (thereby with empty phone) points. Examples: MsgLog Redirected MsgLog Redirected NullPhone GermanPointList Instructs FastLst to consider the affected nodelist as a 3D German style pointlist. Zone 2 is assumed, if not explicitly specified in the "NodeList" statement. This verb is usually used inside an Input Block, so that it affects that nodelist only. WARNING: Be aware that using this statement in the global section or in an Output block affects all the involved nodelists ! Example Input Block: NodeList Points24.??? GermanPointList Nodediff Pr24Diff.??? ArcList Points24.??? 1 ArcDiff Pr24Diff.??? 5 ArcListDesc R24 PointList for day %d (%D), %a format ArcDiffDesc R24 PointDiff for day %d (%D), %a format BeforeCompile <command> Command to be run before compiling the affected nodelist. This statement follows the same rules explained in "External Commands" in section B. The %l parameter is translated to the full pathname of the nodelist. <command> is run from the current directory. AfterCompile <command> Command to be run after compiling the affected nodelist. This statement follows the same rules explained in "External Commands" in section B. The %l parameter is translated to the full pathname of the nodelist. <command> is run from the current directory. SysOpLst Output SysOp data from all the input nodelists to the output list (FidoUser.Lst) and/or index (Sysop.ndx). Example: SysOpLst FidoTxt [<FidoTxt>] Generate an 80 Column Text List of nodes. Nodes included via the "Node,..." method and points are excluded. <FidoTxt> optionally specifies an output file name, which defaults to "NodeList.Txt". If the same file name has already been used for other nodelists, the output is appended. Example: FidoTxt FidoPrn [<FidoPrn>] Generate a 132 Column Text List of nodes. Nodes included via the "Node,..." method and points are excluded. <FidoPrn> optionally specifies an output file name, which defaults to "NodeList.Prn". If the same file name has already been used for other nodelists, the output is appended. Example: FidoPrn IncCoord <CoordLev> The coordinators of the specified and upper levels will be always included, even if excluded by "IncAddr" and "ExcAddr". <CoordLev> can be ZC, RC, NC, HC. Example: IncCoord NC Global Export Section You can use here the statements described in the "Export Global Section" of the "Export Block" (see "Input Block" inside "Output Block"). O U T P U T B L O C K The following verbs define the compilation of a single output binary nodelist. The block begins with a "Output Section", that affects the compilation of all the source (input) nodelists, followed by a sequence of "Input Blocks" that define how to handle each of the source nodelists. The first "output block" can be of a special kind: if the "NoCompile" statement is used instead of "Version7", this block indicates the actions necessary to maintain the specified nodelists, but they are not compiled. Version7 <Path> <Nodex> [<SysopNdx>] Start of a block of config verbs defining the generation of a Version 7 nodelist. You can generate one or more Version 7 nodelists with different names and path for the output files. Each "Version7" statement marks the beginning of a new output-nodelist definition. <Path> is the path where the output binary data and index files are placed. <Nodex> is the file name (no extension) for the data (.DAT) and address-index (.NDX) files. <SysopNdx> is the file name for the sysop-index. When no extension is given, .NDX is assumed if <SysopNdx> is different from <Nodex>, otherwise the .SDX extension is used. If you omit <SysopNdx>, no V7 SysOp-index will be generated. The usual names for V7 files are <Nodex>="NODEX" and <SysopNdx>="SYSOP". All the following verbs, up to the next "Version7" (if any), are related to the preceding "Version7" output files. Examples: ; SysOp Index name Version7 d:\bbs\v7\ NODEX SYSOP ; SYSOP.NDX Version7 d:\bbs\v7\ NODEX NODEX ; NODEX.SDX NoCompile This verb can be used to start the first "Output Block", instead of "Version7". This way the first output block becomes a "NoCompile" block and the indicated nodelists are maintained but not compiled. This is a means for maintaining a NodeList (applying nodediffs, archiving with different archivers etc.) without compiling it. The statements related to nodelist compilation (see Global section C) are obviously illegal in a "NoCompile" block. O U T P U T Section The following verbs affect the compilation of the current output block and must precede the definitions of the input blocks (which start with the Nodelist statement). FidoUserLst [<FidoUserLst>] Generate "fidouser.lst style" text SysOp list. <FidoUserLst> optionally specifies an output file name, which defaults to "FidoUser.Lst". Different output blocks require different names. Example: FidoUserLst SysDup <AddrLst> When a SysOp name is present in various nodes, all the name/address couples are kept in the SysOp lists (fidouser.lst/sysop.ndx). If you want to keep only one address you can use one or more SysDup lines: the SysOps who have the addresses listed in <AddrLst> will be present in the output sysop lists with the specified address only. You can use abbreviated addresses, if you like, provided that the first address of every "SysDup" is complete (FastLst cannot make any assumption for the first item in a list). Example: SysDup 2:332/504 505 336/980 3:25/28.27 Block Specifications You can use here the same statements described in the "Global Section B" and (if this is not a "NoCompile" block) "Global Section C" and "Export Global Section" (see the Export Block below). ADDRESS SPECIFIC STUFF The following verbs define address specific stuff that will affect the compilation of all the source nodelists compiled in the current output block. These statements are illegal in a "NoCompile" block. If you prefer, you can specify this type of information in the "Address Specific Stuff" section of the pertinent input block. WARNING: make sure all addresses have full info (incl. zone). Password <Addr> <Password> Allows to specify <Password> one <Addr> at a time. Version 7 has no limit on password length, however the programs that use it are usually limited to 8 chars. Some (rare) programs have problems with 8 chars and need a maximum of 7 or 6 chars. Example: Password 2:332/504.4 Password PasswordFile <PasswordFile> Allows to include a password file that contains many address/password couples, one per line. In this file you can omit the "Password" keyword. If you like, you can use some "Password" keywords together with one "PasswordFile", however you cannot use more than one "PasswordFile". Please note that the definitions found in this file have effect on the current (Output or Input) block ONLY. FastLst writes to the log file which source or output nodelist is affected by each passwordfile; so, in case of doubts, just check the logs. Example: PasswordFile fidonet.pwd Phone <Addr> [#]<Phone> [<NodeFlags> [<Cost> [<UCost>]]] Allows to override a nodelist phone number and optionally the corresponding "NodeFlags" and costs. If '#' is prefixed, <Phone> is taken verbatim (dial translation not performed, costs default to the values specified with the "CostVerbatimPhone" statement). If '#' is NOT used, <Phone> MUST be in the form used in the source nodelist. The verbatim form may be handy for specifying internet addresses and script names. In the case you want to specify an empty phone, use the '#' alone. <NodeFlags> has the same meaning as in the NodeFlags statement. To specify an overriding empty <NodeFlags>, use a single comma. <Cost> and <UCost> have the same meaning as in the Cost statement. Examples: <Phone> override only: Phone 2:332/501.1 39-59-399999 ; Normal override Phone 2:332/501.0 #499999 ; Verbatim override Phone 3:456/789 # ; Empty phone Phone 1:106/2000 #juge*com ; internet address Phone 1:123/4567 #123*456*789*012 ; IP address Phone 2:245/6789 #"Bob.scr" ; quoted script name <Phone> and <NodeFlags> overrides: Phone 2:332/501.0 #499999 V34,CM ; Flags override Phone 2:332/501.1 39-59-399999 , ; No-flags override <Phone>, <NodeFlags> and <Cost>/<UCost> overrides: Phone 2:332/501 #499999 V34,CM 10 0 ; Different Costs Phone 2:332/502 #mega.com VMP 0 ; <Cost> == <UCost> NodeFlags <Addr> <NodeFlags> Allows to substitute the flags listed in the nodelist entry of <Addr>. If you want to change the CM flag or modem type flags (HST, V32b, ZYX) etc, you can use this verb. Please note that the old flags are lost, so you need to indicate all the necessary flags. Please note that <NodeFlags> might be empty. Example: NodeFlags 2:332/501.0 CM,H16,V32b Flags <Addr> <Flags> The Flags statement allows to set the "user defined" bits in the Flags word of the compiled nodelist entry. These bits are named 5,6,7,8,9,A,B,D,E,F where bit 5 is the 6th bit and F is the 16th bit of the word. These bits are "ORed" with those already set by the "FlagDef" table. If you need to zero some of the bits, please specify the source flags with the "NodeFlags" statement. Example: Flags 2:332/501.0 AB5 ; Set bits 5,A & B. Cost <Addr> <Cost> [<UCost>] <Cost> and <UCost> are in the range 0->65535. Overrides the Cost and User_Cost fields of <Addr> in the compiled nodelist. If no <UCost> is given, it's taken equal to <Cost>. Example: Cost 2:332/501.0 150 SEGMENT SELECTION The following verbs allow to include or exclude selected <NodeList> segments. If you do not use them, the full <NodeList> is compiled. Be aware that the process of checking each address against the list of segments to be included or excluded might slow down the compilation, even if some gain could come from the exclusion of large segments. These statements are obviously illegal in a "NoCompile" block. These statements can be used in an Input block to affect that nodelist only. IncAddr <PartAddrLst> If you want to selectively include nodelist segments, you can use this option: only zones, regions, nets, hubs, nodes, points that are listed in <PartAddrLst> will be present in the output files. You can specify zone, region/net, hub/node and point numbers. Example: IncAddr 1 2:33 2:200/100 3:632 4:801/17 Compiles: zone 1, region 33 of zone 2, hub 100 of net 200 of zone 2, net 632 of zone 3, node 4:801/17 ExcAddr <PartAddrLst> If you want to exclude some segments from the compilation, you can list them in <PartAddrLst>, in the same way as for "IncAddr". You can use either "IncAddr" or "ExcAddr" or both of them to Include only selected segments and exclude sub-segments. Example: ExcAddr 2:332/500 Excludes Hub 500 of net 332 of zone 2. IncSysOp <PartAddrLst> If not used, all the SysOp entries of the compiled segments will be in the output SysOp list/index (if SysOpLst is active). If you want to limit the SysOp entries to selected segments, you can use this verb, listing partial addresses in <PartAddrLst>. SysOps from segments excluded from compilation via "IncAddr" and "ExcAddr" will obviously never be present in the SysOp list/index anyway. Example: IncSysOp 2 Includes only SysOps from zone 2. I N P U T B L O C K The Input Block starts with a "NodeList" statement and continues until the start of the next Input or Output Block (NodeList or Version7 statement respectively) or the end of the configuration file. NodeList <NodeList> [<PartAddr>] Start of a block of config verbs defining the processing of the specified <NodeList> file. You can use many "NodeList" statements to compile several different source nodelists into the same output files specified by the preceding "Version7" statement. Each "NodeList" verb marks the beginning of a new input-nodelist processing-info block. When an address is present in more than one <NodeList> (e.g. you compile both the full nodelist and the faster updated local region or zone segment) only the entry found in the last compiled <NodeList> is put in the indexes. To have the most up-to-date entries in your V7 indexes, please include local segments after the larger list. <NodeList> is the name of the input nodelist. If you don't specify a path, <InputPath> is assumed. If a terminal ".???" is specified, all the files with 3 digits in the place of '???' are examined and that with the latest 3 digit day of the year is chosen for compilation. The optional <PartAddr> is a partial address that must be specified for nodelist segments that do not have full address info. For example, a REGION segment usually starts with the "Region," keyword and does not contain any Zone info: its up to you to tell FastLst which zone we are talking about. Analogously you should provide zone and net info when compiling a Hub segment. The region is assumed equal to the net number of the partial address, the hub equal to the node number. Examples: IMPORTANT: Please note that the following lines represent a list of examples, NOT an example of multiple nodelist compilation. After each "NodeList" verb, you must specify all the statements that affect the compilation of that particular source file. NodeList nodelist.??? ; Fidonet nodelist NodeList region.033 2 ; Region 33 list, zone 2 NodeList region24.??? 2 ; Region 24 list, zone 2 NodeList net.332 2:33 ; Net list, zone 2, region 33 NodeList hub.500 2:332 ; Hub list, zone 2, net 332 NodeList locnode.500 2:332/500 ; Some nodes in zone 2, ; net 332, hub 500 NodeList points.504 2:332/504 ; Points of 2:332/504 ; in "Point," format. NodeList morenode.lst ; Some nodes in the "Node," ; format. No <PartAddr> required ; since the "Node," line gives ; full address info. NodeList ptlist.??? ; Point List in the "Boss," ; format. No <PartAddr> required ; since the "Boss," line gives ; full address info. Input Section The following statements affect the handling of the nodelist specified by the last "NodeList" statement (current Input Block). NodeDiff <NodeDiff> <NodeDiff> is the name of the nodediff file. If you don't specify a path, <InputPath> is assumed. <NodeDiff> must terminate with ".???". FastLst will search for a suitable <NodeDiff>, considering the files that have a 3 digit day of the year in the place of the trailing '???'. Example: NodeDiff NODEDIFF.??? ArcList <ArcList> [<Keep#>] You can specify the name of the archive containing <NodeList>. It is necessary if you use automatic extraction/rearchiving, but it can even be used only to delete old files. <ArcList> is used to extract new nodelists, to compress them using the methods defined in "ArcMethod", to compress the new nodelists after the application of nodediffs. If <ArcList> has a terminating ".???", all the files that have a suitable fixed (.zip, .lzh etc.) or variable (.z10, .z17, .l10, .l17 etc.) extension are considered, taking the digits as the last 2 digits of the day of the year. If you really want to limit search to a specified fixed or variable extension, you can do: "ArcList nodelist.zip", to consider .zip only; "ArcList nodelist.z??", to consider .z?? only. <Keep#> optionally specifies the number of archives to be kept, basing on the day of the year (the modification file date is also used to infer the correct chronological order). If you maintain archives with multiple different extensions (.z??, .l??, etc.) the actual number of files increases, since multiple files with the same day extension count for one. The description associated to the deleted files is removed from FILES.BBS. Example: ArcList nodelist.??? 1 ArcDiff <ArcDiff> [<Keep#>] You can specify the name of the archive containing <NodeDiff>. It is necessary if you use automatic extraction/rearchiving, but it can even be used only to delete old files. <ArcDiff> must terminate with ".???". All the files that have 2 digits in the place of the last 2 '?' are examined, taking the digits as the last 2 digits of the day of the year. If you really want to limit search to a specified extension, you can do: "ArcDiff nodediff.z??", to consider .z?? only. <Keep#> optionally specifies the number of archives to be kept, basing on the day of the year (the modification file date is also used to infer the correct chronological order). In the case of multiple archive extensions, the actual number increases consequently. The description associated to the deleted files is removed from FILES.BBS. Example: ArcDiff nodediff.??? 5 ArcListDesc <Desc> ArcDiffDesc <Desc> You can specify a description to be added to FILES.BBS for the new nodelist and nodediff files created by FastLst. Some parameters are available: %d : the 3 digit day number (0 padded) %a : the archiver name %D : the date, USA format (Feb 10, 1995) %L : the date, Local format Example: ArcListDesc Fido Nodelist for day %d (%D), %a format ArcDiffDesc Fido Nodediff for day %d (%D), %a format Local Specifications You can use here the same statements described in the "Global Section B" and (if we are not in a "NoCompile" block) "Global Section C" and "Export Global Section" (see the Export Block below). ADDRESS SPECIFIC STUFF You can specify here the address specific stuff that is related to the current source nodelist (if not inside a "NoCompile" block). If you have already used the "Output section" for specifying this kind of information, you can skip this section. WARNING: Often you will compile segments of a previously compiled nodelist. For example you could have a "NodeList nodelist.???" block for the world nodelist and then a "NodeList region.033" block for your region's nodelist segment. The majority of entries in the latter will be duplicates of entries already found in the former. However, in the case of duplicates, only the entries found in the last involved "NodeList" block will go to the indexes and be active. This way you can compile the full world nodelist while keeping your segment up-to-date with local segments that get updated faster than the full nodelist. When you have to specify "Address Specific Stuff" for nodes that are present in more than one "NodeList", you must do that in the last involved "NodeList" block (or in the Output Section, of course), otherwise your indications will have no effect. For a list of allowed statements, please see the "Address Specific Stuff" section of the "Output" section above. SEGMENT SELECTION You can use here the same statements described in "Segment Selection" in the Output Section (if not inside a "NoCompile" block). EXPORT Block FastLst can "export" segments of nodelist: e.g. you can export the Region 25 from the world nodelist to a file called Region25.???, where ??? stands for the day of the year. Note that this feature is for exporting segments of nodelist to a dedicated file. To compile segments you should continue using the "Segment Selection" section of FastLst.Cfg. These blocks MUST be at the _END_ of an "Input Block"; there can be multiple Export Blocks in a single Input Block. Obviously the Export Block is available for compiled nodelists only, thus it is illegal inside a "NoCompile" block. The export is done ONLY when a new NodeList is found (or when the file to be exported exists neither in uncompressed nor in archived form), even if the config file is changed. So, you can safely hatch the created arcfile via the AfterArcExport command with no danger of hatching it all the times you change something in the cfg. Under these conditions, if you really want to export anyway, you must use the -i command line switch. IMPORTANT: If you use the same export filename for multiple source nodelists, all the exported segments are appended one another. This way, if you like, you can make FastLst generate a "plain" nodelist file with many different source nodelists in it, just appended one after another. Some people need this feature to create input for some other program. For this feature to work, you need to specify the '+' parameter in the "Export" statement. See "Export Example" below. Export [+] <file> [<PartAddrLst>] The '+' sign must be specified when you want to create a joined list by exporting multiple nodelists to the same export <file>. This way the exported file will be created every time the nodelist is compiled and its timestamp will not be changed to be equal to the source. <file> is the name of the file to which you want to export the selected segment(s). <PartAddrLst> is the partial address list of segments to be exported. Usually it is a single partial address. If omitted, the entire nodelist is exported (useful to create a joined nodelist). This statement marks the start of an "Export Block". Multiple "Export Blocks" are allowed in the same "Input Block". N.B. The Export blocks must be at the _END_ of an input block. See "Export Example" below. Example: Export region25.??? 2:25 Export Section The following verbs define the parameters for the Export specified by the last "Export" statement. ArcExport <arcfile> [Keep#] <arcfile> is the name of the archive file to which you want to compress the exported <file>. [Keep#] is the optional number of archive versions to be kept, basing on the day of the year (the modification file date is also used to infer the correct chronological order). Example: ArcExport region25.??? 2 ArcExportDesc <description> <description> is the description to be applied to FILES.BBS when a new archive is created. Example: ArcExportDesc Region 25 %D, %a format Export Global Section The following verbs can be used in the "Export Section" of an "Export Block", in the "Input Section" of an "Input Block", in the "Output Section" of an "Output Block", in the "Global Section". In few words, they are legal everywhere except for the "NoCompile" block. Depending on their positions, they affect the involved nodelists only. ArcExportMethod <meth>[,<f>] ... Specifies the archive type(s) to be created for the exported file. <meth> is the archiver name as defined in Compress.Cfg. <f> is the optional first letter to be used for variable archive extensions. Example: ArcExportMethod zip lh,H BeforeArcExport <command> AfterArcExport <command> Commands to be run before/after archiving the exported file. <command> can be any type of command (executable file, batch file, internal command, alias, etc.) and supports the %a (full archive name) and %f (name of the file to be compressed, no path) and is run from the directory where %f resides. WARNING: since <command> is executed from the directory where the file to be compressed belongs, you might need to specify the location of the config files used by the programs invoked via <command>. Example: AfterArcExport Hatch %a ExportNeededBeforeKill Specifies that the exported file is needed by the "BeforeKillSource" command. Export Example: NodeList nodelist.??? NodeDiff nodediff.??? ArcList nodelist.??? 2 ArcDiff nodediff.??? 5 ArcListDesc Fido Nodelist for day %d (%D), %a format ArcDiffDesc Fido Nodediff for day %d,(%D), %a format Export region25.??? 2:25 ArcExport region25.??? 1 ArcExportDesc Region 25 %D, %a format ArcExportMethod zip lh AfterArcExport Hatch %a Export region24.??? 2:24 ArcExport region24.??? 1 ArcExportDesc Region 24 %D, %a format ArcExportMethod zip Export Example to generate a joined list: NodeList nodelist.??? NodeDiff nodediff.??? ArcList nodelist.??? 2 ArcDiff nodediff.??? 5 ArcListDesc Fido Nodelist for day %d (%D), %a format ArcDiffDesc Fido Nodediff for day %d,(%D), %a format Export + megalist.Lst NodeList zonelist.??? NodeDiff zonediff.??? ArcList zonelist.??? 2 ArcDiff zonediff.??? 5 ArcListDesc Zonelist for day %d (%D), %a format ArcDiffDesc Zonediff for day %d,(%D), %a format Export + megalist.Lst ArcExport megalist.??? 1 ArcExportDesc MegaList, %a format ArcExportMethod zip lh COMPRESS DEFINITION FILE The file specified in the CompressCfg statement is a sequence of Archive definition blocks, each one starting with "Archiver" and ending with "End Archiver". You can find an example in the Compress.Cfg file included in the distribution pack. The order of the archive definition blocks within this file may be important: when trying to unpack a compressed file, the list of archivers is scanned in a reverse order. In the case of two archivers that use the same identification string (e.g. ARC and PAK), you must specify the archiver that can unpack both (PAK) after the other one (ARC). The compress.cfg file can be shared between DOS/NT and OS/2 applications: the "DOS" and "OS2" keywords are available to distinguish between the commands to be used under DOS/NT and OS/2. O.S. specific archivers or commands must be prefixed with the relevant keyword. IMPORTANT NOTE: The lines that begin with "DOS" or "OS2" are parsed by the DOS/NT and OS/2 versions respectively. If you need the OS/2 version to execute a DOS command, you MUST NOT use the DOS keyword: if you do, it will never parse that line; if you do not, it will execute the DOS command "normally", provided you have installed OS/2's Dos support. See the examples below. Archiver <ARCname> Starts the Archive definition block. <ARCname> is the name used to identify this archiver. Example: Archiver ZIP Extension <ext> Specifies the default extension for the compressed files. Example: Extension ZIP Ident <ofs>,<ID> <ofs> is a decimal integer number representing the offset at which an archive identity marker <ID> must be present. Negative values can be used to indicate offsets from the END of a compressed file. -1 means "the last byte", -2 "the second last byte" and so on. <ID> is a series of hexadecimal figures which represent the bytes of the marker string that must be present at the specified offset of the archive file. Example: Ident 0,504b0304 ; "PK^c^d" Add <command> Specifies the command to add files to an archive. %a and %f are translated to the name of the archive and file to add. Example: Add zip -jk %a %f Extract <command> Specifies the command to extract files from an archive. %a and %f are translated to the name of the archive and file to extract. Example: Extract unzip -qqnjC %a %f View <command> This line is recognized and accepted for compatibility, but not used. End Archiver This statement is used to close a Archive definition. Examples Complete example 1 (you need OS/2 only): Archiver ZIP Extension ZIP Ident 0,504b0304 Add zip -jk %a %f Extract unzip -qqnjC %a %f View unzip -v %a End Archiver Complete example 2 (you need DOS only): Archiver ZIP Extension ZIP Ident 0,504b0304 Add pkzip -a %a %f Extract pkunzip -n %a %f View pkzip -v %a End Archiver Complete example 3 (you need both OS/2 and DOS): Archiver ZIP Extension ZIP Ident 0,504b0304 OS2 Add zip -jk %a %f DOS Add pkzip -a %a %f OS2 Extract unzip -qqnjC %a %f DOS Extract pkunzip -n %a %f OS2 View unzip -v %a DOS View pkzip -v %a End Archiver Complete example 4 (archiver to be used under DOS only): DOS Archiver ZOO DOS Extension ZOO DOS Ident 0,5a4f4f ; "ZOO" DOS Add zoo a: %a %f DOS Extract zoo e:O %a %f DOS View zoo v %a DOS End Archiver Complete example 5 (it's a DOS executable, to be used under DOS or OS/2 indifferently): Archiver ZOO Extension ZOO Ident 0,5a4f4f ; "ZOO" Add zoo a: %a %f Extract zoo e:O %a %f View zoo v %a End Archiver T R O U B L E S H O O T I N G Extraction problem Problem: FastLst does not extract the correct nodelist/nodediff. Solution: Perhaps there is some nodelist/nodediff with corrupted file date. Check your "ArcPath", manually extract to the "InputPath" the required nodelist/nodediff and delete the archive (or reset its file-date so that it is similar to that of the enclosed file). FastLst will automatically rearchive the nodelist/nodediff if you use "ArcMethod"/"ArcDiffMethod", otherwise you can rearchive manually. Out of Memory Problem: FastLst runs out of memory (Dos versions). Solution: DOS 16: - try freeing as much conventional memory as possible. - reduce the "MsgSize" buffer, if you have used that statement in the configuration. - be aware that 530K of free conventional memory is the minimum requirement. - If you have a 386sx or better, you can use the 32 bit version that has less stringent conventional memory requirements, but you will need at least 800KB of extended memory. DOS 32: - give more DPMI memory to FastLst - enable the temporary file method: use the -t command line switch or the "FileMode" statement in the configuration file. Problems with Empty Phone entries Problem: Some program behaves oddly while accessing entries that contain an empty phone number. Solution: The problem might be caused by the cost that is assigned to empty-phone nodes (65535 by default). Try using the "CostNullPhone" global statement to give lower costs. Example: CostNullPhone 900 0 Empty SysOp List Problem: The SysOp List is empty, even if I have specified its name in the configuration. Solution: You must use the SysOpLst statement to specify which nodelists must be considered for output to the SysOp list. Slow processing Problem: FastLst works very slowly (OS/2 version). Solution: Perhaps you are compiling a large nodelist or set of nodelists on a system with few MegaBytes of free physical RAM, so that OS/2 needs to extensively use virtual memory. Try using the "temporary file" method: specify the -t command line switch or use the "FileMode" statement in the configuration file. System performance degradation Problem: FastLst loads the system excessively, so that other OS/2 tasks don't perform properly (OS/2 version). Solution: Use the "Priority Idle" statement in the configuration file, so that FastLst receives its time slices only when other processes with higher priority are idle. I want maximum speed Problem: I run FastLst while the communications are off, so I would like it to run as fast as possible even if it is in the background and other tasks are active (OS/2 version). Solution: Use the "Priority High 31" statement in the configuration file, so that FastLst receives the maximum priority for "non time-critical" processes. Archived Diffs are not applied Problem: FastLst does not apply the archived Diffs. Solution: Remember that "InputPath <path>" is the default path for lists and diffs, while "ArcPath <path>" is the one for archives. Please compare your Compress.cfg with the example one, check the paths and try the commands manually. Check the day-extensions and time-stamps of the relevant files. Dos/32 DOS4GW exception Problem: The Dos/32 version of FastLst aborts with an exception from the Dos extender. Solution: Try using the 16 bit version. If this works, try again the 32 bit version, but starting with a "clean" config.sys and autoexec.bat (the Dos extender might be incompatible with some of your loaded drivers or TSRs. Dial Scripts and VMODEM addresses Problem: How can I put script names or internet addresses in the place of a phone number ? Solution: You may use the "Phone" statement in the "verbatim" form (phone string prefixed with '#'). You may find handy to specify flag and cost overrides in the same statement. See the "Phone" statement. Note: you will usually need to specify a modem type for Vmodem calls. Example: Let's suppose the following Modem Type table is defined: TypeDef ISDNC 1 V34 2 VMP 3 End You may use a Phone override of this kind: Phone 2:345/678 #domain.com VMP,CM 10 0 And a ModemTrans (for Binkley's VMODEM line): ModemTrans 0 ModemTrans 1 ModemTrans 2 ModemTrans 3 ATDT# ; Vmodem Region and zone-level Export Problem: How can I export a Region segment together with the zone-level entries ? Solution: The zone level entries have the Region/Net field equal to the zone number; you can use the Export statement in the following way: export MyR33.??? 1:1 2:2 2:33 3:3 4:4 5:5 6:6 Support ? Problem: I cannot find the solution to my problems. Solution: - Try linking the APWORKS support echo - Try asking your local supporter - Try asking the author directly You can find the addresses in the ReadMe.1st file. S H A R E W A R E If you like this program and continue using it, you should pay the author for his work, as per the ShareWare concept of distribution. Please see LICENSE.DOC and REGISTER.DOC for information. Thank you for your interest in FastLst.