1comm(n) Remote communication comm(n)
2______________________________________________________________________________
6
8 comm - A remote communication facility for Tcl (8.3 and later)
9
11 package require Tcl 8.3
12 package require comm ?4.5.7?
14 ::comm::comm send ?-async? ?-command callback? id cmd ?arg arg ...?
16 ::comm::comm self
18 ::comm::comm interps
20 ::comm::comm connect ?id?
22 ::comm::comm new chan ?name value ...?
24 ::comm::comm channels
26 ::comm::comm config
28 ::comm::comm config name
30 ::comm::comm config ?name value ...?
32 ::comm::comm shutdown id
34 ::comm::comm abort
36 ::comm::comm destroy
38 ::comm::comm hook event ?+? ?script?
40 ::comm::comm remoteid
42 ::comm::comm_send
44 ::comm::comm return_async
46 $future return ?-code code? ?value?
48 $future configure ?-command ?cmdprefix??
50 $future cget -command
52_________________________________________________________________
54
56 The comm command provides an inter-interpreter remote execution facil‐
57 ity much like Tk's send(n), except that it uses sockets rather than the
58 X server for the communication path. As a result, comm works with mul‐
59 tiple interpreters, works on Windows and Macintosh systems, and pro‐
60 vides control over the remote execution path.
61 These commands work just like send and winfo interps :
63 ::comm::comm send ?-async? id cmd ?arg arg ...?
65 ::comm::comm interps
66 This is all that is really needed to know in order to use comm
69 COMMANDS
71 The package initializes ::comm::comm as the default chan.
72 comm names communication endpoints with an id unique to each machine.
74 Before sending commands, the id of another interpreter is needed.
75 Unlike Tk's send, comm doesn't implicitly know the id's of all the
76 interpreters on the system. The following four methods make up the
77 basic comm interface.
78 ::comm::comm send ?-async? ?-command callback? id cmd ?arg arg ...?
80 This invokes the given command in the interpreter named by id.
81 The command waits for the result and remote errors are returned
82 unless the -async or -command option is given. If -async is
83 given, send returns immediately and there is no further notifi‐
84 cation of result. If -command is used, callback specifies a
85 command to invoke when the result is received. These options
86 are mutually exclusive. The callback will receive arguments in
87 the form -option value, suitable for array set. The options
88 are: -id, the comm id of the interpreter that received the com‐
89 mand; -serial, a unique serial for each command sent to a par‐
90 ticular comm interpreter; -chan, the comm channel name; -code,
91 the result code of the command; -errorcode, the errorcode, if
92 any, of the command; -errorinfo, the errorinfo, if any, of the
93 command; and -result, the return value of the command. If con‐
94 nection is lost before a reply is received, the callback will be
95 invoked with a connection lost message with -code equal to -1.
96 When -command is used, the command returns the unique serial for
97 the command.
98 ::comm::comm self
100 Returns the id for this channel.
101 ::comm::comm interps
103 Returns a list of all the remote id's to which this channel is
104 connected. comm learns a new remote id when a command is first
105 issued it, or when a remote id first issues a command to this
106 comm channel. ::comm::comm ids is an alias for this method.
107 ::comm::comm connect ?id?
109 Whereas ::comm::comm send will automatically connect to the
110 given id, this forces a connection to a remote id without send‐
111 ing a command. After this, the remote id will appear in
112 ::comm::comm interps.
113 EVAL SEMANTICS
115 The evaluation semantics of ::comm::comm send are intended to match
116 Tk's send exactly. This means that comm evaluates arguments on the
117 remote side.
118 If you find that ::comm::comm send doesn't work for a particular com‐
120 mand, try the same thing with Tk's send and see if the result is dif‐
121 ferent. If there is a problem, please report it. For instance, there
122 was had one report that this command produced an error. Note that the
123 equivalent send command also produces the same error.
124 % ::comm::comm send id llength {a b c}
126 wrong # args: should be "llength list"
127 % send name llength {a b c}
128 wrong # args: should be "llength list"
129 The eval hook (described below) can be used to change from send's dou‐
132 ble eval semantics to single eval semantics.
133 MULTIPLE CHANNELS
135 More than one comm channel (or listener) can be created in each Tcl
136 interpreter. This allows flexibility to create full and restricted
137 channels. For instance, hook scripts are specific to the channel they
138 are defined against.
139 ::comm::comm new chan ?name value ...?
141 This creates a new channel and Tcl command with the given chan‐
142 nel name. This new command controls the new channel and takes
143 all the same arguments as ::comm::comm. Any remaining arguments
144 are passed to the config method. The fully qualified channel
145 name is returned.
146 ::comm::comm channels
148 This lists all the channels allocated in this Tcl interpreter.
149 The default configuration parameters for a new channel are:
151 "-port 0 -local 1 -listen 0 -silent 0"
153 The default channel ::comm::comm is created with:
156 "::comm::comm new ::comm::comm -port 0 -local 1 -listen 1 -silent 0"
158 CHANNEL CONFIGURATION
161 The config method acts similar to fconfigure in that it sets or queries
162 configuration variables associated with a channel.
163 ::comm::comm config
165 ::comm::comm config name
167 ::comm::comm config ?name value ...?
169 When given no arguments, config returns a list of all variables
170 and their value With one argument, config returns the value of
171 just that argument. With an even number of arguments, the given
172 variables are set to the given values.
173 These configuration variables can be changed (descriptions of them are
175 elsewhere in this manual page):
176 -listen ?0|1?
178 -local ?0|1?
180 -port ?port?
182 -silent ?0|1?
184 -interp ?interpreter?
186 -events ?eventlist?
188 These configuration variables are read only:
190 -chan chan
192 -serial n
194 -socket sockIn
196 When config changes the parameters of an existing channel (with the
198 exception of -interp and -events), it closes and reopens the listening
199 socket. An automatically assigned channel id will change when this
200 happens. Recycling the socket is done by invoking ::comm::comm abort,
201 which causes all active sends to terminate.
202 ID/PORT ASSIGNMENTS
204 comm uses a TCP port for endpoint id. The interps (or ids) method
205 merely lists all the TCP ports to which the channel is connected. By
206 default, each channel's id is randomly assigned by the operating system
207 (but usually starts at a low value around 1024 and increases each time
208 a new socket is opened). This behavior is accomplished by giving the
209 -port config option a value of 0. Alternately, a specific TCP port
210 number may be provided for a given channel. As a special case, comm
211 contains code to allocate a a high-numbered TCP port (>10000) by using
212 -port {}. Note that a channel won't be created and initialized unless
213 the specific port can be allocated.
214 As a special case, if the channel is configured with -listen 0, then it
216 will not create a listening socket and will use an id of 0 for itself.
217 Such a channel is only good for outgoing connections (although once a
218 connection is established, it can carry send traffic in both direc‐
219 tions). As another special case, if the channel is configured with
220 -silent 0, then the listening side will ignore connection attempts
221 where the protocol negotiation phase failed, instead of throwing an
222 error.
223 EXECUTION ENVIRONMENT
225 A communication channel in its default configuration will use the cur‐
226 rent interpreter for the execution of all received scripts, and of the
227 event scripts associated with the various hooks.
228 This insecure setup can be changed by the user via the two options
230 -interp, and -events.
231 When -interp is set all received scripts are executed in the slave
233 interpreter specified as the value of the option. This interpreter is
234 expected to exist before configuration. I.e. it is the responsibility
235 of the user to create it. However afterward the communication channel
236 takes ownership of this interpreter, and will destroy it when the com‐
237 munication channel is destroyed. Note that reconfiguration of the com‐
238 munication channel to either a different interpreter or the empty
239 string will release the ownership without destroying the previously
240 configured interpreter. The empty string has a special meaning, it
241 restores the default behaviour of executing received scripts in the
242 current interpreter.
243 Also of note is that replies and callbacks (a special form of reply)
245 are not considered as received scripts. They are trusted, part of the
246 internal machinery of comm, and therefore always executed in the cur‐
247 rent interpreter.
248 Even if an interpreter has been configured as the execution environment
250 for received scripts the event scripts associated with the various
251 hooks will by default still be executed in the current interpreter. To
252 change this use the option -events to declare a list of the events
253 whose scripts should be executed in the declared interpreter as well.
254 The contents of this option are ignored if the communication channel is
255 configured to execute received scripts in the current interpreter.
256 REMOTE INTERPRETERS
258 By default, each channel is restricted to accepting connections from
259 the local system. This can be overridden by using the -local 0 config‐
260 uration option For such channels, the id parameter takes the form { id
261 host }.
262 WARNING: The host must always be specified in the same form (e.g., as
264 either a fully qualified domain name, plain hostname or an IP address).
265 CLOSING CONNECTIONS
267 These methods give control over closing connections:
268 ::comm::comm shutdown id
270 This closes the connection to id, aborting all outstanding com‐
271 mands in progress. Note that nothing prevents the connection
272 from being immediately reopened by another incoming or outgoing
273 command.
274 ::comm::comm abort
276 This invokes shutdown on all open connections in this comm chan‐
277 nel.
278 ::comm::comm destroy
280 This aborts all connections and then destroys the this comm
281 channel itself, including closing the listening socket. Special
282 code allows the default ::comm::comm channel to be closed such
283 that the ::comm::comm command it is not destroyed. Doing so
284 closes the listening socket, preventing both incoming and outgo‐
285 ing commands on the channel. This sequence reinitializes the
286 default channel:
287 "::comm::comm destroy; ::comm::comm new ::comm::comm"
290 When a remote connection is lost (because the remote exited or called
293 shutdown), comm can invoke an application callback. This can be used
294 to cleanup or restart an ancillary process, for instance. See the lost
295 callback below.
296 CALLBACKS
298 This is a mechanism for setting hooks for particular events:
299 ::comm::comm hook event ?+? ?script?
301 This uses a syntax similar to Tk's bind command. Prefixing
302 script with a + causes the new script to be appended. Without
303 this, a new script replaces any existing script. When invoked
304 without a script, no change is made. In all cases, the new hook
305 script is returned by the command.
306 When an event occurs, the script associated with it is evaluated
308 with the listed variables in scope and available. The return
309 code (not the return value) of the script is commonly used
310 decide how to further process after the hook.
311 Common variables include:
313 chan the name of the comm channel (and command)
315 id the id of the remote in question
317 fid the file id for the socket of the connection
319 These are the defined events:
321 connecting
323 Variables: chan, id
324 This hook is invoked before making a connection to the remote
326 named in id. An error return (via error) will abort the connec‐
327 tion attempt with the error. Example:
328 % ::comm::comm hook connecting {
331 if {[string match {*[02468]} $id]} {
332 error "Can't connect to even ids"
333 }
334 }
335 % ::comm::comm send 10000 puts ok
336 Connect to remote failed: Can't connect to even ids
337 %
338 connected
341 Variables: chan, fid, id, host, and port.
342 This hook is invoked immediately after making a remote connec‐
344 tion to id, allowing arbitrary authentication over the socket
345 named by fid. An error return (via error ) will close the con‐
346 nection with the error. host and port are merely extracted from
347 the id; changing any of these will have no effect on the connec‐
348 tion, however. It is also possible to substitute and replace
349 fid.
350 incoming
352 Variables: chan, fid, addr, and remport.
353 Hook invoked when receiving an incoming connection, allowing
355 arbitrary authentication over socket named by fid. An error
356 return (via error) will close the connection with the error.
357 Note that the peer is named by remport and addr but that the
358 remote id is still unknown. Example:
359 ::comm::comm hook incoming {
362 if {[string match 127.0.0.1 $addr]} {
363 error "I don't talk to myself"
364 }
365 }
366 eval Variables: chan, id, cmd, and buffer.
369 This hook is invoked after collecting a complete script from a
371 remote but before evaluating it. This allows complete control
372 over the processing of incoming commands. cmd contains either
373 send or async. buffer holds the script to evaluate. At the
374 time the hook is called, $chan remoteid is identical in value to
375 id.
376 By changing buffer, the hook can change the script to be evalu‐
378 ated. The hook can short circuit evaluation and cause a value
379 to be immediately returned by using return result (or, from
380 within a procedure, return -code return result). An error
381 return (via error) will return an error result, as is if the
382 script caused the error. Any other return will evaluate the
383 script in buffer as normal. For compatibility with 3.2, break
384 and return -code break result is supported, acting similarly to
385 return {} and return -code return result.
386 Examples:
388 [1] augmenting a command
390 % ::comm::comm send [::comm::comm self] pid
393 5013
394 % ::comm::comm hook eval {puts "going to execute $buffer"}
395 % ::comm::comm send [::comm::comm self] pid
396 going to execute pid
397 5013
398 [2] short circuiting a command
401 % ::comm::comm hook eval {puts "would have executed $buffer"; return 0}
404 % ::comm::comm send [::comm::comm self] pid
405 would have executed pid
406 0
407 [3] Replacing double eval semantics
410 % ::comm::comm send [::comm::comm self] llength {a b c}
413 wrong # args: should be "llength list"
414 % ::comm::comm hook eval {return [uplevel #0 $buffer]}
415 return [uplevel #0 $buffer]
416 % ::comm::comm send [::comm::comm self] llength {a b c}
417 3
418 [4] Using a slave interpreter
421 % interp create foo
424 % ::comm::comm hook eval {return [foo eval $buffer]}
425 % ::comm::comm send [::comm::comm self] set myvar 123
426 123
427 % set myvar
428 can't read "myvar": no such variable
429 % foo eval set myvar
430 123
431 [5] Using a slave interpreter (double eval)
434 % ::comm::comm hook eval {return [eval foo eval $buffer]}
437 [6] Subverting the script to execute
440 % ::comm::comm hook eval {
443 switch -- $buffer {
444 a {return A-OK}
445 b {return B-OK}
446 default {error "$buffer is a no-no"}
447 }
448 }
449 % ::comm::comm send [::comm::comm self] pid
450 pid is a no-no
451 % ::comm::comm send [::comm::comm self] a
452 A-OK
453 reply Variables: chan, id, buffer, ret, and return().
456 This hook is invoked after collecting a complete reply script
458 from a remote but before evaluating it. This allows complete
459 control over the processing of replies to sent commands. The
460 reply buffer is in one of the following forms
461 · return result
463 · return -code code result
465 · return -code code -errorinfo info -errorcode ecode msg
467 For safety reasons, this is decomposed. The return result is in ret,
469 and the return switches are in the return array:
470 · return(-code)
472 · return(-errorinfo)
474 · return(-errorcode)
476 Any of these may be the empty string. Modifying these four variables
478 can change the return value, whereas modifying buffer has no effect.
479 callback
481 Variables: chan, id, buffer, ret, and return().
482 Similar to reply, but used for callbacks.
484 lost Variables: chan, id, and reason.
486 This hook is invoked when the connection to id is lost. Return
488 value (or thrown error) is ignored. reason is an explanatory
489 string indicating why the connection was lost. Example:
490 ::comm::comm hook lost {
493 global myvar
494 if {$myvar(id) == $id} {
495 myfunc
496 return
497 }
498 }
499 UNSUPPORTED
502 These interfaces may change or go away in subsequence releases.
503 ::comm::comm remoteid
505 Returns the id of the sender of the last remote command executed
506 on this channel. If used by a proc being invoked remotely, it
507 must be called before any events are processed. Otherwise,
508 another command may get invoked and change the value.
509 ::comm::comm_send
511 Invoking this procedure will substitute the Tk send and winfo
512 interps commands with these equivalents that use ::comm::comm.
513 proc send {args} {
516 eval ::comm::comm send $args
517 }
518 rename winfo tk_winfo
519 proc winfo {cmd args} {
520 if {![string match in* $cmd]} {
521 return [eval [list tk_winfo $cmd] $args]
522 }
523 return [::comm::comm interps]
524 }
525 SECURITY
528 Something here soon.
529 BLOCKING SEMANTICS
531 There is one outstanding difference between comm and send. When block‐
532 ing in a synchronous remote command, send uses an internal C hook
533 (Tk_RestrictEvents) to the event loop to look ahead for send-related
534 events and only process those without processing any other events. In
535 contrast, comm uses the vwait command as a semaphore to indicate the
536 return message has arrived. The difference is that a synchronous send
537 will block the application and prevent all events (including window
538 related ones) from being processed, while a synchronous ::comm::comm
539 send will block the application but still allow other events to get
540 processed. In particular, after idle handlers will fire immediately
541 when comm blocks.
542 What can be done about this? First, note that this behavior will come
544 from any code using vwait to block and wait for an event to occur. At
545 the cost of multiple channel support, comm could be changed to do
546 blocking I/O on the socket, giving send-like blocking semantics. How‐
547 ever, multiple channel support is a very useful feature of comm that it
548 is deemed too important to lose. The remaining approaches involve a
549 new loadable module written in C (which is somewhat against the philos‐
550 ophy of comm) One way would be to create a modified version of the
551 vwait command that allow the event flags passed to Tcl_DoOneEvent to be
552 specified. For comm, just the TCL_FILE_EVENTS would be processed.
553 Another way would be to implement a mechanism like Tk_RestrictEvents,
554 but apply it to the Tcl event loop (since comm doesn't require Tk).
555 One of these approaches will be available in a future comm release as
556 an optional component.
557 ASYNCHRONOUS RESULT GENERATION
559 By default the result returned by a remotely invoked command is the
560 result sent back to the invoker. This means that the result is gener‐
561 ated synchronously, and the server handling the call is blocked for the
562 duration of the command.
563 While this is tolerable as long as only short-running commands are
565 invoked on the server long-running commands, like database queries make
566 this a problem. One command can prevent the processing requests of all
567 other clients for an arbitrary period of time.
568 Before version 4.5 of comm the only solution was to rewrite the server
570 command to use the Tcl builtin command vwait, or one of its relatives
571 like tkwait, to open a new event loop which processes requests while
572 the long-running operation is executed. This however has its own per‐
573 ils, as this makes it possible to both overflow the Tcl stack with a
574 large number of event loop, and to have a newer requests block the
575 return of older ones, as the eventloop have to be unwound in the order
576 of their creation.
577 The proper solution is to have the invoked command indicate to comm
579 that it cannot or will not deliver an immediate, synchronous result,
580 but will do so later. At that point the framework can put sending the
581 actual result on hold and continue processing requests using the main
582 event loop. No blocking, no nesting of event loops. At some future date
583 the long running operation delivers the result to comm, via the future
584 object, which is then forwarded to the invoker as usual.
585 The necessary support for this solution has been added to comm since
587 version 4.5, in the form of the new method return_async.
588 ::comm::comm return_async
590 This command is used by a remotely invoked script to notify the
591 comm channel which invoked it that the result to send back to
592 the invoker is not generated synchronously. If this command is
593 not called the default/standard behaviour of comm is to send the
594 synchronously generated result of the script itself to the
595 invoker.
596 The result of return_async is an object. This object, called a
598 future is where the result of the script has to be delivered to
599 when it becomes ready. When that happens it will take all the
600 necessary actions to deliver the result to the invoker of the
601 script, and then destroy itself. Should comm have lost the con‐
602 nection to the invoker while the result is being computed the
603 future will not try to deliver the result it got, but just
604 destroy itself. The future can be configured with a command to
605 call when the invoker is lost. This enables the user to imple‐
606 ment an early abort of the long-running operation, should this
607 be supported by it.
608 An example:
610 # Procedure invoked by remote clients to run database operations.
612 proc select {sql} {
613 # Signal the async generation of the result
614 set future [::comm::comm return_async]
616 # Generate an async db operation and tell it where to deliver the result.
618 set query [db query -command [list $future return] $sql]
620 # Tell the database system which query to cancel if the connection
622 # goes away while it is running.
623 $future configure -command [list db cancel $query]
625 # Note: The above will work without problem only if the async
627 # query will nover run its completion callback immediately, but
628 # only from the eventloop. Because otherwise the future we wish to
629 # configure may already be gone. If that is possible use 'catch'
630 # to prevent the error from propagating.
631 return
632 }
633 The API of a future object is:
636 $future return ?-code code? ?value?
638 Use this method to tell the future that long-running
639 operation has completed. Arguments are an optional return
640 value (defaults to the empty string), and the Tcl return
641 code (defaults to OK).
642 The future will deliver this information to invoker, if
644 the connection was not lost in the meantime, and then
645 destroy itself. If the connection was lost it will do
646 nothing but destroy itself.
647 $future configure ?-command ?cmdprefix??
649 $future cget -command
651 These methods allow the user to retrieve and set a com‐
652 mand to be called if the connection the future belongs to
653 has been lost.
654 COMPATIBILITY
656 comm exports itself as a package. The package version number is in the
657 form major . minor, where the major version will only change when a
658 non-compatible change happens to the API or protocol. Minor bug fixes
659 and changes will only affect the minor version. To load comm this com‐
660 mand is usually used:
661 package require comm 3
663 Note that requiring no version (or a specific version) can also be
666 done.
667 The revision history of comm includes these releases:
669 4.5.7 Changed handling of ports already in use to provide a proper
671 error message.
672 4.5.6 Bugfix in the replacement for vwait, made robust against of
674 variable names containing spaces.
675 4.5.5 Bugfix in the handling of hooks, typo in variable name.
677 4.5.4 Bugfix in the handling of the result received by the send
679 method. Replaced an after idle unset result with an immediate
680 unset, with the information saved to a local variable.
681 The after idle can spill into a forked child process if there is
683 no event loop between its setup and the fork. This may bork the
684 child if the next event loop is the vwait of comm's send a few
685 lines above the after idle, and the child used the same serial
686 number for its next request. In that case the parent's after
687 idle unset will delete the very array element the child is wait‐
688 ing for, unlocking the vwait, causing it to access a now missing
689 array element, instead of the expected result.
690 4.5.3 Bugfixes in the wrappers for the builtin update and vwait com‐
692 mands.
693 4.5.2 Bugfix in the wrapper for the builtin update command.
695 4.5.1 Bugfixes in the handling of -interp for regular scripts. The
697 handling of the buffer was wrong for scripts which are a single
698 statement as list. Fixed missing argument to new command comm‐
699 SendReply, introduced by version 4.5. Affected debugging.
700 4.5 New server-side feature. The command invoked on the server can
702 now switch comm from the standard synchronous return of its
703 result to an asynchronous (defered) return. Due to the use of
704 snit to implement the future objects used by this feature from
705 this version on comm requires at least Tcl 8.3 to run. Please
706 read the section ASYNCHRONOUS RESULT GENERATION for more
707 details.
708 4.4.1 Bugfix in the execution of hooks.
710 4.4 Bugfixes in the handling of -interp for regular and hook
712 scripts. Bugfixes in channel cleanup.
713 4.3.1 Introduced -interp and -events to enable easy use of a slave
715 interp for execution of received scripts, and of event scripts.
716 4.3 Bugfixes, and introduces -silent to allow the user to force the
718 server/listening side to silently ignore connection attempts
719 where the protocol negotiation failed.
720 4.2 Bugfixes, and most important, switched to utf-8 as default
722 encoding for full i18n without any problems.
723 4.1 Rewrite of internal code to remove old pseudo-object model.
725 Addition of send -command asynchronous callback option.
726 4.0 Per request by John LoVerso. Improved handling of error for
728 async invoked commands.
729 3.7 Moved into tcllib and placed in a proper namespace.
731 3.6 A bug in the looking up of the remoteid for a executed command
733 could be triggered when the connection was closed while several
734 asynchronous sends were queued to be executed.
735 3.5 Internal change to how reply messages from a send are handled.
737 Reply messages are now decoded into the value to pass to return;
738 a new return statement is then cons'd up to with this value.
739 Previously, the return code was passed in from the remote as a
740 command to evaluate. Since the wire protocol has not changed,
741 this is still the case. Instead, the reply handling code
742 decodes the reply message.
743 3.4 Added more source commentary, as well as documenting config
745 variables in this man page. Fixed bug were loss of connection
746 would give error about a variable named pending rather than the
747 message about the lost connection. comm ids is now an alias for
748 comm interps (previously, it an alias for comm chans). Since
749 the method invocation change of 3.0, break and other exceptional
750 conditions were not being returned correctly from comm send.
751 This has been fixed by removing the extra level of indirection
752 into the internal procedure commSend. Also added propagation of
753 the errorCode variable. This means that these commands return
754 exactly as they would with send:
755 comm send id break
758 catch {comm send id break}
759 comm send id expr 1 / 0
760 Added a new hook for reply messages. Reworked method invocation
763 to avoid the use of comm:* procedures; this also cut the invoca‐
764 tion time down by 40%. Documented comm config (as this manual
765 page still listed the defunct comm init!)
766 3.3 Some minor bugs were corrected and the documentation was cleaned
768 up. Added some examples for hooks. The return semantics of the
769 eval hook were changed.
770 3.2 A new wire protocol, version 3, was added. This is backwards
772 compatible with version 2 but adds an exchange of supported pro‐
773 tocol versions to allow protocol negotiation in the future.
774 Several bugs with the hook implementation were fixed. A new
775 section of the man page on blocking semantics was added.
776 3.1 All the documented hooks were implemented. commLostHook was
778 removed. A bug in comm new was fixed.
779 3.0 This is a new version of comm with several major changes. There
781 is a new way of creating the methods available under the comm
782 command. The comm init method has been retired and is replaced
783 by comm configure which allows access to many of the well-
784 defined internal variables. This also generalizes the options
785 available to comm new. Finally, there is now a protocol version
786 exchanged when a connection is established. This will allow for
787 future on-wire protocol changes. Currently, the protocol ver‐
788 sion is set to 2.
789 2.3 comm ids was renamed to comm channels. General support for comm
791 hook was fully implemented, but only the lost hook exists, and
792 it was changed to follow the general hook API. commLostHook was
793 unsupported (replaced by comm hook lost) and commLost was
794 removed.
795 2.2 The died hook was renamed lost, to be accessed by commLostHook
797 and an early implementation of comm lost hook. As such, com‐
798 mDied is now commLost.
799 2.1 Unsupported method comm remoteid was added.
801 2.0 comm has been rewritten from scratch (but is fully compatible
803 with Comm 1.0, without the requirement to use obTcl).
804
806 John LoVerso, John@LoVerso.Southborough.MA.US
807 http://www.opengroup.org/~loverso/tcl-tk/#comm
809
811 Please see the file comm.LICENSE that accompanied this source, or
812 http://www.opengroup.org/www/dist_client/caubweb/COPYRIGHT.free.html.
813 This license for comm, new as of version 3.2, allows it to be used for
815 free, without any licensing fee or royalty.
816
818 · If there is a failure initializing a channel created with
819 ::comm::comm new, then the channel should be destroyed. Cur‐
820 rently, it is left in an inconsistent state.
821 · There should be a way to force a channel to quiesce when chang‐
823 ing the configuration.
824 The following items can be implemented with the existing hooks and are
826 listed here as a reminder to provide a sample hook in a future version.
827 · Allow easier use of a slave interp for actual command execution
829 (especially when operating in "not local" mode).
830 · Add host list (xhost-like) or "magic cookie" (xauth-like)
832 authentication to initial handshake.
833 The following are outstanding todo items.
835 · Add an interp discovery and name->port mapping. This is likely
837 to be in a separate, optional nameserver. (See also the related
838 work, below.)
839 · Fix the {id host} form so as not to be dependent upon canonical
841 hostnames. This requires fixes to Tcl to resolve hostnames!
842 This man page is bigger than the source file.
844
846 Tcl7.5 under Windows contains a bug that causes the interpreter to hang
847 when EOF is reached on non-blocking sockets. This can be triggered
848 with a command such as this:
849 "comm send $other exit"
851 Always make sure the channel is quiescent before closing/exiting or use
854 at least Tcl7.6 under Windows.
855 Tcl7.6 on the Mac contains several bugs. It is recommended you use at
857 least Tcl7.6p2.
858 Tcl8.0 on UNIX contains a socket bug that can crash Tcl. It is recom‐
860 mended you use Tcl8.0p1 (or Tcl7.6p2).
861
863 Tcl-DP provides an RPC-based remote execution interface, but is a com‐
864 piled Tcl extension. See http://www.cs.cor‐
865 nell.edu/Info/Projects/zeno/Projects/Tcl-DP.html.
866 Michael Doyle <miked@eolas.com> has code that implements the Tcl-DP RPC
868 interface using standard Tcl sockets, much like comm.
869 Andreas Kupries <andreas_kupries@users.sourceforge.net> uses comm and
871 has built a simple nameserver as part of his Pool library. See
872 http://www.purl.org/net/akupries/soft/pool/index.htm.
873
875 This document, and the package it describes, will undoubtedly contain
876 bugs and other problems. Please report such in the category comm of
877 the Tcllib SF Trackers [http://source‐
878 forge.net/tracker/?group_id=12883]. Please also report any ideas for
879 enhancements you may have for either package and/or documentation.
880
882 send(n)
883
885 Copyright (c) 1995-1998 The Open Group. All Rights Reserved.
886 Copyright (c) 2003-2004 ActiveState Corporation.
887 Copyright (c) 2006-2007 Andreas Kupries <andreas_kupries@users.sourceforge.net>
888comm 4.5.7 comm(n)