March 2004 - Boost - lists.boost.org

[Threads] Reactive Objects
by scott 05 Mar '04

05 Mar '04

Hi, I have been working on a small library that has become reasonably useful and possibly of interest to someone else. It's also become fairly large. I am quite interested in this library getting "boosted" but even given that there is substance to the library its scope is a bit of a problem; its quite large. It was started as a project with requirements very similar to those associated with the ActiveObject pattern; http://citeseer.nj.nec.com/lavender96active.html (reference from J. Garland) and takes many cues from the concepts and primitives found in SDL; http://www.sdl-forum.org/ Its possible to characterize this library as the basic building blocks for clean, robust multi-threading <ahem>. It could also be characterized as a subset of TAO, i.e. CORBA. While its a very small subset <sheepish grin>, maybe these comments gives scale to the scope issue. A breakdown of the incidental technologies tackled by this library follows; * multi-threading building block classes * network representation of structured data (variants) * saving and loading of network data (persistence) * interactive machines (inter-thread messaging) * stateless and stateful machine representation * code generation of network data and machine skeletons * message forwarding (distributed messaging) * transparent remote machines (proxies for inter-process messaging) * support for multiple encodings, e.g. binary, text, XML... To flesh this list out into something less opaque, a set of file fragments follows. These are the result of applying the library to the test case pushed around in the recent thread "Simple Active Object". They are necessarily abbreviated; // --------------------------------- // db_server.tms (the schema) // record configuration { // Persistent portion of the db_server string data_source integer maximum_workers integer maximum_clients } record open {} // Messages between the server and its clients record close {} record select { string sql_statement } record busy {} state INTIAL, // Pre-defined READY, // Normal operation OVERLOADED, // All workers busy OUT_OF_SERVICE // Something went wrong // Define the messages expected in the // different states transition INITIAL void // Standard startup transition READY open, // Another client connecting close, // Client closing configuration, // Reconfiguration select // An SQL select statement .. // --------------------------------- // db_server.h (auto-generated header) // .. #define db_server_configuration_hash 0x03493ac9 #define db_server_open_hash 0x7a6ffa00 #define db_server_close_hash 0x1093ac66 #define db_server_select_hash 0x004410ac #define db_server_busy_hash 0x74529711 struct typed03493ac9 { .. string &data_source; long &maximum_workers; long &maximum_clients; .. }; struct db_server { .. enum { INITIAL, READY, OVERLOADED, OUT_OF_SERVICE }; typedef typed03493ac9 configuration; typedef typed7a6ffa00 open; typedef typed1093ac66 close; typedef typed004410ac select; typedef typed74529711 busy; virtual int transition( typed_state<INITIAL> ) = 0; // void message virtual int transition( typed_state<READY>, typed_bind<open> & ) = 0; virtual int transition( typed_state<READY>, typed_bind<close> & ) = 0; virtual int transition( typed_state<READY>, typed_bind<configuration> & ) = 0; virtual int transition( typed_state<READY>, typed_bind<select> & ) = 0; int machine( typed_word & ); // The auto-generated switch statements .. }; .. // --------------------------------- // my_server.cpp (application declaration of db_server) // #include "db_server.h" class my_server : public state_machine<db_server> { typed_memory_folder home; db_server configuration; address_group client; virtual int transition( typed_state<INITIAL> ); virtual int transition( typed_state<READY>, typed_bind<open> & ); .. }; // my_server.cpp (application definition of db_server) // my_server::my_server() : home( "configuration\\my_server" ) { try // Persistence { load( home, configuration ); // Recover previous image } catch( ... ) { // Not found. Set factory defaults configuration.data_source = ""; configuration.maximum_workers = 4; configuration.maximum_clients = 100; save( home, configuration ); } } int my_server::transition( typed_state<INITIAL> ) { if( data_source.empty() ) { return OUT_OF_SERVICE; } if( ... ) // Attempt to connect to RDBMS { return READY; } return OUT_OF_SERVICE; } int my_server::transition( typed_state<READY>, typed_bind<open> &s ) { client.insert( sender() ); reply( configuration ); } .. int my_server::transition( typed_state<READY>, typed_bind<select> &s ) { if( ... ) // Find an idle worker { // Submit request to worker .. return READY; } reply( busy() ); return READY; } Hopefully these fragments have been enough to sketch out the intent of the library and where it currently is "at". As a bunch of materials, it is a .lib, headers and a .exe (the schema parser+generator) There are (serious) pretentions to platform independence but current materials are for Windows. I am very happy to work towards a boosted version of a subset of this work cos basically it seems too large to be a prospect otherwise. This subset would be in the area of mutli-threading building block classes, i.e. an "alternate threading model". However, after the recent exchange of related and interesting messages, I felt compelled to "come clean" with some background. So, to Matthew and Mark. Are you interested in dicsussing the possible implementation of; * simple active objects (completion of current alternate threading model) * simple reactive objects (completion of current with two-way calling of methods) * something else? Cheers, Scott

5 12

More versatile any_cast
by David Abrahams 05 Mar '04

05 Mar '04

Someone wanted to be able to put a derived class into a boost::any and use any_cast to get a base class pointer out. Enclosed is a patch to do it, FWIW. Disclaimer: I have no opinion on whether this should be done or not. -- Dave Abrahams Boost Consulting www.boost-consulting.com

1 0

RE: [boost] Re: [Threads] Simple active object wrapper, take 2
by Darryl Green 04 Mar '04

04 Mar '04

> -----Original Message----- > From: Matthew Vogt [mailto:mvogt@juptech.com] > Sent: Wednesday, 3 March 2004 8:44 AM > To: boost(a)lists.boost.org > Subject: [boost] Re: [Threads] Simple active object wrapper, take 2 > > > scott <scottw <at> qbik.com> writes: > > > Yep. All straight now. I know understand your breakdown of calls > > to callbacks is very similar to that inherent in SDL. And > yes, you/we/it > > are fully asynchronous. > > > > I'm assuming you prefer this to the reactive object version? Which > > does a pretty good job of hiding the comms Oh yeah, and doesnt > > require selection of the method to call in either the receipient > > or the client (on response), i.e. at the point of making a method > > call. > > > > Maybe we're down to a matter of taste? > > Yes, I guess it depends on whether you're > familiar/comfortable with the message > passing paradigm, or whether you're more comfortable with C++ > shenanigans to > obscure the messy details. It may be the sign of a warped mind but queueing some form of bound functor-like object has almost always seemed to be (broadly) what I wanted to do in any message passing scheme, whether or not fancy machinery existed to do it. However, there are times when it is useful to have the messaging defined independently of either the sender or the receiver, with both being written to conform to the messaging interface. I've tended to think of the tightly/early bound, just queue a functor, and the message passsing + switch case as 2 completely independent patterns, but maybe they should be built on a common framework (at a level above the detail that they both ivolve queuing some sort of object). The AO pattern doesn't require that you identify the object you want to invoke the method on in the proxy, because you send the proxy to that object, and only that object. So one could drop the "this" parameter from the proxy. That would potentially allow the proxy to be routed/forwarded in a more general reactive object model (I think this was one of Scotts requirements)? I'm not 100% sure how useful this is, but it might be nice for the class of problem where you do want concurrent execution - of as many completely independent instances of the object as is "useful". The client/master doesn't care which instance - the next available one will do nicely. Further, it would seem possible to exclude any binding to a particular member function from the proxy itself and instead allow "free proxies" to be defined. To avoid confusion, I'm going to drop the work proxy for these "free proxies" and call them "messages". These would just define a function signature to match, not a particular (member) function. It is not clear to that the type of the "this" parameter need be defined by the message - excluding it would offer better separation between message sender and receiver. The sender need only know that it wants to notify or invoke something by sending the message. Such message types (just a tuple) could be defined with no reference to the wrapper or receiving object type at all. As you have said, the downside is that there is some form of type-switch implied in the receiver (eg the receiver wrapper would accept a boost::variant<message_type_1, message_type_2...> from the activation list and apply a generic visitor to do the dispatch). Maybe this can be avoided if there is a separate queue per message type? The additonal queue/scheduler complexity would probably wipe out any potential advantage though. So a receiver wrapper generator would need to, given the set of message types and corresponding member function pointers to handle them, generate the variant and corresponding visitor. I certainly like this better than a hand coded switch, and it is something like a simplified/limited form of the compile time (actually code generation time) generation of message demux code that is part of scotts existing framework (I think - I only skimmed scotts descriptions and possibly guessed too much from vaguely similar, but more primitive things I've used before). > I would postulate, however, that wrapping the details to > allow binding at the > call site (or message post) is slightly superior, due to the > removal of the > dispatch. Of course, this minor gain may easily be outweighed by the > contortions required to make it work, and the reduction in > the transparency > of the code... The tightly/call site bound case can be written as a simple layer on top of the lose/late bound system. Message type becomes variant<proxy> and the visitor for the one and only type dispatches to a wrapper function that just does the stuff that was done in the thread_function loop in your AO code. I'm not sure that for a reasonably small (any sane?) interface the performance difference due to the demux will be terribly large anyway, so maybe the specialisation isn't needed for that alone. However I've so far ignored the return value in this - the return type could obviously be defined in the message, but what should define how/where it should be delivered? It seems that the caller should define how/where it wants results delivered at the time of invoking the request - which it can do by passing some sort of result handler object in the message instance. Of course this is what the synch object in your proxy is. Perhaps its interface can be made more generic to allow its operation to be policy based. Broadening the definition from "return value" to "message delivery notification" with the sender supplied message delivery notifier object passed the return value from the handler doesn't feel too weird to include at the low level, and shouldn't cost anything for a void return without notification. Interestingly, the option to generate a a notification that a message with a handler returning void has been processed doesn't seem like a strange feature when looked at this way (a void future is a bit odd though). > > > > I hadn't realised how integral the state machine was to your > > > design until I > > > read your other post. Still, if you are doing dispatch on > > > both message code > > > and object state, it can be simplified to bind the message to > > > a per-message > > > code dispatch function which then dispatches on object state. > > > > Characterizing of reactive objects as a framework for state machines > > would be a little bit sad. I'm not sure it was your intention to say > > that so just in case others are listening, I will elaborate from > > my POV. > > > > Sorry, I read too much into the example code in the other > message. I assumed > the messaging would be an independent layer, but that the > design you were > using must have the communicating objects modelled as state > machines. Thanks > for clearing that up. > This observer didn't see the state machine aspect as the exclusive use, more an important use case reflected in the design. Would that be fair to say? > <snip> > > > The messaging facillity has _no understanding_ of what the > recipient does > > with a message. Proxies and state machines were intended to > highlight > > the benefits of that design. > > > > To summarize; the Reactive Objects that I currently > consider our target > > would communicate via a similar mechansim and there would be no > > assumption that a Reactive Object was a state machine, or a > proxy, or > > anything except an object that accepts messages. My questions that I've been thinking out loud about (always dangerous in public) were: What is a message? What is acceptance? Sorry this was so long - am I barking up the wrong tree here or not? > > Yes, that all sounds good. > > > Cheers, > > Scott > > Matt > > _______________________________________________ > Unsubscribe & other changes: > http://lists.boost.org/mailman/listinfo.cgi/bo> ost > ########################################################################## This e-mail is for the use of the intended recipient(s) only. If you have received this e-mail in error, please notify the sender immediately and then delete it. If you are not the intended recipient, you must not use, disclose or distribute this e-mail without the author's prior permission. We have taken precautions to minimise the risk of transmitting software viruses, but we advise you to carry out your own virus checks on any attachment to this message. We cannot accept liability for any loss or damage caused by software viruses. ##########################################################################

3 2

Re: [boost] Re: GUI library - another one
by Beman Dawes 04 Mar '04

04 Mar '04

At 12:10 PM 3/4/2004, David Turner wrote: >... A button in one window isn't necessarily >the same thing as a button in another window; just as a bold font in one >document isn't necessarily the same as a bold font in another. I'd rewrite that as "The implementation of a button in one window isn't necessarily the same thing as the implementation of a button in another window..." To modify Dave's example to make a point: button OK_button("OK"); window_one("Warning: now entering twilight zone one") [ text_field("my favorite color is:"), OK_button | button("cancel") | attribute_x(1) ] window_two("Warning: now entering twilight zone two") [ text_field("my favorite color is:"), OK_button | button("cancel") | attribute_x(2) ] Perhaps the implementation of OK_button in window_one is very different from window_two. But they both have OK_buttons, and at the interface level these are the same button. > >One cannot say in general that elements can be exchanged between >documents. Similarly, one cannot say in general that widgets can be >exchanged between windows. I understand that the implementation details can't be exchanged. But we aren't taking about implementations of buttons, we are talking about the abstract concept of an OK_button. That certainly could be transferred, couldn't it? More importantly, you could build up a widget library which could be used regardless of the target window. Sorry if I'm missing something here; I'm not much of a GUI programmer. --Beman

6 5

Re: GUI library - another one
by Jesse Booher 04 Mar '04

04 Mar '04

>>> dave(a)boost-consulting.com 03/04/04 08:59AM >>> > button("click me") doesn't have to be a widget. Think expression > templates. It can just build up a type to be used by window to add a > button. Well said. This exemplifies the reason why most gui toolkits are so horrible to use, they allow implementation technicalities to dictate interface. Having to decide where an object is going to live before any of its properties can be set is a clear violation of object oriented design principles but we live with these kind of problems all the time for supposed technical reasons.

7 12

Re: [boost] Re: GUI library - another one
by Jesse Booher 04 Mar '04

04 Mar '04

David Turner wrote: > At a broader level, I'd argue that there is a good reason to use this > document/elements design because the type of the widgets have to match the > type of window they're going to go into. If we think of a button object as a particular implementation designed explicitly to support a particular window's interface requirements, this is would make sense. However, we might step back from the implementation particulars and consider a button to be a collection of properties such as caption, color, etc. that might at some point get bound to a window at which point the button object could become linked to a widget implementation appropriate to that context. This opens up the possibility that an object will specify properties that the implementation to which it gets attached does not support, but this is one of the corner stones of object oriented design. An object (e.g. button) advertises it properties and it's up to the user of the object (e.g. window) to realize those properties as best it can. Regards, Jesse Booher

1 0

RE: [boost] Re: First lambda
by Powell, Gary 04 Mar '04

04 Mar '04

>>Yeah, but do you propose to make local names visible inside the anonymous >>function, or not? If you do, you open a big can of worms for implementations >>(recursive nested functions, need I say more?), if you don't, BLL can still >>do things that anonymous functions can not. Anonymous functors could help >>some, but their syntax won't be quite as compact as your example above. > Now, to complicate matters, what if I wish to access v in the > inner anonymous function? Can you also do that? Example: > std::for_each(vv.begin(), vv.end(), > void (vector<int> const& v) > { > std::for_each(v.begin(), v.end(), > void (int x) > { > cout << x << endl; > do_fomething_with(v); // <<--here > } > } > ); > Joel de Guzman This gets ugly. As you want to "import" only one the variables you declare. For the previous example with std::accumulate. e.g.. implementation. template <class InputIterator, class T, class BinaryOperation> T accumulate<InputIterator v, InputIterator l, T i, BinaryOperation bin_op) { while (v != l) i = bin_op(i, *v++); return i; } int i = 0; unsigned long n = std::accumulate( xs.begin(), xs.end(), 0UL, auto (auto lhs, auto rhs) { return ( lhs + rhs.count) + i; // Which "i"!!!! } ); And this is where BLL wins big, it has a mechanism for importing variables into the fn. And no I don't have a good solution for this. -Gary-

1 0

RE: [boost] Re: First lambda
by Powell, Gary 04 Mar '04

04 Mar '04

> "Shannon Stewman" <stew(a)uchicago.edu> wrote in message > news:20040304025205.GD832@uchicago.edu... > >>On Wed, Mar 03, 2004 at 01:34:50PM -0800, Powell, Gary wrote: >> >> >>>unsigned long n = std::accumulate( >>> xs.begin(), xs.end(), 0UL, >>> long (long lhs, X const &rhs) // inline fn with no > > name. > >>> { >>> return lhs + rhs.count; >>> } >>> ); >> >>This seems to cast out the type inference present in BLL. I think >>that's one of the nicer concepts in the lambda library, and a real >>boon to programming. In my limited experience with SML/OCaml, type >>inference seems a better, safer, and more convenient way to program, >>albeit requiring a lot more compiler intelligence. As BLL shows, > > though, > >>this is quite possible, though perhaps limited, with existing C++. >> > > > How about: > > unsigned long n = std::accumulate( > xs.begin(), xs.end(), 0UL, > auto(long lhs, const auto& rhs) // inline fn with no > name. > { > return lhs + rhs.count; > } > ); > I would prefer also eliding the argument types: > > unsigned long n = std::accumulate( > xs.begin(), xs.end(), 0UL, > lambda(lhs, rhs) > { > return lhs + rhs.count; > } > ); > > Most of the time, you also do not know the type of the > arguments before hand. > Joel de Guzman There are still direct templates (even without implied ones.) The syntax would be a bit verbose though. unsigned long n = std::accumulate( xs.begin(), xs.end(), 0UL, template<class T, class S) auto (T const &lhs, S const &rhs) { return lhs + rhs.count; } );

1 0

circular_buffer v3.6
by Jan Gaspar 04 Mar '04

04 Mar '04

Hello all! New version of the circular_buffer is available. It can be found at http://groups.yahoo.com/group/boost/files/circular_buffer_v3.6.zip Release notes: - fixed bug regarding not destroyed elements - requirements for element types were changed to CopyConstructible concept - changed debug support classes - no dynamic memory allocation needed - changed debug support for invalidated iterators: assertion raised also when used iterator which points to the overwritten element - added allocator& get_allocator() /* NONCONST */ {...} method - added functionality to set_capacity() and resize() methods - changed circular_buffer_space_optimized constructor signatures Regards Jan __________________________________ Do you Yahoo!? Yahoo! Search - Find what you�re looking for faster http://search.yahoo.com

1 0

RE: [boost] GUI library - another one
by Jesse Booher 04 Mar '04

04 Mar '04

>>> hurdm(a)sig.com 03/03/04 10:30PM >>> >> Even the "why not" version seems overly complex compared to a console >> iostreams "Hello, world". Why not just: >> >> winout << title("Example window") << button("Hello, world!") >> << wait_for(delete_signal); >> >> I'm not necessarily pushing that exact approach, but just pointing out >> more >> C++-like alternatives. > > I understand you're not pushing that approach. It couldn't quite work > like that as you're trying to structure a tree rather than a sequence. Not that I'm pushing the approach either, but if all your objects that can be parents have an insertion operator then building a tree is simple a matter of adding parentheses: tree << a << b << © << d << e << (f << g) ); or breaking it up f << g; c << d << e << f; tree << a << b << c;

1 0