| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Comments:
<0> Go bug #php.
<1> joples: I have played with it and the CGI module has support for them
<2> robkinyon: There's always another level higher. Well, maybe not above cheezychat. That's almost Bitch->dwim().
<3> or write your porgram in perl ;)
<1> joples: but that was a long long time ago
<4> dngor: Heh. I guess what I'm saying is define the idea of a message
<4> that is completely abstracted from the sender or the recipient
<5> sorry...
<6> Ikarus: ok, i will go through the documentation
<6> Ikarus: thanks a lot
<4> that's probably 80%
<1> joples: np, Server Side Push is still fun to toy with
<4> then, each framework implements how messages are actually implemented, but the users of the framework still only use the abstract message
<6> Ikarus: sounds like great way to implement a progress bar :]
<2> robkinyon: My current thought on the matter is that messages and frameworks need some coupling to be effective. In POE::Stage, they work together to provide something like continuations, although the stages (not the messages) do most of that work.
<4> dngor: The idea here is that you have the concept of a message and the concept of an entity that can send and/or receive messages
<4> implementing that will give you some scaffolding upon which to build something
<7> I "use" Bloglines. That is to say, I created an account, added a thousand feeds, made sure to read everything the next day, and never logged in again because of all the unread items.
<4> be extremely simple - Stem works on the idea that there is a method corresponding to the message that is called. If it returns something, then that's the response
<4> nothing fancy at all
<4> but, it's extremely flexible to build something more on top of it
<8> something like IKC implemented under POE::Stage is what sounded interesting at the time i was looking around for something
<9> I have %hash = foo(); How can I make foo return so I can do a check if (!%hash) ?
<8> too many somethings:)
<10> Aankhen``, ever try anything else?
<7> jsn: Nah.
<9> if I return 0 it complains ofc
<11> balony: return an emtpy list
<9> LeoNerd: %hash = (); ?
<11> balony: You have to return a list. Moreover, since you're using it as a hash, an even-sized list
<2> robkinyon: Sounds rather basic. I'm doing something a little more complex. Messages define return paths for different types of response. An object doesn't need to ***ume anything about its caller's interface.
<11> "return" on its own is usually best - () in list context, undef in scalar
<11> Both of these things are false
<4> balony: "return;"
<9> ah, ok
<4> dngor: Start with simple, then build upon it.
<9> im used to having "null" or NULL :)
<4> the key is to have the terminology in place before you get complicated
<4> that's the major design win I'm seeing in Stem
<12> GumbyBRAIN: why don't you make yourself at home and stay with me
<13> Let me make it yourself.
<14> GumbyBRAIN: hi
<13> Http://www.telegraph.co.uk/1/hi/england/southern_counties/5081744.stm - i love it.
<10> I want a swanky desktop blog reader which will merge all the feeds and stream them by in one window. If I click on one, it should open to expanded content
<10> :-D
<2> robkinyon: I always do, but at some point it expands so that the majority of the issues are edge cases. That's why I abandoned three prototypes.
<14> jsn: z0mg ajax
<4> dngor: If you're hitting a bunch of edge cases with messaging, then you're getting too complicated
<4> Stem's messaging protocol is implementation-independent
<4> he claims that his messages can communicate over SOAP, XML-RPC, straight TCP/IP, or UDP
<10> sili, I don't want it to be as heavy as a browser
<2> Oh, that I'm not worried about. I mean use cases.
<4> or IPC
<4> usecase: "Provide a mechanism by which to call a method with arguments in some other object"
<4> YAGNI on the rest
<10> sili: Don't be mean, I am pretty serious about writing for the internet, and although i've been doing it for years, I never bothered with a RSS aggregator. Now I want one.
<2> I totally disagree. POE's been used a LOT over the years, and people need a lot of stuff. Gloss over it at your peril.
<14> jsn: i like lisp.
<4> dngor: there's two distros involved
<4> one is the actual implementation
<4> the other is the high-level outline that the implementation "implements"
<10> sili: Probably a sympton of your unhealthy addiction to emacs. You are still young, n?vim? while there is still time.
<4> i'm speaking of the latter
<8> sili: what was the last app you wrote in lisp?
<10> I guess I'd be forced to do (n?vi|vim) for a perl regex to match all three names (vi, vim, nvi). Is there a better way?
<8> nvi?
<2> robkinyon: Oh, I'm talking about actual code.
<15> POE is a lot like perl at this point. It does stuff people need at the cost of inelegance at some level--but the cost hasn't been high enough to throw it away and start over.
<2> Yaakov: No need to throw it away. It's pluggable at nearly every level.
<8> are you going to moose it?
<4> don't moose POE yet
<4> wait about 3 months
<8> :)
<15> Well, that's part of the problem. Some things PROBABLY ought to migrate to core DESIGN. But the thing is, it isn't a good argument AGAINST POE, nor does it seem to be enough to really get people going on something to supercede it. So, in other words, POE's warts don't matter to people looking to get stuff done.
<12> Yaakov: except if the warts get in the way!
<12> which is rare
<15> Well, yes--that's what I am saying... POE has evolved into what people need.
<2> robkinyon: I'm considering Moose'ing POE::Stage, which doesn't have widespread adoption yet. What's happening in 3 months?
<4> dngor: It'll be more solid
<16> What's moose?
<4> Moose is still in the "early adopter" phase, according to stevan
<4> retardo: Look on CPAN
<16> Fine ,but what is it?
<2> Oh. Sounds like API shifting.
<4> dngor: Not really (we use it at $work), but stevan wants to leave his options open in case something crazy comes up
<4> don't anticipate it, but better to have people wait three months
<4> the main API regarding cl*** definition is stable and has been for several months
<2> POE::Stage doesn't need Moosey joy. Maybe I should put away the paint can. :)
<4> he's working on optimizations and adding the ability to close cl***es.
<4> Catalyst and CGI::Application are both moving to Moose/Cl***::MOP, so it's mostly stable
<4> but, it's better to wait until them (and Jifty) have been using it and figured out the edge cases for you
<2> Yeah. I first heard about Moose from the Catalyst people who hang in irc.perl.org #poe.
<17> why am I returning a scalar with this string... it's within a foreach loop... my $site_name = grep(/\w:/, $_);
<15> Why are you using grep?
<16> StylusEater_Work: If you write code that ***igns the result of grep to a scalar variable, then grep will return a scalar value.
<17> doh!
<16> If you want grep to return a list, you have to supply a list or an array for it to go into.
<12> GumbyBRAIN: lists arrays and scalars oh my!
<13> Oh.. Kind of like my sources to a variable (need it for arrays in perl? Is dbd quote() function anymore coz the place they are transmiting them.
<17> I have the following: http://rafb.net/paste/results/9jydmj25.html
<18> The paste 9jydmj25 has been moved to http://erxz.com/pb/1843
<15> StylusEater_Work: It _looks_ funny to use grep in that context. What do you WANT it to return?
<12> looks to me like he's making a log parser
<17> the string that matches \w:
<17> simcop2387: yes
<12> StylusEater_Work: grep doesn't do that
<19> GumbyBRAIN, stupid people
<13> does it have a job in a group of people.
<12> hey Elisabeth
<15> StylusEater_Work: m// with ().
<19> hey simcop2387
<17> I guess it just returns a scalar 0 or 1 yes?
<17> if it finds the pattern
<12> nope
<12> perldoc -f grep
<18> Type 'perldoc -f grep' in your shell or go to http://perldoc.perl.org/functions/grep.html
<16> I believe it returns the number of mathcing items.
<15> StylusEater_Work: perldoc perlretut
<18> StylusEater_Work: Type 'perldoc perlretut' in your shell or go to http://perldoc.perl.org/perlretut.html
<17> bbiab
<10> It returns a list of strings, not a number of matches
<10> eval: print join "," grep { /abc/ } qw(Dabc Ccba Babc Aabb)
<18> jsn: Error: syntax error at (eval 134) line 15, near ""," grep"
<10> eval: print join ",", grep { /abc/ } qw(Dabc Ccba Babc Aabb)
<18> jsn: Dabc,Babc 1
<17> how would you recommend I strip a trailing : ???
<17> or better yet...everything after a :
<16> $string =~ s/:.*//;
<15> Hello, retardo version of MJD.
<16> Hellow, Yaakov version of Yaakov
<15> Heh.
<15> retardo: What is Moose?
<16> I don't know.
<15> Hmm... probably isn't too important then!
<16> It's another OOP system.
<14> oops.
<12> GumbyBRAIN: silly sili sings saprano
<13> Sili do you want to leave in a long shot.
<14> ok
<14> GumbyBRAIN: sili
<13> just make a db_file, because the chances of holland/ivorians being decent.
<12> he doesn't respond too well to alliteration
<20> GumbyBRAIN: take it to the fridge!
<13> Rounded up to in the fridge.
<3> GumbyBRAIN: rounding error
<13> (Which is the error and error and won't install.
<12> GumbyBRAIN: we don't get hindu's harmonizing in our halls
<13> Oh, the ^x prefix is in our halls.
<17> so how about this now... http://rafb.net/paste/results/sYJgkw32.html
<18> The paste sYJgkw32 has been moved to http://erxz.com/pb/1844
Return to
#perl
or
Go to some related
logs:
KDE... configure: error:
in the prefix, you've chosen, are no KDE headers insta
ubunu live cd mount
ubuntu apt-get tcp-wrapper
perl INPUT_RECORD_SEPERATOR
Get directory path from FILE* c++
checking battery status ubuntu
#python
channel 2: open failed: administratively prohibited: open failed cvs
#mysql
#perl
|
|