| |
| |
| |
|
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 30
Comments:
<0> if you want it to block, remove the timeout argument...
<1> What's the default value? Zero?
<0> toxik: no... None evaluate to False, in this case, like 0
<2> still IF that would've been the case, it isn't supposed to return it in the read list
<2> yeah yeah, whatever, I removed it anyway
<3> from the docs: "When the timeout argument is omitted the function blocks until at least one file descriptor is ready. A time-out value of zero specifies a poll and never blocks."
<4> CardinalFang: Default return value, supposing there isn't an explicit return? That would be None
<2> benji: still not the problem here
<1> Heh. No, the default timeout.
<5> Damn.
<5> Still my rendering sorting algothrim is not going well.
<0> CardinalFang: there's no default... it blocks until there's something ready
<2> that's the thing. it's returing a socket immediately
<0> toxik: the socket is ready then...
<2> no, it isn't
<2> calling accept() on it returns an error
<2> and it shouldn't be ready unless someone connected, and trust me, no one did
<2> not 10 times in a row in any given moment
<1> I'm accustomed to the Unix select() semantics. NULL = forever. 0 = poll, never block.
<6> hey all, if i upgraded new libmysqlclient to 4.1, how can I reconfigure the DBI so I can connect ot it w/o getting "Client does not support authentication protocol" ?
<2> CardinalFang: I was too, but apparently it's wrong - still I removed it so it doesn't matter
<4> CardinalFang: NULL _is_ 0, at least as far as C is concerned.
<0> toxik: no, it's right... I am wrong
<2> just leave that last paremeter subject, it's not what is being wrong
<1> Arafangion: It's a const struct timespec * .
<0> toxik: your socket is not listening... try calling listen(backlog) first
<2> werneck: oh .. silly me >_>
<2> I removed that startListener() crap and I didn't think of adding the listening part back :P sorry
<0> toxik: you forgot bind too
<2> yeah, that whole def is gone
<2> thanks again
<0> np
<2> werneck: It works, thanks a bunch again - truly gratefull
<1> L33TrDaNu: Why 4.1? Isn't 5.0 Teh Goodness?
<7> is there a function that can be called which gives a read-eval-print loop from stdin/out with the locals where where its called?
<6> CardinalFang: cause 4.1 is in debian sarge
<1> Ah.
<2> L33TrDaNu: this problem arises when your mysql client library is out of date
<2> L33TrDaNu: upgrade your libmysql
<6> CardinalFang: but.. i downlaoded python use apt-get install... i am thinking i have to recompile drivers for Python::DBI ?
<6> toxik, libmysqlcient is upgraded to use 4.1
<2> if they are linked; very probably
<6> toxik: ah ok, that makes sense I suppose
<1> L33TrDaNu: I'm not claiming _anything at all_ about that auth problem. i don't have a clue about it. I was merely curious about the version, for selfish reasons.
<2> yeah, I'm originally a PHP developer (which is why you are having all these weird questions from me :P) and I've seen that error A LOT of times, and I don't mean like once a day, I mean .. at least 10 times
<6> CardinalFang: you selfish bastard =) (jk
<6> )
<6> =)
<1> I'll have to embrace MySQL very soon, with a p***ion, so I'm just trying to get a feel for its usage.
<2> why?
<2> MySQL is a toy DBMS
<1> It was a toy. It's getting better. 5.0 is pretty good. 5.1 will be even better.
<2> how many times have I heard that too... :P
<2> I mean, just look at those laughable constraints
<1> I say this as a PostgreSQL devotee, who is starting work at MySQL AB in two weeks.
<2> Oh.. I guess you do have a clue then :P
<3> I disagree toxik, it is a tool that has its place (as all tools do)
<2> my impression of mysql aren't the best tbh
<1> toxik: Mine either. It was awful when I first used it, long ago. (~1996?) These days, it's as good as PostgreSQL in most areas, and better in some. It was playing catch-up. Now it's just p***ing it.
<1> Anyway -- sorry to hijack the topic.
<2> hijack away, I just want to discuss ;-)
<2> another python question then: how do I know if an object is an instance of this or that subcl***?
<1> isinstance() http://python.org/doc/lib/ #builtin functions
<8> hi to all
<6> CardinalFang: ok, problem is that , the server is running Debian stable and the developer is running Debian unstable
<2> argh, this is the third time I ask today: How do you unindent in vim?
<8> i have a GtkTextView i'd like to have a variable to store the current char number insered in the box... putting a signal on key pressed i had bad result any ideas?
<9> toxik: backspace?
<2> doesn't work
<1> L33TrDaNu: Newer client libraries should work with older servers, I would think. Sorry, but I doubt I'll be much help, yet.
<2> it just erases til the current indent level
<1> toxik: <<
<9> toxik: try using insert mode...
<2> AcidReign: arf.. I know what I'm talking about
<2> CardinalFang: Thanks a bunch
<2> hmm
<2> I have another question about subcl***ing sockets
<0> toxik: but, honestly... unless you really need it, don't check if an object is an instance of this or that cl***...
<0> toxik: instead, check if it has the method/attribute/interface you need...
<2> When you call accept from a subcl***ed socket, how do I then make that returned socket an instance of a subcl***ed cl*** instead of the original socket object?
<2> werneck: good idea
<1> We call it "duck typing". Check to see that the object you're holding has a bill, feathers, and webbed feet. Don't test its DNA for duckness.
<2> werneck: then issubcl***() is better?
<0> toxik: ops... that's why I used a BaseSocket cl*** instead of subcl***ing socket on that code
<2> ..oh.. :P
<2> werneck: and thus the reason for having a socket attribute?
<10> i'm looking for a parser for config files (forget "ConfigParser"). PyPi tells me "cfgparse", "config", "ConfigObj", "Dict4Ini", "tconfpy". any recommendations? i only need simple functionality and the config files should be human readable.
<11> wxPython cl*** for a single line text box?
<0> toxik: redefine the accept method on the listenerSocket subcl***, and inside it, wrap the accepted socket in a clientSocket instance
<2> so first overload the accept(), from it return a clientSocket(socket.accept())?
<3> biopyte, I've only used ConfigParser, why do you find it unacceptable?
<0> toxik: yes... that's it
<2> werneck: okay, I'll see if I can get that to work
<2> (shouldn't really be hard/long?)
<0> toxik: don't forget that socket.accept returns a (socket, addr) tuple
<2> good point
<0> toxik: no... it's easy...
<2> I got to 3 lines of code including the sign
<10> benji, maybe i have not understood the underlying principle of this module, but its making trouble all the time. if i read a config file, modify it and write it back i loose my data or have to apply "tricks".
<3> sorry, biopyte, can't help you there: "it works for me" :)
<10> biopyte, i'm messing around with it quiet long now.
<10> i mean, it should be trivial, not worth even mentioning
<2> hmm, issubcl***() can only operate on cl***es, how do I know if an object is a subcl*** of something then?
<10> benji, whats your basic code, if you do the following: "read config file, modify options, write back" ?
<12> took: an object isn't a subcl*** of something if it isn't a cl***
<12> took: i suspect you mean isinstance
<8> i have a GtkTextView i'd like to have a variable to store the current char number insered in the box... putting a signal on key pressed i had bad result any ideas?
<12> toxik: sorry, that was for you...
<0> toxik: isinstance(), but what exactly do you want ? check if a socket returned by select is a listener or client ?
<3> I don't do the "write back" bit, so I may not be running into the same problems you are, other than that I'm just using the published API in a straight-forward way
<2> werneck: exactly that
<2> joedj: yes, I'm having a hard time to express myself :\
<10> benji, you use it only for reading, not for writing options?
<3> right, humans write, sofware reads; I don't have a need for the sofware to write to the config
<13> what is the best python framework to do web applications (like rails for ruby) ?
<2> Ruby on Rails
<2> (is the name)
<10> benji, well, thats a different case. read-only works for me here, too.
<2> werneck: so I'm supposed to use isinstance() anyway?
<3> I do know that the built-in ConfigParser has some short-comings for that use case though, I'd recommend ConfigObject just because it seems to be the most popular
<14> Hmmm, anybody heard of managing CUPS with Python?
<14> I know there is a CUPS API, just wondering if there is a python module for it
<10> benji, k
<3> kristalino, it depends a lot on your app
<13> benji, the most similar to Rails, that's what i need
<0> toxik: yes, it's ok to use it... but as CardinalFang said, this is not the best solution... check for some attribute, not for the object's type...
<3> don't know that, kristalino; but that seems like a silly requirement
<2> werneck: like what, interface?
<1> toxik: if hasattr(critter, "quack") and hasattr(critter, "feathers") ...
<0> toxik: yes...
<2> well.. thinking of the lines of speed, just cheking for feathers would be OK as long as it's unique
<2> CardinalFang: wouldn't it be better to check for what I'm going to use? Like, f.ex. checking if it has an accept() method
<1> Give that man a Kwepie doll! He wins!
<0> toxik: that's the idea, but since both subcl*** socket, both have accept method
<15> ok i spent the last 3h trying to display a time nicely in c++... you can laught :)
<2> werneck: that's also a good thought
<15> and still not working :) some int64 issue i dont get :)
<2> that was smart..
<2> socket.accept() from listenerSocket.accept() calls itself >_>
<2> how would I call the parent then?
<1> Kl***name.method(self)
<2> yeah, and it decided to call itself instead
<1> That was my answer. What's the relationship between your objects/cl***es?
<2> http://deadbeefbabe.org/paste/2834
<2> it apparently decides to call the overriden method instead
Return to
#python
or
Go to some related
logs:
lun stuck at 0h problem
ubuntu dapper wireless wep 128
raaga on ubuntu
Debian + Marvell 88E1111
#web
z990 flops
#ubuntu
perl pros cons
Login.app XF86Config-4
#python
|
|