| |
| |
| |
|
Page: 1 2
Comments:
<0> hrm
<0> OK, I trashed my install of fink and have re-installed python, pyGTK2 and all's well
<0> but...
<0> I get this message
<0> from _gtk import *
<0> RuntimeError: could not open display
<0> oops, sorry hund
<1> I'm sure it's easier for people to help if you include more context in a paste.
<0> ok one sec
<0> http://rafb.net/paste/results/yR8AK555.html < this is what I tried running
<0> and I got a message from gtk saying it couldn't connect to display and I'm running OS X
<0> and yes I have/had X11 running
<0> now is there some option I need to p*** for it to connect to X11 ?
<1> And xterm or another X11 app from the same shell runs fine?
<0> let me check I'm pretty sure yea
<1> If so, I can't help much.
<0> if it didn't?
<1> Then I could tell you to ask an OSX person about getting X11 to work. :)
<0> what I do know is that the python compiler install from http://www.activestate.com/ had pythonw (for GUIs)
<0> but I used fink to grab the latest python instead :)
<0> ah heh...X11 works but hrm
<0> ok diff question
<0> how come X11 seems to load a different 'shell 'than oh duh brb
<0> ok right
<0> X11 seems to load a different .bashrc whatever (env variables
<0> i.e. if I run python there it loads up v2.3.5, In Terminal I get the new 2.4.2 (installed in /sw which is the Fink directory)
<1> Did you install pyGTK from Fink?
<0> yup
<1> Fink doesn't seem as streamlined as it should be, it would seem. People keep having trouble with it.
<2> yep
<2> the only exposure I have to it is people complaining about it here
<0> well it was straight forward tbh
<2> which they do frequently
<0> did a fink install python, then pygtk2-py24 whatever
<0> now I'm doing a fink of glade2
<0> woot got it working
<0> 'export display=:0' :D
<3> i love when you pull on the bowl/bong/whatever just right and a little smoke ring pops off the top of the bowl... or out of the carb...
<2> uh, that's nice
<4> does anyone remember a long time ago when I asked about getting errors for nonblocking sockets?
<4> I know how to do it now :p
<2> what did you do
<4> sock.getsockopt(socket.SOL_SOCKET, socket.SO_ERROR)
<2> get?
<4> apparently..
<4> this is what "man connect" told me to do
<5> ate..too..much
<1> Two fingers down your throat, then. :)
<5> exactly (ate chicken fingers)
<6> re
<5> I am ***uming this is invalid: popen2.popen2('cmd /c fix.bat')
<5> i mean, I know fix.bat works, and cmd /c works, hehe
<5> os.system will work
<5> (security isn't an issue if someone's about to respond with that)
<6> I don't know about popen2 on Windows really but are you sure the command line is p***ed to the shell?
<6> on Unix, it is
<6> but basically if it works similarly on Windows platform, you could just popen2.popen2 ("fix.bat")
<5> yeah, that didn't work with popen2
<6> or if I'm right, you could try popen2.popen2 ("c:\winnt\cmd.exe /c fix.bat")
<5> not sure why
<6> (with double backslashes, though)
<5> it doens't matter now, it works andI'm done with that client's bs
<5> :)
<6> ah, all right
<5> thanks though
<6> I was just guessing for fun then :)
<5> haha
<7> uhm.. how does that whole "p*** by reference" thing work in python? i have called a function with an empty list as a parameter, hoping to fill it up with data - but to no avail [vertices = [] \n load(vertices) <- yields empty set after load]
<6> basically everything is "p***ed by reference", or pointed to
<7> any idea on why that doesn't work then?
<4> when you do ***ignment on a variable, it replaces the variable with a reference to a new thing instead of modifying the old thing
<6> it should work; however, are you ***igning to the variable pointing to the list in load() ?
<6> def add (seq): seq.append (1)
<6> foo = []
<6> add (foo)
<6> print foo
<6> That prints [1]
<7> yes
<7> hmz.. i wrote some test-function, and it worked there.. :/
<6> then off you go narrowing down the problem :)
<7> the list i'm trying to fill is a list with lists, does that have consequence?
<6> basically no
<6> hard to say without seeing the code
<6> though
<7> i'll paste it in a sec..
<6> => http://rafb.net/paste/
<7> http://rafb.net/paste/results/WrCStt89.html
<7> the problem is at line 190 something - even accessing Edges[0][0] yields empty set
<7> but if i try accessing it at the end of the loadgraphfile-thingy - it works just fine
<6> you ***ign to Edges on line 150
<7> i'm sorry about the messy code, and whatever junk you may find.. it's my first attempt at python, if that makes it any easier to forgive :p
<7> yes. initially to -1 on line 150
<6> basically, when loadGraphFile() is called, a variable called "Edges" is _bound_ to the list you give to it as a parameter
<7> later on i fill the defined edges with weights, thus leaving each undefined edge as -1
<7> that was what i was hoping for
<6> if you then ***ign to "Edges", the variable now points to the new list
<7> uhm.. i don't follow
<6> the old list, given as a parameter, becomes inaccessible and isn't touched anymore
<7> in main i make a list, i then p*** it as a parameter to loadgraphfile
<7> in loadgraphfile i alter the list
<6> no, you first replace the value of "Edges" with a new list on line 150
<6> which is what you alter
<7> uh
<7> damn
<7> heh
<6> it'd be pretty easy to just return the filled in lists and never give any of them as parameters anyway
<7> in a tuple you mean?
<7> like "return (vertices, edges)"?
<6> or alternatively, you could rewrite line 150 as an iteration that appends a list of no_nodes "udef"s to the "Edges" list no_nodes times
<6> Yes. (Or just return vertices, edges
<6> )
<6> and then call it like:
<7> indeed - that was kind of embarr***ing :S
<6> Vertices, Edges = loadGraphFile (sys.argv[1])
<7> i'll do it like that i guess - seems more appropriate
<6> we can go through the appending thing too, but it feels unpythonic
<6> i've never probably done that myself in Python. Very commonly in C++ though :)
<7> i'm not following?
<7> what's with the append() thing?
<6> well, instead of line 150 you could do something like:
<6> (if I remember correctly, I closed the browser already)
<8> how does python free objects that are circuarily referenced.. it just has reference counts..?
<6> for i in range (no_nodes):
<6> Edges.append ([udef for _ in range (no_nodes)])
<7> yes, but that's the same thing - right?
<7> except you can p*** edges by reference and fill it like that
<6> Neutrino: not the same thing. In the latter version you indeed modify the original list p***ed in as a parameter. The "return" approach creates a new list which is then returned
<6> ludde: by following the references in order to track whether a set of objects refer to themselves only
<6> (and not any GC root)
<7> hmm.. i misunderstood your intention about stating " go through append thing - feels unpythonic" - i thought you felt that my making of "edges" inside loadgraphfile was unpythonic, not the p***ing of an array to be filled
<6> yeah
<8> yason: I thought python didn't have a garbage collector as-such?
<8> that sounds like a garbage collector
<6> Neutrino: p***ing in arrays is unpythonic. Additionally, it means that the function has side-effects which is generally bad
<7> uh.. i like side-effects :p
<6> ludde: Python has a GC based on reference counting with cycle detector
<8> okay
<6> Neutrino: side-effects make your functions functions of the calling state
<7> i've altered it to return a tuple - it works fine now :)
<6> Neutrino: it's a good way to work :)
<7> thank you for your help btw
<7> :)
<6> np
<6> once in a while I learn something myself too
<6> while teaching
<1> Learning by teaching is underrated.
<6> Ditto. Propelled by sayings such as: "Those who can, do. Those who can't, teach."
<6> It's true in a sense: if you already know something really well, you're better off doing it and you'd probably be too hasty to teach well anyway
<1> Figuring something out because someone's waiting is far more effective for me than figuring it out because I'll need it in an exam two months later.
<6> but everybody should have lots of things "on the works" all the time, making them good candidate subjects for teaching (and thus learning more)
<6> Yes. And sometimes you think you know something but once you start explaining it, you'll find a lot of questions yourself first
Return to
#python
or
Go to some related
logs:
scramble 8300hd
#computers
#computers
#windows
#stocks
#sex
#worldcup
#firebird
#dsl
extra time for ejectulation
|
|