Pike-devel June 2012

pike-devel@lists.lysator.liu.se

7 participants
18 discussions

Pike 7.8 Release (last call)
by H. William Welliver III 12 Jun '12

12 Jun '12

Does anyone have any objections to 7.8.680 becoming the next official release of Pike? Do note that machine code is non-functional on (at least more modern instances of) i386. If I don't hear any objections in 24 hours, I will attempt to update the website, etc, with the new information. Best, Bill

9 18

Starting with USB() instead of USB.devices()
by Stephen R. van den Berg 10 Jun '12

10 Jun '12

I now have a cmod which is in a subdir src/post_modules/USB, which gives me an USB.devices(), USB.Iterator(), USB.device() and USB.transfer(). The usual way this works, is that one starts with: object dev=USB.devices(); and then you do a foreach() which gets the Iterator, which delivers the devices, etc. Now, what if I'd like it to start with: object dev=USB(); (and then the rest with foreach() etc. same as above). Are there any aesthetical or technical reasons not to prefer this? (I'm all for brevity). If there are no obvious objections, what would be the easiest way to implement this? Is this doable in the cmod? Are there any existing modules that did it this way? -- Stephen. "There are three types of people in this world: those who make things happen, those who watch things happen and those who wonder what happened."

2 4

PIKEVAR object foo;
by Stephen R. van den Berg 09 Jun '12

09 Jun '12

If I declare this inside a PIKECLASS in a .cmod, I need the canonical: PIKEFUN void setfoo(void|object foo) { if(THIS->foo) free_object(THIS->foo); if((THIS->foo=foo)) { debug_malloc_touch(THIS->foo); Pike_sp--; /* don't decrease references */ } else pop_stack(); } to set it. The question is, should I also free_object() it in the destructor, or does Pike already take care of that somewhere? -- Stephen. How many weeks are there in a lightyear?

4 10

quick_add_*_constant() not so quick
by Stephen R. van den Berg 05 Jun '12

05 Jun '12

Is there a reason why quick_add_integer_constant takes a char* and a length, instead of a (much quicker and cleaner) struct pike_string* which would enable it to work more efficiently in collaboration with MK_STRING() ? I see that the functions do not seem to be used anywhere, so I propose to change the (internal) interface definition for them to actually make them quick. -- Stephen. "There are 10 types of people in the world. Those who understand binary and those who do not."

3 5

IP options
by Martin Nilsson (Opera Mini - AFK!) ＠ Pike (-) developers forum 05 Jun '12

05 Jun '12

I'm experimenting with IP options and have added the following to the file object. void file_get_ip_options(INT32 args) { char buffer[44]; socklen_t len = sizeof(buffer); int ret; pop_n_elems(args); ret = getsockopt(FD, IPPROTO_IP, IP_OPTIONS, buffer, &len); if(ret!=0) { push_int(0); return; } push_string( make_shared_binary_string( buffer, len ) ); } void file_set_ip_options(INT32 args) { int ret; if(args!=1 || TYPEOF(Pike_sp[-args]) != PIKE_T_STRING) Pike_error("Illegal arguments to set_ip_options.\n"); ret = setsockopt(FD, IPPROTO_IP, IP_OPTIONS, Pike_sp[-args].u.string->str, Pike_sp[-args].u.string->len); pop_n_elems(args); push_int( ret==0 ); } I can successfully set the NOP IP options with Stdio.File f = Stdio.File(); f->open_socket(); f->set_ip_options("\1"); and can see it set in wireshark. However, I'm unable to read it out on the server side. Is it my poor understanding of C or of network code that is the problem?

1 0

Threading in callbacks
by Stephen R. van den Berg 05 Jun '12

05 Jun '12

Suppose my libusb does a callback to the C callback function I provided, which, in turn, will want to run a Pike callback function the user supplied. What if the usblibrary decides to call my callback during a time where threads are allowed? Is there any way for me to detect this? What to do about this? Conditionally run something like: static void usbcallback(...) { if(threadsareallowed) THREADS_DISALLOW(); -- Run the Pike callback function. -- if(threadsareallowed) THREADS_ALLOW(); } Is this even going to work? -- Stephen. "There are 10 types of people in the world. Those who understand binary and those who do not."

2 2

cmod parser and two methods with the same name
by Stephen R. van den Berg 04 Jun '12

04 Jun '12

It throws the parser into an infinite recursion, and results is rather cryptic error messages. It's not high priority, but if anyone knows his way around that thing, it might be worth checking for this. -- Stephen. "Very funny, Mr. Scott. Now beam down my clothes!"

2 3

libusb and autoconf
by Stephen R. van den Berg 03 Jun '12

03 Jun '12

Just in case someone happens to be working on it, I'm in the process of implementing a native interface to libusb-1.0 from within Pike. The sad part is that actually programming the interface is not that much work, generating proper configure.in files *is* however a big PITA. Has anyone ever considered using cmake or jam instead of autoconf ? I'm not saying I know it to be better, but it just seems like it can't get any worse either (besides simplicity, it definitely could be faster, the current configure business is a resource hog, timewise). -- Stephen. "Always remember that you are unique. Just like everyone else."

3 10

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Pike-devel June 2012