Pike-devel

pike-devel@lists.lysator.liu.se

2 participants
3046 discussions

Multiple function prototype declarations
by Stephen R. van den Berg 23 Sep '16

23 Sep '16

I seem to recall that in C, one can define multiple function declarations for the same function. Is it possible to do the same at Pike level? Example, I have a function declared as void foo(mixed ... bar); Now I want to check, and enforce two different declarations: void foo(string s, mixed ... bar); void foo(function f, string s, mixed ... bar); Can I, at Pike level, declare somehow that the above two prototypes are the only two valid prototypes for this function? So that Pike will give me a syntax error if I try to do foo(23); ? -- Stephen.

2 2

Re: Setting socket options: so many, uhh, options!
by Chris Angelico 15 Sep '16

15 Sep '16

On Tue, Mar 10, 2015 at 10:05 PM, Mirar @ Pike developers forum <10353(a)lyskom.lysator.liu.se> wrote: > Wait, I said REUSEPORT? What's the difference to REUSEADDR? Mysteries > of TCP sockets... Here's a decent explanation, I think: http://stackoverflow.com/questions/14388706/socket-options-so-reuseaddr-and… ChrisA

3 6

Compile error in 8.1
by Pontus Östlund 14 Sep '16

14 Sep '16

Hi guys! Does anyone know how to solve this? make[2]: *** No rule to make target '/home/poppa/src/pike-git/devel/src/hashtable.h', needed by 'cpp.o'. Stop. Makefile:149: recipe for target '_make_in_builddir' failed make[1]: *** [_make_in_builddir] Error 2 Makefile:66: recipe for target 'compile' failed make: *** [compile] Error 2 It seems to have happend after commit dcf57dec52f15cda0dc07bbaf2198a250b37030c Regards ----------------------------- Pontus Östlund Developer • Roxen AB +46 70-662 81 69 www.roxen.com <http://www.roxen.com/> | twitter.com/roxen <https://twitter.com/roxen>

2 3

Winsock error 10038 when backending sockets and stdin
by Chris Angelico 12 Sep '16

12 Sep '16

On Unix, you can happily select() on any file descriptors, including stdin and sockets. On Windows, that doesn't appear to be the case. Here's a simple form of netcat: object sock; void sockread(mixed id, string data) { //Restrict to ASCII printables for simplicity write(filter(data, lambda(int c) {return has_value("\t\r\n", c) || (c >= 32 && c < 128);})); } void console(mixed id, string data) {sock->write(data);} int main() { sock = Stdio.File(); sock->connect("minstrelhall.com", 221); sock->set_read_callback(sockread); Stdio.stdin->set_read_callback(console); return -1; } The important aspects of this are read callbacks on both a socket and Stdio.stdin, and returning -1 from main. Comment out the stdin callback and all is well. Keep it in, and the program runs fine on Unix, but on Windows, bombs out: /home/zino/hack/pike8-rel/pike/src/backend.cmod:5496: Fatal error: Filedescriptor 0 (IS CONSOLE) caused fatal error 10038 in backend. Backtrace at time of fatal: -:1: Pike.Backend(0)->`()(3600.0) This happens with both 7.8 and 8.0 stables on Windows 7, though I doubt it'll be any different on other Pikes or Windowses. Is there any good way to respond to console and sockets? For now, I think I'll just write this off as "Windows sucks" and disable the console (it's only a debug/emergency feature anyway). ChrisA

1 0

Documentation breakage
by Magnus Holmgren, Millnet/Lysator/Debian/Mensa ＠ Pike developers forum 29 Jul '16

29 Jul '16

Just to make sure we're talking about the same thing, the patch I posted has been applied, but the patch I didn't post hasn't. 5c6616146a1f1cdb18a8ed6f973a63ee73df9278 is the current version of Sign.pike on the 8.0 branch, isn't it? The fix on the 8.1 branch is in b238c808a70a4c671a2a7ecbdd083e41f469f41b. --- a/lib/modules/__builtin.pmod/Nettle.pmod/Sign.pike +++ b/lib/modules/__builtin.pmod/Nettle.pmod/Sign.pike @@ -99,9 +99,9 @@ string(7bit) jose_sign(string(8bit) mess //! @returns //! Returns @expr{0@} (zero) on failure, and an array //! @array -//! @item mapping(string(7bit):string(7bit)|int) 0 +//! @elem mapping(string(7bit):string(7bit)|int) 0 //! The JOSE header. -//! @item string(8bit) 1 +//! @elem string(8bit) 1 //! The signed message. //! @endarray //!

1 0

Documentation breakage
by Magnus Holmgren, Millnet/Lysator/Debian/Mensa ＠ Pike developers forum 29 Jul '16

29 Jul '16

When preparing a new Debian package of 8.0.240, I noticed that some documentation disappeared, apparently due to the following. The error in Nettle.Sign is already corrected but not the others, AFAICT, and I'm guessing that @exp should be @expr. Patch below, which I home someone can apply right away. ERROR: <Invalid error container: Tools.AutoDoc.AutoDocError(SourcePosition(File: /build/pike8.0-8.0.240/src/modules/Gmp/mpz_glue.c, lines: 668..668), "DocParser", "@exp cannot be used like this: @exp{ ... @}")> ERROR: <Invalid error container: Tools.AutoDoc.AutoDocError(SourcePosition(File: /build/pike8.0-8.0.240/lib/modules/__builtin.pmod/Nettle.pmod/Sign.pike, lines: 102..102), "DocParser", "@item is not allowed inside @array (allowed children are: @elem)")> ERROR: <Invalid error container: Tools.AutoDoc.AutoDocError(SourcePosition(File: /build/pike8.0-8.0.240/lib/modules/Crypto.pmod/ECC.pmod, lines: 286..286), "DocParser", "@item is not allowed inside @array (allowed children are: @elem)")> ERROR: <Invalid error container: Tools.AutoDoc.AutoDocError(SourcePosition(File: /build/pike8.0-8.0.240/lib/modules/Crypto.pmod/RSA.pmod, lines: 368..368), "DocParser", "@item is not allowed inside @array (allowed children are: @elem)")> --- a/lib/modules/Crypto.pmod/ECC.pmod +++ b/lib/modules/Crypto.pmod/ECC.pmod @@ -283,9 +283,9 @@ class Curve { //! @returns //! Returns @expr{0@} (zero) on failure, and an array //! @array - //! @item mapping(string(7bit):string(7bit)|int) 0 + //! @elem mapping(string(7bit):string(7bit)|int) 0 //! The JOSE header. - //! @item string(8bit) 1 + //! @elem string(8bit) 1 //! The signed message. //! @endarray //! --- a/lib/modules/Crypto.pmod/RSA.pmod +++ b/lib/modules/Crypto.pmod/RSA.pmod @@ -365,9 +365,9 @@ class State { //! @returns //! Returns @expr{0@} (zero) on failure, and an array //! @array - //! @item mapping(string(7bit):string(7bit)|int) 0 + //! @elem mapping(string(7bit):string(7bit)|int) 0 //! The JOSE header. - //! @item string(8bit) 1 + //! @elem string(8bit) 1 //! The signed message. //! @endarray //! --- a/src/modules/Gmp/mpz_glue.c +++ b/src/modules/Gmp/mpz_glue.c @@ -665,7 +665,7 @@ static void mpzmod_create(INT32 args) /*! @decl int cast_to_int() *! Casts the object to an integer. *! @deprecated - *! Use @exp{(int)@} instead. + *! Use @expr{(int)@} instead. */ static void mpzmod_get_int(INT32 args) {

2 2

Re: Notes from the impromptu developer meetup 2016-07-07
by Chris Angelico 21 Jul '16

21 Jul '16

On Fri, Jul 22, 2016 at 1:05 AM, Martin Bähr <mbaehr(a)realss.com> wrote: > Excerpts from Chris Angelico's message of 2016-07-22 00:46:18 +1000: >> Hah. Actually, I've been pushing more people onto Pike who aren't >> developers, and unfortunately there are a number of trickinesses to >> working on Windows. I've largely come to the conclusion that "Windows >> sucks" is a valid excuse for nearly any bug that I can't repro on >> Linux, but for all those non-developers who want stuff to "just work", >> getting them to install Debian might be a lot of hassle... > > would the linux environment in windows 10 help with that? > pike should work on it, though gtk support won't (unless you get an X > server) so maybe a dealbreaker for you :-( > Lacking GTK would be a killer; so would any additional complexity that lands on the end user, as they're already tending to balk at "install this .msi, then get this .zip and extract it, then double-click gypsum.pike". Also, doesn't help with all the people on Win 7/8. So it's probably going to stay with the native Windows builds for now. That said, though, I don't think the glob changes would break anything I do. ChrisA

1 0

Notes from the impromptu developer meetup 2016-07-07
by Peter Bortas 21 Jul '16

21 Jul '16

We had that impromptu Pike dev meetup on the 7th, and I took notes on some of the things we discussed. Dumping them here before we all forget about it. In attendence: Zino, Nilsson, Grubba, Marcus, Tobi, Arne Status report ------------- Per tradition grubba made a short status report about what has changed since the last meeting: - Some non-enabled new SSL server stuff is not yet compatible with Chrome - We enforce C89 now - GMP is required and always loaded in 8.1 - Threads are allowed while compiling in 8.0 - pragma #dissasemble/#dissassemble (new in 8.1) 8.2 TODO -------- We went though what everyone wants in before we split branch 8.2. In no particular order: - Compiles on Windows For now and until he protests this is zino's problem. - Something about Mpz. Don't remember the details. - Something about lfuns proxy-requested by Per. Don't remember the details. - Add MPI support zinos project. Not really a blocker. - Stop using declarations on the top of the function in C? Debatable - Get the fastcall, fast frames improvements in - glob("C:\*.gif") alt 1: Add Future.glob() depricate glob() import Future alt 2: Consider this a Windows problem and call up the 3 Windows users by phone. alt 3: //Never do this Add heuristics. alt 4: add 8.2::glob dspecial depricate glob constant glob=8.2::glob alt 5: Add String.glob() or similar depricate glob() Maybe reintroduce glob() in a later release We where leaning towards 5. - Stop dumping modules - Move/Remove the Markdown non-module Random notes ------------ - Make a list of all modules and make a note about why each of them are included. Remove the ones with no motivation for them. Nilsson is responsible for the list-making. - Check if anything runs on eureka, otherwise, kill it - Check if we have anything on pelix, otherwise kill it - TravisCI / Coverage - Opera Critic (https://github.com/jensl/critic) Someone said they would have a look att setting that up for Pike last meeting. No one could remember who. - Xenofarm Dates are broken somewhere on the buildmachine which prevents some of the clients from building. - _Roxen module, move the useful stuff somewhere and kill it Make one canonical HTTP module. - Rewrite dirnodes to make them have fewer confusing edge-cases nilsson, grubba - dot-operator is confusing as hell - Use merges to make it easier to track that bugfixes has been merged in which branch Arne will check. - Remove configurability of Calendar Nisson - Make Calendar discoverable for users - Take a look at Image.Orient - Take a look at Image.Graph - Documentation for stuff that inherits ::this_program - Ticket support? (I don't remember what this was about at all) -- Peter Bortas

2 1

Pike 8.0 rel 4 is out
by Peter Bortas 11 Jul '16

11 Jul '16

Pike 8.0 rel 4 is now out. https://pike.lysator.liu.se/pub/pike/latest-stable/Pike-v8.0.276.tar.gz Other builds: https://pike.lysator.liu.se/pub/pike/latest-stable/Pike-v8.0.276-Darwin-15.… https://pike.lysator.liu.se/pub/pike/latest-stable/Pike-v8.0.276-win32-oldl… Changes since Pike 8.0.240 (release 3) ---------------------------------------------------------------------- New features ------------ o _Roxen.HeaderParser feed() can now be forced not to normalize (ie lower_case) the header names. o Protocols.WebSocket Added client mode. Bug fixes --------- o Documentation Fixed some AutoDoc markup typos. o Fixed assertion failure in multiset handling. Indexing of non-empty multisets containing no objects or functions, with an object implementing lfun::`<() et al could cause assertion failures. o Filesystem.Monitor Fixed potential recursive mutex locks on destroy(). Regards, -- Peter Bortas

1 0

f_hash() and endianess
by Arne Goedeke 06 Jul '16

06 Jul '16

The documentation of the new f_hash() claims that its return value is architecture independent. SipHash reads its input as a series of little-endian 64 integers, which means that the hash value depends on endianess for wide strings. I have at some point written byte order independent versions of siphash for both 16 and 32 bit strings. Those can be found in commit 21f218e51d55d8c07aff9db0cf45e4ac83050a5e as part of the bloom filter branch. If nobody objects, I will merge that commit into pike 8.1 and integrate it with f_hash(). Something unrelated: I am going to be in linköping from tomorrow until saturday next week. Anyone up for a mini pike meetup? Arne

2 3

Jump to page:

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