Pike-devel October 2004

pike-devel@lists.lysator.liu.se

12 participants
49 discussions

Facets
by Martin Nilsson (DivX Networks) ＠ Pike (-) developers forum 23 Oct '04

23 Oct '04

Mikael Amborns facet implementation is finally checked into Pike. Facets allows you to enforce that a multiple inheritance is used when functionality is split into several classes. One current misfeature is that the facet group object needs to exist before (e.g. loader wrapper) or be created during compilation (e.g. .pmod file). Usage example: > add_constant( "Bunny", FacetGroup() ); (1) Result: 0 > class White { facet Color:Bunny; } > class Black { facet Color:Bunny; } > class Small { facet Size:Bunny; } > class Large { facet Size:Bunny; } > class Snuffy { inherit White; inherit Small; } > class Sooty { inherit Black; } Compiler Error: 1:Error in some product classes Compiler Exception: Product class does not inherit from all facets. _static_modules.Builtin()->facet_group: _static_modules.Builtin()->facet_group()->check_product_classes() > class Checkers { inherit Black; inherit White; } Compiler Error: 1:Unable to add product class Compiler Exception: Product class can only inherit from one class in every facet. _static_modules.Builtin()->facet_group: _static_modules.Builtin()->facet_group()->add_product_class(65713,0,65701) Another problem is that it leaks all kinds of stuff (parser stack, objects and strings)...

4 12

Bug in precompile.pike
by Marcus Agehall (PacketFront) ＠ Pike (-) developers forum 22 Oct '04

22 Oct '04

There seems to be a bug in precompile.pike in 7.7. I checked out a fresh copy from CVS this morning and it failed to compile the SQLite module which Nilsson checked in last night. I looked at the code and found that on line 145, it does store = GET_RESOBJ_STORAGE(res); which the C-compiler later complains about. Some investigation shows that precompile.pike gives the following define for that macro: #define GET_RESOBJ_STORAGE ((struct ResObj_struct *)(o->storage+ResObj_storage_offset) which is almost the same as the OBJ2_RESOBJ macro, besides that OBJ2_RESOBJ takes a parameter (o) and has balanced parenthesis. I do not understand the intended difference between these defines and thus I will leave it up to someone else to sort this out. For now, I'll change the code inte SQLite module to use the OBJ2_RESOBJ() macro.

2 4

C documentation
by Martin Nilsson (DivX Networks) ＠ Pike (-) developers forum 21 Oct '04

21 Oct '04

One topic discussed at the Pike Camp was documentation. With the beginners book just around the corner and the module tree documentation almost complete, the lack of documentation of the C API was deemed the most important area to improved. Thus we decided to immideatly try out doxygen for this, and as an experiment we documented array.[ch] and generated this: http://pike.ida.liu.se/generated/manual/c_api/html/index.html To not clash with autodoc, only the /** syntax is used, and not the /*! one. Comments?

4 4

Ncurses
by Kaos 19 Oct '04

19 Oct '04

Howdi, I've tried to make a relase on modules.gotpike.org, but it keeps saying I've not provided any source when uploading my tar-archive. So, anyone interested to see what've been done this far, please use the following link: http://kaos.explosive.se/files/Stdio_Ncurses-0.4.tar.gz There's still alot to do.. but basic operation of ncurses should work. Tested (somewhat) on both 7.6.25 and 7.7.9. For a compiled refdoc, refer to: http://www.explosive.se/~kaos/doc/ncurses/ex/predef_3A_3A/Stdio/Ncurses.html // Andreas

2 2

Roadmap
by David Gourdelier 19 Oct '04

19 Oct '04

Hello, I have a few simple questions regarding Pike developpment: What is the roadmap for Pike developpement? If yes, can it be published on the Pike website? If there is no roadmap, can it be discussed on this mailing list ? / David

2 1

Problems with SDL.h
by Marcus Agehall (PacketFront) ＠ Pike (-) developers forum 16 Oct '04

16 Oct '04

Several times over the last 24h my buildtree has failed to compile, complaining about the SDL.h file which it doesn't know how to generate. Apparently, SDL.h is marked as a dependency in the SDL module and make apparently doesn't look for the dependencies in the standard include dirs. Removing the builddir and rebuilding from scratch usually works. Anyone else seen this problem?

1 0

Strange segfault
by David Gourdelier 16 Oct '04

16 Oct '04

Hi, I have a strange backtrace with Pike 7.6.25 that I didn't see before: #0 0x080e3e22 in gc_mark_object_as_referenced (o=0x99a00b3) at object.c:1721 1721 DOUBLEUNLINK(first_object, o); (gdb) bt #0 0x080e3e22 in gc_mark_object_as_referenced (o=0x99a00b3) at object.c:1721 #1 0x08111263 in run_queue (q=0x823d2e8) at queue.c:43 #2 0x080b42ac in do_gc (ignored=0x82cc4f8, explicit_call=0) at gc.c:2844 #3 0x0809e34b in low_call_callback (lst=0x823d254, arg=0x0) at callback.c:167 #4 0x08074f38 in low_mega_apply (type=APPLY_LOW, args=1, arg1=0x8415b3c, arg2=0x3) at apply_low.h:215 #5 0x08072f35 in jump_opcode_F_CALL_OTHER (arg1=1) at interpret_functions.h:1941 #6 0x0844b6ea in ?? () #7 0x00000001 in ?? () #8 0x00000001 in ?? () #9 0x0842caa0 in ?? () #10 0x0842caa8 in ?? () #11 0x40c3268c in ?? () #12 0x40c32614 in ?? () #13 0x08136d0f in append_path_unix (s=0x0, path= {ptr = 0x842caa8 "\004\211\025(\002", shift = 1086531212}, len=138594976) at combine_path.h:263 #14 0x080761d4 in o_catch ( pc=0x844b569 "¡0Ò#\b\203@\034\024\213H$\213\035 Ò#\b\213Q\004\211S\004\213\t\211\vf\203ù\aw\002ÿ\002\203Ã\b\211\035 Ò#\bèµKÂÿ\205À\017\205Ñ") at interpret.c:2050 #15 0x0807062c in jump_opcode_F_CATCH () at interpret_functions.h:1236 #16 0x0844b563 in ?? () #17 0x00000003 in ?? () #18 0x0891a708 in ?? () #19 0x00000003 in ?? () #20 0x40c32864 in ?? () #21 0x00000008 in ?? () #22 0x00000000 in ?? () #23 0x099c6cb8 in ?? () #24 0x40c327b0 in ?? () #25 0x40cf7000 in ?? () #26 0x40c327f0 in ?? () #27 0x080761d4 in o_catch ( pc=0x880dc8f "¡(Ò#\b\213\r Ò#\b\211\b\203À\004£(Ò#\b\213\0350Ò#\b\017¿S>\203Â\033\211T$\b\213S8\211T$\004\211\f$èÌQ\215ÿ¡ Ò#\b\203À\b£ Ò#\bÇ\004$") at interpret.c:2050 Previous frame identical to this frame (corrupt stack?) I don't know what produced it and how to reproduce it:( Ideas? / David

2 1

atexit
by Martin Nilsson (DivX Networks) ＠ Pike (-) developers forum 15 Oct '04

15 Oct '04

It looks like atexit is broken. Some recent change prevents the Redline cleanup code to always run when you exit Hilfe, messing up the terminal. Who did something here last? mast? grubba?

4 3

ncurses
by Kaos 15 Oct '04

15 Oct '04

Hi, We're about to use ncurses from pike, and thus need a module to interface with the ncurses lib. My question then is: does anyone have (and want to share) a ncurses module? Otherwise, we'll develop one ourselves (will put it on modules.gotpike.org as soon as we have something usefull). Or is there a better way(tm) ? // Andreas

6 14

longjmps in pike
by Marcus Agehall (PacketFront) ＠ Pike (-) developers forum 15 Oct '04

15 Oct '04

While looking at the code in dbg_main(), I saw that pike uses longjmps. I suspect the use for this is the exception handling stuff. Am I on the right track? Can someone elaborate a bit on the usage as to where it is used in pike and how things are supposed to interact..?

4 5

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 October 2004