Pike-devel

pike-devel@lists.lysator.liu.se

3 participants
3048 discussions

Start a nNew thread

sort
by Johan Sundstr�m (folksk�dare) ＠ Pike (-) developers forum 28 Apr '03

28 Apr '03

Pike's sort is not stable.

11 55

Array.flatten
by Martin Nilsson (har bott i google) ＠ Pike (-) developers forum 28 Apr '03

28 Apr '03

In order to not escalate the CVS war, I'll present my solution for how to deal with cyclic arrays in Array.flatten here: //! Flatten a multi-dimensional array to a one-dimensional array. //! @trows //! @[flatten] throws an error if @[a] is cyclic. array flatten(array a) { array ret=({}); multiset seen = (<>); void low_flatten(array aa) { foreach(aa, mixed b) if(arrayp(b)) { if(!sizeof([array]b)) continue; if(seen[b]) error("Trying to flatten a cyclic array.\n"); seen[b]=1; low_flatten([array]b); } else ret += ({ b }); }; low_flatten(a); return ret; } (aside from the fact the programs like the one below hangs pike) void foo() { void bar(array i) { error("gazonk\n"); }; array a = ({ 0 }); a[0]=a; bar(a); }

5 5

How does this work?
by Marcus Agehall (Tr�dl�s) ＠ Pike (-) developers forum 27 Apr '03

27 Apr '03

Yesterday I checked in a patch for the MySql module which corrected the prototype for Mysql.mysql.create(). The reason I found that broken prototype, was that when fiddeling with the mysqls dbtype (see checkins from yesterday) the Sql module failed to compile since Sql.mysqls (which inherits Sql.mysql) overloads create(). The compilation failure was due to too many arguments passed to Mysql.mysql.create(). How can it be that that prototype has been broken for quite some time and Sql.mysql hasn't failed to compile? Can someone enlighten me a bit on how this works? The feeling I got from the problem, is that it might be some weird bug, but it might also be that I fail to see the difference in the inheritance.

1 0

Canonical order function
by Martin Stjernholm, Roxen IS ＠ Pike developers forum 26 Apr '03

26 Apr '03

The new ADT interface and the new multisets use a function that works like `< to define order while Array.sort_array uses `>. This inconsistency is silly and should be resolved one way or the other. So I ask for opinions which operator you'd like to see as the canonical ordering function in Pike. Reasons to choose `>: o No compatibility issue with Array.sort_array (it's not a compatibility problem to change it for multisets and in Petas ADT interface since those haven't been released). o It maps somewhat better to a three-way compare: If a < b: cmp(a, b) == -1 (a > b) == 0 If a == b: cmp(a, b) == 0 (a > b) == 0 If a > b: cmp(a, b) == 1 (a > b) == 1 <= Same sign. C.f: If a < b: cmp(a, b) == -1 (a < b) == 1 <= Opposite sign. If a == b: cmp(a, b) == 0 (a < b) == 0 If a > b: cmp(a, b) == 1 (a < b) == 0 Reasons to choose `<: o The operator < is almost always used in mathematics to describe order relations, not >. Thus it feels more natural, at least for me.

1 0

Valgrind
by Marcus Agehall (Tr�dl�s) ＠ Pike (-) developers forum 26 Apr '03

26 Apr '03

I seem to have trouble running pike under valgrind on a RedHat 7.3 system. I get the following when using --valgrind: /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:332: Bad type in assignment. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:332: Expected: mapping(string:multiset(string)) /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:332: Got : int(-1..-1) /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:1026: Class definition failed. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:1130: Must return a value for a non-void function. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:1187: Class definition failed. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:1189: Class definition failed. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:1285: Must return a value for a non-void function. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:1323: Class definition failed. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:1688: Class definition failed. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:2560: Class definition failed. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:3083: Class definition failed. /home/agehall/Pike/7.5/build/linux-2.4.18-27.7.xsmp-i686/master.pike:3139: Class definition failed. ERROR: Compilation failed. Running pike outside of valgrind works just fine. It's a newly compiled but not installed Pike 7.5 (fresh from CVS) that I'm trying to run. Any ideas on how to get it to work?

4 5

Re: sort
by Fredrik (Naranek) Hubinette (Real Build Master) ＠ Pike (-) developers forum 25 Apr '03

25 Apr '03

> o destructive sorting when array has one ref is no longer > possible (fairly rare anyways) It seem I remembered wrong, it's sorted destructively when there is only one argument to sort(), which is of course a bit more common.

1 0

Sad.....
by Xavier Beaudouin 25 Apr '03

25 Apr '03

Hello there, I wanted to drop a note that recent thread about IPv6 support and Pike, has a bad result (for me, for all the rest of Pike devel team I dunno). Marek has decided he will stop to commit, submit patches and even participate in pike development, he removed all his mails from all Pike related mailing lists (pike / pike-devel / roxen) and will only maintaint debian pike package. Everybody here may think about that. Not the fact that some good pike developper goes away, but the fact that the Pike Community (sic) is maybe (personnaly I am allmost sure) too mutch elitist maybe no soo opened as it claims to. What is the main problem with Pike : too small community, too less documentation, too mutch skill is needed to join into it. Why do I say that : I discovered Spinner in early 96.... I took 6 years to under some Pike coding and how to make some C modules. On others languages (says Perl, Python, ...) it is far more easy to get a good skill in months instead of years...... Why we can't promote Pike ? - lack of docs - too mutch obscure things - sometimes compatibilty is not really compatibility - it is NOT embeddable.... (so we can't make a mod_pike for apache). I am allmost fed up myself with some behavior of Pike people... that I have decided to make my own things in my projects instead of giving my code to Pike community (eg some of this code are now in _Caudium module inside Caudium 1.3).... You can say it is sad, but when I send messages to pike devel ml some of them are thrashed, others are ignored or flammed. Now Pike Community loose one of their member.... Maybe others members can come RSN.... /Xavier

12 26

ADTs third time around
by Peta, jo det �r jag ＠ Pike developers forum 25 Apr '03

25 Apr '03

I'we compiled a new version of how eventual new ADTs should work. I was planning to put out a document with some new ADTs to but that has been delayed quite a few times, so I decide to put up only the old 'Common API for new ADTs in Pike' again, but with a few changes. Many the ADT methods are now optional, instead the number of iterators has grown. The document size has also grown so you can find it as a PDF at http://w1.313.telia.com/~u31318241/CommonAPI.pdf

2 2

Re: Sad.....
by Xavier Beaudouin 24 Apr '03

24 Apr '03

Le jeudi, 24 avr 2003, à 22:45 Europe/Paris, Martin Nilsson (har bott i google) @ Pike (-) developers forum a écrit : >>> - sometimes compatibilty is not really compatibility >> >> IMHO, it is not a good idea to support outdated and rare platforms >> which >> are in use by two or three individuals on the planet... > > Being able to compile on outdated and rare platforms is in itself of > little value. It does however pay off every time there is a brand new > platform we want to support. E.g. we have fairly good support for 64 > bit > architechtures thanks to our support for rare platforms. Though I am > not > sure Xavier meant portability. s/portability/compatibility/ Was speaking about pike itself. Eg #pike7.2 somethimes is not enougth to help comptability inside some code.... :/ >>> - it is NOT embeddable.... (so we can't make a mod_pike for >>> apache). >> And it will never be, I believe. At least it will not be done by >> the Pike >> team. Why? "Nobody needs it" > > Well, its on top of my wishlist. Good ::)

1 0

IPv6 diff for Pike 7.5.7 (today's snapshot)
by Marek Habersack 24 Apr '03

24 Apr '03

The attached diff is not finished. I've just rediffed and adjusted whatever I could, but I have no more time today for keep working on it. The get_inet_addr function in system.c needs more work to get the families right (basically a few switch statements in there). The patched pike compiles on Linux/Debian, it will compile on any linux with glibc 2.2+, should compile fine on FreeBSD/NetBSD/Solaris. I'm attaching the latest Pike 7.3 diff for the stuff which I used as the basis for the 7.5 one. You can use it for reference wherever I goofed up (which I undoubtedly did). If you do any tweaks on the patch, please send me a diff against the latest 7.5 CVS, thanks, marek

4 9

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