Pike-devel

pike-devel@lists.lysator.liu.se

3041 discussions

Pike & dumping
by David Gourdelier 06 Apr '04

06 Apr '04

Hello, I'm currently trying to make Caudium startup faster. For now it takes 21 seconds for it to load all the modules on my desktop and I found a little improductive to wait so long when I develop on it and restart it often. I tried to dump the Caudium's pmods as pmod.o files with the Tools.Standalone.dumpit function. The dump works good but the startup is not faster so I wonder, how does Pike load .pmod.o ? Does it really make things faster? Do I need a special thing to load .pmod.o if the pmods are not within the Pike official path for modules ? (thought there are in the path given with -M). Thank you for your answer. / David Gourdelier

3 3

7.4.44, Mird, Solaris 9, /opt/SUNWspro/bin/cc och "restrict"
by Christer Bor�ng, CD (Procurator Odiosus Ex Infernis) ＠ Pike (-) developers forum 06 Apr '04

06 Apr '04

"restrict" verkar vara ett reserverat ord, vilket gör att Mird-modulen inte kompilerar. Filen src/modules/Mird/libmird/hashtrie.c försöker använda en variabel som heter restrict (runt rad 443-525). När jag byter namn på den till my_restrict så fungerar det hela.

2 1

__dtoxmode()
by Marcus Agehall (Scanian) ＠ Pike (-) developers forum 06 Apr '04

06 Apr '04

At line 318 in fdlib.c Pike 7.4.44 (source tarball) calls a function called __dtoxmode(). Where is this function defined? I'm attempting to compile Pike using MinGW32 but it fails in several places. If anyone else wants to give it a shot, just download Mingw, MSys and the win32api update from www.mingw.org and start compiling.. I have succeded in patching the code so that Pike compiles and loads the master, but there seems to be several problems within fdlib since it seems like file operations fail. fdlib was a problematic section and I made the first attempt by just commenting out problematic lines from the code. I will now redo the attempt and try to fix problems in a more correct way.

1 0

The very first argument
by Martin Nilsson (provokat�r) ＠ Pike (-) developers forum 04 Apr '04

04 Apr '04

I just had a backtrace that started with modules/security/htaccess.pike:690 Error: Bad argument -2147483648 to HtUser. Looks like some sort of int interpretation confusion (in language?).

3 3

leak
by Martin Nilsson (provokat�r) ＠ Pike (-) developers forum 02 Apr '04

02 Apr '04

Has anyone taken a look at the call_out leak in the backend?

3 3

fatal poll error
by Bill Welliver 02 Apr '04

02 Apr '04

I'm getting fatal errors while trying to connect to a tcp port using 7.5.23: /home/hermes/bwellive/projects/Pike/7.5/src/backend.cmod:403: Fatal error: Failed to register fd 4 for 0x0041 with /dev/poll (errno:13). Backtrace at time of fatal: object(src/module.c:62).___files()->Fd: object(src/modules/files/file.c:3909)->set_read_callback(object(/usr/local/pike/7.5.23/lib/module s/Stdio.pmod/module.pmod:102)->__stdio_read_callback) object(/usr/local/pike/7.5.23/lib/modules/Stdio.pmod/module.pmod).File: object(/usr/local/pike/7.5.23/lib/modules/Stdio.pmod/module.pmod:102)->set_read_callback() Any idea why these would be _fatal_ errors (ie pike bails out)? Bill

2 7

ADD_FUNCTION() macro on pike 7.5 ?
by Xavier Beaudouin 31 Mar '04

31 Mar '04

Hello there, Does ADD_FUNCTION macro still works outside the program scope on the module level in a Pike C module (not in a cmod) ? We have currently issues in our Caudium C module... and it seems that the following code has stopped to work : main.c /* ... */ void pike_module_init( void ) { /* some other add_function_constant() */ add_function_constant( "http_date", f_http_date, "function(int|void:string)", 0); /* */ init_external_module(); /* */ } in external.c we have : void init_external_module (void) { ADD_FUNCTION("whatever",f_whatever, tFunc(tVoid, tInt), 0); } Before (on 7.2 and in 7.4) it has used to work, but today on Pike 7.5 it has suddenlt stopped to work... Is this normal ? /Xavier -- Xavier Beaudouin - Unix System Administrator & Projects Leader. President of Kazar Organization : http://www.kazar.net/ Please visit http://caudium.net/, home of Caudium & Camas projects

12 36

Calendar question
by David Gourdelier 30 Mar '04

30 Mar '04

Hi again, It seems to me that the autodoc of the Calendar module is incorrect or there is a problem in the Calendar.Time* classes. The doc say you can use Calendar.TimeRange("unix", int unixtime) but when I try with either pike 7.5 or 7.2, I have: > Calendar.TimeRange("unix", time()); Calling undefined function. ISO.TimeRange: TimeRange()->create_unixtime_default(1080599245) /usr/local/pike/7.2.504/lib/modules/Calendar.pmod/TimeRanges.pmod:81: TimeRange()->create("unix",1080599245) ISO: ISO->TimeRange() -:1: ___Foo4711() > Calendar.TimeRange("unix", time()); Calling undefined function. ISO.TimeRange: TimeRange()->create_unixtime_default(1080599068) /usr/local/pike/7.5.23/lib/modules/Calendar.pmod/TimeRanges.pmod:78: TimeRange()->create("unix",1080599068) /usr/local/pike/7.5.23/lib/modules/Calendar.pmod/TimeRanges.pmod:11: ISO->TimeRange() HilfeInput:1: HilfeInput()->___HilfeWrapper() There is the same problem with Calendar.Time.TimeofDay. When looking at the code it seems indeed that the create_unixtime_default function is not defined in either TimeRange nor TimeofDay classes but defined in clases that inherited these one. So I wonder if the doc lie or if the code is wrong ? I also changed the doc for http://pike.ida.liu.se/generated/manual/modref/ex/predef_3A_3A/Calendar/Tim…, replacing "unixtime" by "unix". / David

3 3

Changelog and newsletter
by Bill Welliver 30 Mar '04

30 Mar '04

Okay folks, here's the deal: I'll volunteer to periodically (4-6 times a year should developments dictate) put together a Pike Update. As I see it, things that would be worthwhile in such a production might include: Pike development updates (what's happening, what the plan is) Pike Tips Usage Articles Conference Information Pike-based application updates and information etc. In order to do this, I need some help. I know that there is at least a little resistance to the idea, but having a changelog updated as changes get made would really help me (and other potentially interested parties) figure out what is going on. The reasons I think that a changelog is necessary and beneficial include: 1. A new "feature" or change may involve 5 - 10 or more cvs checkins, the log entries of which may not hint at the actual change being made. 2. Sometimes, even those of us who might follow the changes religously might have trouble figuring out what's going on, or the significance of that change 3. Waiting until a release is made to find out what's happening, while dramatic, is really a bit of a problem from a planning perspective: people _using_ the software should really know what's changing as it happens. 4. Having an awareness of a change lets people plan for it when it's released 5. Allows interested parties to test and (heaven forbid) document the change, possibly helping you, the developer out. I know that there are some people who feel that this is just extra work. To those, I say that someone is making the entry at the end, why not take a few seconds and describe the change (and take credit for it) when it's added. Not every checkin will prompt an entry, only those that complete a change or addition. If there's any doubt about whether a change merits inclusion, err on the side of over documenting and include it. An example of a few changelog entries I have made that represent a few hours of work on my part: - SSL3 client can now present certificates to a server; server can request and validate a client certificate chain presented. (Bill Welliver) - SSL3 module now checks presented certificate chains for validity by examining effectivity, issuers, and completeness of certification chain. (Bill Welliver) - Added method Tools.X509.verify_certificate_chain() that can verify the validity of a chain of X509 certificates. (Bill Welliver) This took me a total of 2 minutes to do, and might have taken less if the work were fresh in my mind. The third entry was one I included that would probably be cut from a release change log, but is certainly not wasted space. I'd like to see a policy added for 7.7 that says if you check in a completed new or changed feature (module, major work, language behavior, etc), you add a brief entry describing what you've done. Any thoughts? Who can make this happen? If anyone has any comments or would like examples of what I envision, please feel free to contact me. Thanks, Bill

1 0

Looking for translators
by David Gourdelier 29 Mar '04

29 Mar '04

Hi, The Camas team is looking for translators willing to translate the next Camas release in the language you know. For now the languages for which we have an up to date translations are: . English . French . Spanish . Polish And thus the Camas team is looking for anyone who knows a language except these. Mainly we are looking for: . German . Italian * . Swedish * . Finish * . Czech * . Dutch * . Estonian * . Hungarian * . Portuguese * . Russian The languages with a * are languages for which the translation is not up to date for this release. The translation is very easy and it'll require you only a text editor and a few minutes. If you are willing to translate, just send an email to me and I'll send you a file that you'll just need to translate and send me back. Thank you for your help. Best regards. -- David Gourdelier

2 1

Jump to page:

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