Pike-devel

pike-devel@lists.lysator.liu.se

3041 discussions

Re: _main call in OCPikeInterpreter.m
by Martin Stjernholm, Roxen IS ＠ Pike developers forum 06 Jun '08

06 Jun '08

It also does a SETJMP and several lines of error handling, only to be followed by a block that effectively executes no code that could have use for it. So it's not only priming for something that happens somewhere else. To me it looks just as if the very call to _main itself is missing there. And no, _main no longer uses that, that's why the function is gone. If someone wants the environment in that (bulky) form, (s)he now has to do it by converting the mapping returned by __builtin._getenv.

1 0

_main call OCPikeInterpreter.m
by Martin Stjernholm, Roxen IS ＠ Pike developers forum 05 Jun '08

05 Jun '08

I've just committed a fix so that putenv and getenv actually do what they claim to do. As part of this, master()->_main no longer takes an environment mapping. I think I've made the corresponding change for that in OCPikeInterpreter.m too, but someone who knows what's going on there better take a look at it, because to me the call to master()->_main seems to be missing altogether.

2 2

7.8 blockers
by Martin Nilsson (Opera Mini - AFK!) ＠ Pike (-) developers forum 01 Jun '08

01 Jun '08

For Tools.Standalone.dump constant dont_dump_program=1 is a solution.

1 0

Setting things to NULL after free
by Martin Stjernholm, Roxen IS ＠ Pike developers forum 01 Jun '08

01 Jun '08

It's been customary to set pointers to NULL after freeing them in object (and module) exit functions. I'd like to question whether that really makes much sense. The object storage is only going to be freed just afterwards anyway (same goes for dynamically loaded modules), so the only bugs that this kind of "hygiene" could help nail is if destruct_object or exit_dynamic_load misbehaves. That's a bit unlikely by now, and anyway it's an awful lot of code everywhere just for that. The assignments are not even inside #ifdef PIKE_DEBUG as such belts-and-suspenders things usually are. Or am I missing something here? Also, when a pointer is not supposed to be used anymore, I usually do something like this instead: free (x->foo); #ifdef PIKE_DEBUG x->foo = (void *) (ptrdiff_t) -1; #endif That is more likely to trig a helpful coredump if the pointer is being used incorrectly.

1 0

pike version numbering
by Martin Bähr 31 May '08

31 May '08

hi, since i have seen a few references to pike version 7.10 i'd like to discuss the current versioning scheme. i'd like to suggest that the next version is numbered as pike 8. the main reason for that is that if we continue with 7.10, 7.12, etc, we will NEVER reach a state where the changes are large enough to get to pike 8 because after 7.10 and further switching to 8 will create expectations where pike 8 would be significantly different from pike 7. in other words: 7.8, 7.10 then will be our major versions, and the 7 in front would become rather meaningless. a succession like 7.6, 7.8, 8.0 seems natural and won't surprise anyone. ie: no one will expect a huge set of changes there. on the other hand the changes from 7.6 to 7.8 seem quite significant (syntax additions like getters and setters, etc) already, that i even think it may be waranted to go directly from 7.6 to 8. in any case however, we should never go to 7.10. greetings, martin.

3 2

Blocks inside expressions
by Martin Nilsson (Opera Mini - AFK!) ＠ Pike (-) developers forum 31 May '08

31 May '08

That points towards the lack of need for inlined expression blocks, but it may just be that people have been conscious about the lambda overhead or thought a function would be more apropriate.

1 0

Blocks inside expressions
by Martin Stjernholm, Roxen IS ＠ Pike developers forum 31 May '08

31 May '08

Aha, cute. Yes, that use would be exclusive with this one, unfortunately.

1 0

Blocks inside expressions
by Martin Stjernholm, Roxen IS ＠ Pike developers forum 31 May '08

31 May '08

I don't know. What would these generators be?

1 0

Blocks inside expressions
by Mirar ＠ Pike developers forum 31 May '08

31 May '08

Sounds good to me. Is that keyword used?

2 1

Blocks inside expressions
by Martin Stjernholm, Roxen IS ＠ Pike developers forum 31 May '08

31 May '08

Not only, as said. It seems like a waste of time to fix inlining of that particular kind of lambdas (and I think it wouldn't be nearly as simple as this) when the end result is only less versatile and clumsier to write. Other kinds of lambdas, like e.g. those commonly used with map() and filter() would still not be possible to inline, btw. For that to work we'd have to change map(), filter(), etc to special builtin constructs.

1 0

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