lua-l archive

Search lua-l

This index contains 143,615 documents and 1,774,615 keywords. Last update on 2023-03-09 .

341. Re: a new proposal for environments (score: 3)

Author: Roberto Ierusalimschy <roberto@...>

Date: Thu, 25 Feb 2010 17:52:52 -0300

Again maybe I was not very clear. The specificity of the situation was not only the sandboxing, but the absence of the io lib. With the io lib, you should be able to do a loadfilein like here: loadi

342. Re: a new proposal for environments (score: 2)

Author: Duncan Cross <duncan.cross@...>

Date: Thu, 25 Feb 2010 20:19:12 +0000

I apologise for selectively quoting like that, it was bad form. (I am not entirely convinced that the only reason you would ever want to use a "loadfilein" is sandboxing, but then in truth I don't ha

343. Re: a new proposal for environments (score: 2)

Author: Alexander Gladysh <agladysh@...>

Date: Thu, 25 Feb 2010 12:06:54 +0300

Perhaps it would be better to separate the two needs? Say, debug.* for debugging, meta.* for metaprogramming. Not sure if it is possible though... <...> Cool idea. But I have a question. So, _ENV is

344. Re: a new proposal for environments (score: 2)

Author: Stephen Kellett <lua@...>

Date: Wed, 24 Feb 2010 23:50:28 +0000

Roberto Ierusalimschy wrote: At the same time, we are convinced that there are situations where those functions may be quite useful. To reconcile both views, we are reviewing our position that the de

345. a new proposal for environments (score: 2)

Author: Roberto Ierusalimschy <roberto@...>

Date: Wed, 24 Feb 2010 20:38:47 -0300

After that long discussion about setfenv/getfenv, we are convinced that both functions should belong to the debug API. Both clearly break a function's abstraction, tampering with the "original meanin

346. Re: setfenv/getfenv (score: 2)

Author: Nick Gammon <nick@...>

Date: Thu, 11 Feb 2010 11:17:23 +1100

This proposed solution assumes you want io.open to be available. If I want to replicate: func = loadfile(path) setfenv(func, env) func() ... but have sandboxed out io.open, is there another solution?

347. Re: The problem with stack-based environments... (score: 3)

Author: joao lobato <btnfdp.lobato@...>

Date: Sun, 31 Jan 2010 19:35:20 +0000

Lua provides mechanisms, you write the modules. Couldn't you wrap every method of your plugin such that plugin:method() would actually do local results local table = table in sandbox do results = tab

348. The problem with stack-based environments... (score: 3)

Author: Mark Hamburg <mark@...>

Date: Sun, 31 Jan 2010 11:09:57 -0800

As the primary proponent of stack-based-environments -- though I will identify myself less as a proponent and more as someone who has tried get clarity around what in-do-end seems to be implying (as

349. What are the use cases of in-do-end? (score: 3)

Author: GrayFace <sergroj@...>

Date: Wed, 27 Jan 2010 19:24:42 +0600

So far I've read mentions of several use cases, but I don't understand how exactly it may be useful in them. Here's what I read about: 1) Obvious usecase: something like Deplhi's 'with'. This is not

350. Re: Lua 5.2 and in-do-end (score: 2)

Author: Mark Hamburg <mark@...>

Date: Mon, 25 Jan 2010 15:38:55 -0800

That's a nice way to do the analysis. "Stack-based environments" however is probably still clearer than "dynamic binding" though it refers to mechanism rather than effect. Mark

351. Re: Lua 5.2 and in-do-end (score: 2)

Author: Edgar Toernig <froese@...>

Date: Mon, 25 Jan 2010 23:36:43 +0100

I agree, especially as the scope is not dynamic at all. By lexical rules it's either local or global. The "binding" of globals is dynamic. Locals: early binding. Globals in function-based environment

352. RE: Lua 5.2 and in-do-end (score: 2)

Author: mark@...

Date: Mon, 25 Jan 2010 13:22:22 -0700

in-do-end only seems to be interesting for sandboxing with stack-based rather than function-based environments. (I'm thinking that terminology may be more descriptive than "dynamic scoping".) Mark

353. Lua 5.2 and in-do-end (score: 2)

Author: joao lobato <btnfdp.lobato@...>

Date: Sat, 23 Jan 2010 21:57:53 +0000

Hello list The in-do-end construct in Lua 5.2 compelled me to subscribe to the list (rather than simply browsing through the recent archives). Maybe this post should have gone to a pre-existing threa

354. Re: Multiple lua instances (score: 3)

Author: ingmar wirths <ingmania@...>

Date: Tue, 19 Jan 2010 17:49:47 +0100

2010/1/19 Rob Kendrick <rjek@rjek.com>: Allright, but since these restrictions are all implemented in lua, how can i enforce them, given that the user should be permitted to modify his ai? From the p

355. Re: Multiple lua instances (score: 3)

Author: Rob Kendrick <rjek@...>

Date: Tue, 19 Jan 2010 16:42:04 +0000

On Tue, 19 Jan 2010 17:34:25 +0100 ingmar wirths <ingmania@googlemail.com> top-quoted: The idea is that if a Lua function can't see another Lua function, it can't call it, and it can't magic up a ref

356. Re: Multiple lua instances (score: 3)

Author: ingmar wirths <ingmania@...>

Date: Tue, 19 Jan 2010 17:34:25 +0100

Thanks for your help so far, i guess i have to study luas sandboxing facilities. What still gives me frown, is that these sandboxing facilities are all implemented in lua. A user, who wants to bypass

357. Re: Multiple lua instances (score: 2)

Author: David Given <dg@...>

Date: Tue, 19 Jan 2010 16:03:45 +0000

Lua instances are reasonably lightweight. It's feasible to run thousands of concurrent Lua states; it's a rather extreme thing to want to do, and you'll need hefty memory requirements, but it *can* b

358. environment as upvalue (score: 2)

Author: "Juris Kalnins" <juris@...>

Date: Tue, 19 Jan 2010 09:08:00 +0200

Tried to implement this, the patch came out quite small, and not even complete, e.g. getfenv is simply missing. I'm sure I got it wrong at places, but I like the idea of not having separate concept o

359. Re: setfenv/getfenv (score: 3)

Author: Mark Hamburg <mark@...>

Date: Wed, 13 Jan 2010 22:47:38 -0800

While there are certainly cases where dynamic scoping has seemed appealing, I think it's in-do-end that seems to bring it to the forefront since it provides linguistic (rather than just library) supp

360. Re: setfenv/getfenv (score: 2)

Author: Fabio Mascarenhas <mascarenhas@...>

Date: Wed, 13 Jan 2010 22:11:25 -0200

I think function environments are a better fit for Lua's role as an extension language, though. Having functions called by the sandboxed functions get access to the non-sandboxed environment is ok, a