[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: [lua-l] Re: LuaSocket 3.0-rc1... Please release a 3.0 rock, at a convenient time?
- From: sur-behoffski <sur_behoffski@...>
- Date: Tue, 29 Oct 2019 11:31:52 +1030
On 2019-10-18 01:13, Bruno Silvestre wrote:
This problem came from LuaSocket. Luarocks is installing it as dependency.
from LuaSocket 3.0-RC1, file src/option.c:40,41
char msg[45];
sprintf(msg, "unsupported option `%.35s'", name);
It was fixed in 'master' branch:
char msg[57];
sprintf(msg, "unsupported option `%.35s'", name);
regards
--
bruno
FYI, I run Gentoo GNU/Linux, and have "luasocket" installed both as:
1. An OS-stable package:
=dev-lua/luasocket-3.0_rc1-r4
and
2. A LuaRocks Rock:
luasocket
3.0rc1-2 (installed) - /usr/local/lib/luarocks/rocks-5.1
My impression is that Gentoo marking "3.0-rc1-r4" as Stable is a strong vote for
releasing the set of sources, plus the patch sets that comprise the -r4 set, as
a full LuaSocket 3.0 Rock.
Presumably the simple array size expansion, to eliminate the chances of a
buffer overflow on sprintf, would be included in Gentoo's -r4 patches, and thus
would address the compiler warnings.
Gentoo's "eix luasocket" describes dev-lua/luasocket's Homepage as:
http://www.tecgraf.puc-rio.br/~diego/professional/luasocket/
Sorry to be pushy, but I hope that this is a fairly simple and easy change. If
the developer's workload is high, and it takes many months before this change is
attempted, I would still be grateful: I very much enjoy the Lua language and
extended ecosystem; my reasons for pushing back on warning messages is because
these can be off-putting for newcomers, leading to an impression of a
second-class ecosystem. In particular, I believe that LuaRocks 3.2.1 has now
settled into a space where it works better than 2.x for being a central package
manager, and so am now trying to make it more of a central unifying component
(e.g. std.normalize, std.strict, bit32, ldoc, luaposix etc) for many of my
scripts.
cheers,
s-b etc.