Re: [ANN] stdlib 36 released

["Gary V. Vaughan" <gary@...>]

Just a quick note that I've fixed all the issues below in the just released stdlib v37:

On Jan 16, 2014, at 11:04 PM, Gary V. Vaughan <gary@vaughan.pe> wrote:
>> On Jan 16, 2014, at 10:06 PM, Dirk Laurie <dirk.laurie@gmail.com> wrote:
>> 
>> 2014/1/16 Gary Vaughan <gary@vaughan.pe>:
>> 
>>> I am happy to announce release 36 of stdlib.
>> 
>>> ** New features:
>>> 
>>> - Modules have been refactored so that they can be safely
>>>  required individually, and without loading themselves or any
>>>  dependencies on other std modules into the global namespace.
>> 
>> I have tried this under Lua 5.2.3.
>> 
>> - `std=require"std"` does not return a table containing std.set etc
>>  which is assigned to a variable `std` but gives the error message
> 
> That is a TODO item.

Now a done item.  Thanks for the prod.

>> /usr/local/share/lua/5.2/std/debug.lua:6: module 'std.debug_init' not found:

Added a workaround for the LuaRocks install bug v36 tripped in the new v37
release.

>> - `set=require"set"` adds the names `proper_subset` and `new` to the
>> global namespace. The former is fairly harmless, the latter is disastrous.
>> 
>> I have not tried any other submodules.
> 
> Yikes! Thanks for the heads up. Until very recently stdlib basically scribbled all over the global namespace, and moving to encapsulate everything in tables is a work in progress - not yet at the point where I've actually audited for outliers. But std.set is supposed to be one of the clean modules, so I'll fix  that (and any others I find) and put out a new release shortly.

It turns out that after adding some extensive specl examples to every module to specify no symbol leaking, those two (new and proper_subset) were the only two culprits in all of stdlib! Also fixed in the new release.  Thanks again :)

Cheers,
-- 
Gary V. Vaughan (gary AT vaughan DOT pe)

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail