[OPEN-ILS-DEV] ***SPAM*** Re: ***SPAM*** RE: ***SPAM*** Re: ***SPAM*** Re: SHA1 code in sha.c

[Jason Stephenson]

Quoting Steve Wills <steve.wills at lyrasis.org>:

> It's also bundled in my Apple osX where my iPhone development  
> environment lives, not to stray TOO far afield. *g*

You might be interested in what I've done before: A message digest  
category on NSData.

:)

>
> Stev3
> Lyrasis
>
> -----Original Message-----
> From: open-ils-dev-bounces at list.georgialibraries.org  
> [mailto:open-ils-dev-bounces at list.georgialibraries.org] On Behalf Of  
> Thomas Berezansky
> Sent: Tuesday, December 22, 2009 9:25 AM
> To: open-ils-dev at list.georgialibraries.org
> Subject: Re: [OPEN-ILS-DEV] ***SPAM*** Re: ***SPAM*** Re: SHA1 code in sha.c
>
> Additional note:
>
> Several things, such as postgresql (client and server) and apache with
> SSL, already require that the openssl implementation be installed on
> the system. The SSH implementation installed on the systems will
> likely also require it. Adding another makes no sense, especially when
> we already have openssl's sitting there from everything else. Given
> that we can SSH across endianess and int sizes I assume that all of
> that is covered properly in the openssl implementation.
>
> Thomas Berezansky
> Merrimack Valley Library Consortium
>
>
> Quoting Jason Stephenson <jstephenson at mvlc.org>:
>
>> Quoting Scott McKellar <mck9 at swbell.net>:
>>
>>>
>>>
>>> --- On Mon, 12/21/09, Jason Stephenson <jstephenson at mvlc.org> wrote:
>>>> Why reinvent the wheel? Why not just
>>>> farm the work out to libcrypto? Last I checked OpenSSL
>>>> typically has a SHA1 implementation that you don't have to
>>>> fiddle with.
>>>>
>>>> Jason
>>>
>>> I have no intention of reinventing any wheels.  I did a quick Google and
>>> saw the OpenSSL project among others.  What I don't know is how portable
>>> they are.  The ones I looked at so far don't say very much on their
>>> websites about architectures, int sizes, endianness, or other aspects of
>>> portability.  I'd rather not have to decipher a dozen different
>>> implementations of an algorithm I don't understand in order to figure
>>> out which ones are better than what we've got.
>>
>> OpenSSL is about as portable as you can get, and is pretty much
>> guaranteed to be installed on any machine that can host OpenILS.
>>
>>
>>>
>>> If anybody already knows of a suitable implementation, or knows of one
>>> that we should avoid, you may be able to save me some time.
>>>
>>> Scott McKellar
>>>
>>>
>>
>>
>>
>
>
>