[Vm-dev] Selecting distributed plugins

Robert robert.withers at pm.me
Sun Mar 1 19:48:17 UTC 2020


I do not believe it is a good idea to have multiple implementations of 
these primitives running about. Why didn't Croquet leverage the work 
done in Cryptography? A question for the ages. Can we consolidate them 
now, under the Cryptography plugins, please? Those are tested against 
the Cryptography code. As far as 64-bit, I do not know, I imagine they 
have not been ported. I am CCing the Cryptography team in case someone 
would be interested in doing that work for the team. I still offer the 
DES and TripleDES options for encryption in ParrotTalk. I am also using 
SHA256. What would be really sweet is if someone could implement SHA512...

There are also the DateAndTuime errors in 5.3RC3, and a failure of

MultiByteFileStream>>#primSize:

in

Squeak5.3rc2-19422-64bit which I wrote about to the dev list.


Kindly,
Rabbit

On 3/1/20 2:39 PM, Levente Uzonyi wrote:
>
> On Fri, 28 Feb 2020, Nicolas Cellier wrote:
>
>> Robert kindly asked me whether it would be possible to include DES and
> MD5 plugins in OpenSmalltalk VM distribution.
>> We currently generate src/plugins code for those 2, but don't compile
> them.
>
> There's also the SHA256Plugin which didn't see much use for sure.
> MD5 is covered by the CroquetPlugin, and DES is pretty much outdated.
> I suspect that neither of these plugins are ready for 64-bit compilation,
> and they probably yield plenty of warnings because of typing issues.
>
>> This raises the question: who decides which plugin gets
> included/compiled/distributed or not? On which criterion?
>
> IIRC these were included with the interpreter VM, but only after the Terf
> VM branch was created. Since the latter became the current main branch, we
> do not have those plugins in today's builds.
>
> Levente
>
> P.S.: I've got an unreleased package which implements MD5/SHA1/SHA2 hashes
> in Smalltalk, optionally using DSAPrims and CroquetPlugin for the first
> two. If there's interest, I can upload it somewhere.



More information about the Vm-dev mailing list