<div dir="ltr"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
I think that's a good project, though a test vector language does not<br>
necessarily need to be coupled to a HacSpec language, I think.
</blockquote><div><br></div><div>I agree. But as we need test vectors anyway I think it makes sense have a common way specify them. It just makes things easier.<br></div><div>We'll see if If the "test vector language" becomes anything more.<br><span class="gmail-im"></span><br><span class="gmail-im"></span><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span class="gmail-im">
</span>FWIW, I like the idea of using a rigid subset of JSON for test<br>
vectors, so it's easy to parse *either* with JSON or with dumb,<br>
hand-written parsers.
</blockquote><div><br></div><div>The idea is really to keep this as simple as possible without making it too hard to interpret. <br></div><br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
NIST ACVP and Wycheproof are also looking into JSON test vectors for<br>
crypto, so if you want to go further in that direction you could try<br>
to get some of those people here.
</blockquote><div><br></div><div>Thanks for the pointers. I'll try to get in touch and see if they're interested. <br></div></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Feb 14, 2018 at 9:05 AM, Trevor Perrin <span dir="ltr"><<a href="mailto:trevp@trevp.net" target="_blank">trevp@trevp.net</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Sun, Feb 11, 2018 at 6:40 PM, Franziskus Kiefer<br>
<<a href="mailto:franziskuskiefer@gmail.com">franziskuskiefer@gmail.com</a>> wrote:<br>
><br>
> to make test vectors easier to use hacspec should also specify a common way<br>
> of writing them.<br>
<br>
</span>I think that's a good project, though a test vector language does not<br>
necessarily need to be coupled to a HacSpec language, I think.<br>
<span class=""><br>
<br>
> * Is JSON the right format to use here or do you think there are better<br>
> formats?<br>
<br>
</span>FWIW, I like the idea of using a rigid subset of JSON for test<br>
vectors, so it's easy to parse *either* with JSON or with dumb,<br>
hand-written parsers. You could imagine an interface to test simple<br>
devices like smartcards that don't have a full JSON parser. That's<br>
the approach we're taking with Noise:<br>
<br>
<a href="https://github.com/noiseprotocol/noise_wiki/wiki/Test-vectors" rel="noreferrer" target="_blank">https://github.com/<wbr>noiseprotocol/noise_wiki/wiki/<wbr>Test-vectors</a><br>
<a href="https://github.com/centromere/test-vector-spec/blob/master/output/test-vectors.pdf" rel="noreferrer" target="_blank">https://github.com/centromere/<wbr>test-vector-spec/blob/master/<wbr>output/test-vectors.pdf</a><br>
<br>
<br>
NIST ACVP and Wycheproof are also looking into JSON test vectors for<br>
crypto, so if you want to go further in that direction you could try<br>
to get some of those people here.<br>
<br>
<a href="https://github.com/usnistgov/ACVP/blob/master/artifacts/acvp_protocol.txt" rel="noreferrer" target="_blank">https://github.com/usnistgov/<wbr>ACVP/blob/master/artifacts/<wbr>acvp_protocol.txt</a><br>
<a href="https://github.com/google/wycheproof/tree/master/testvectors" rel="noreferrer" target="_blank">https://github.com/google/<wbr>wycheproof/tree/master/<wbr>testvectors</a><br>
<span class="HOEnZb"><font color="#888888"><br>
Trevor<br>
</font></span></blockquote></div><br></div>