<br><br><div class="gmail_quote">2009/7/16 Gervase Markham <span dir="ltr"><<a href="mailto:gerv@mozilla.org">gerv@mozilla.org</a>></span><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="im">On 15/07/09 17:11, Alex Wulms wrote:<br>
> How about making crc32c the default, that must be supported by any crcsync<br>
> server and client but foresee in the protocol that other algorithms might be<br>
> added in future if pilot testing by early adapters might show that crc32<br>
> gives too many clashes in practice or if crc32 works fine for most people but<br>
> in a specific deployment with a specific usage scenario it turns out to be<br>
> too weak.<br>
<br>
</div>It may or may not be wise to allow changes of algorithms. But surely we<br>
can calculate mathematically the chances of a collision using various<br>
forms of CRC? We don't have to guess. Let's not complicate things<br>
unnecessarily, let's do the maths and pick the best trade-off between<br>
size and clashes. That might even be 16-bit, I don't know. If 1 in<br>
10,000 had to be retransmitted rather than 1 in 10,000,000 but there was<br>
a speed gain for the other 9,999, it might still be worth it.<br>
<div class="im"></div></blockquote><div><br>I agree we need to do the numbers (and also that supporting multiple crc choices is probably a bad idea, maybe add a hook for graceful modification in a v2 of the protocol if it is found to be needed.)<br>
<br>The issue with the numbers is that in the case of say a credit card transaction, if we have a strong hash failure we cannot simply retransmit the data, which means the user will see an error page. So this means it is not a simple trade-off of data saved vs. retransmissions. I will sit down later today and see if I can come up with some numbers and then run them past the list...<br>
<br>Toby<br> </div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div class="im"><br>
> Or am I now overcomplicating things?<br>
<br>
</div>I think so :-)<br>
<br>
Gerv<br>
<div><div></div><div class="h5">_______________________________________________<br>
Http-crcsync mailing list<br>
<a href="mailto:Http-crcsync@lists.laptop.org">Http-crcsync@lists.laptop.org</a><br>
<a href="http://lists.laptop.org/listinfo/http-crcsync" target="_blank">http://lists.laptop.org/listinfo/http-crcsync</a><br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br>This email is intended for the addressee only and may contain privileged and/or confidential information<br>