Tootfinder

Don’t ever stop learning…
TIL: ‘* ’ has a VERY special meaning in Perl regular expressions.
I have been working with regex since 1983 and with Perl regex since 1993 and it took a stream of pre-release fixes for potential runaways in SpamAssassin for me to look at the new code and go “HUH?” and go delving in the docs.
(and yes, SA code additions ALWAYS get a REAL code review. I don’t give a 1 on code I don’t understand. NO ‘LGTM’ reviews!)
#Sysadminnery

🎯 Faster decoding with DSpark draft tokens and confidence-scheduled verification (default max draft length: 5). SWA bounded replay cuts decoder prefill work by replaying only a bounded suffix
🔒 Batch-invariant inference keeps results identical across batch compositions, deterministic sampling returns identical outputs for identical inputs and seeds (both NVIDIA only). Structured outputs constrain generation to regex, JSON schema or EBNF grammar

#TIL: a single 2-part regex in Perl can sometimes be reduced from quadratic to linear time by decomposing it into a series of 2 simpler regexes.
This may not matter to you if you don't have to deal with adversarially invalid input data. But I work with email, so...
#Sysadminnery

Ever tried to validate the XML of a DMARC rua report?
"DMARC RFC 7489 points to the XML target namespace http://dmarc.org/dmarc-xml/0.1, which happens to contain a rua.xsd file. This is not the XML Schema from RFC 7489, but an older one. [...] the RFC 7489 schema is broken due to an incomplete namespace declaration.
Both schemata use an improper regex for IP addresses, which will trigger validation errors on valid IPv6 addresses.
Bottom line: Both schemata are garbage."
https://wander.science/projects/email/pyrua/