Good morning
π
Quiet afternoon at work yesterday, because of Slack being down
Good Morning! Yes, Slack being down was quite refreshing
Morning π Must admit, it felt weird working in a remote team without the usual method of chat.
rcfotd:
clojure.core/rsubseq
([sc test key] [sc start-test start-key end-test end-key])
sc must be a sorted collection, test(s) one of <, <=, > or
>=. Returns a reverse seq of those entries with keys ek for
which (test (.. sc comparator (compare ek key)) 0) is true
mΓ’ning!
morning
Morning
Morn'
anyone have any views on juxt/tick
vs vanilla cljc.java-time
? (the interval algebra stuff in tick looks neat, and will let me delete some code, but i'm not exactly sure yet what other things tick adds over cljc.java-time
)
Portability maybe?
what dyu mean @thomas?
As in clj and cljs implementations. Not sure what either provide nor what you need of course
But that could be a deciding factor
ah, i see - juxt/tick
builds on top of cljc.java-time
- both of them are clj+cljs compatible, so i don't think compatibility is an issue in this instance
i think what i'm looking for is a "you should use juxt/tick
rather than cljc.java-time
and it's worth the extra dependency and bitrot risk because ..."
which is no doubt something that someone who has used the lib/s for a while would easily be able tell me, but i can't seem to get it from the docs
i guess as somebody that's about to have to rip out a datetime lib in our project cos it's on the verge of rot... i'm a little suspicious of the non native option
but obv i am currently in typescript/js not clj(s) so i know it's apples and oranges
tl;dr tick offers some nice extras, at the cost of some stackoverflow-ability. the main thing on the horizon is the new Temporal api for js. when that's in browsers, I think that'd be preferred over the current js java.time impl (js-joda). not sure how that'll play out but hopefully tick protocols could be extended to that so it'll 'just work' for the majority of the existing tick api. cljc.java-time could maybe be made to work for Temporal but even if it could, it looks like a lot more work ... so less likely to happen hope that's helpful π
morning
good morning... just
aha - i hadn't noticed you were also a committer on tick @henryw374 - that would have been another useful datapoint π
given the likely better future-resilience of tick, i think i'll go with that - thanks @henryw374!
:thumbsup:
oooh just watched the channel 4 "alternative christmas message"
still in uncanny valley territory
seems like the lighting on her face isn't quite right
Have you watched Sassy Justice on YouTube?
That is really well done
but how much is bias because I know it to be fake
oh, no I have not
It features a certain well known soon-to-be-maybe-yes-maybe-no President of the US of A.
the lighting on his face is always wierd of course
presumably by design
<https://www.youtube.com/watch?v=9WfZuNceFDM>
presumably it depends on quantity and topology of neural net nodes
where more is not necessarily better
are there rules of thumb emerging about optimum toplogies for different tasks?
I guess we get to the point where slebs are using the deep fakes for real time blemuish removal
and suddenly noone knows what they really look like anyway
heh, the lack of professional makeup during the pandemic is noticeable
I guess deep fake makeup artistry is not yet a widespread thing
that Michael Caine somehow looks like the G man from an ingame "Half Life" cut scene
I agree the Trump is really good though
Julie Andrews is good too
The scarey thing is, that is now. Give it another 2-3, 5 years and it'll be perfect
true dat
although I'd rather they spent that computation budget on figuring out nuclear fusion...
climate crisis what now
much more important to spend money on serving ads and propping up late stage capitalism than saving the planet ofc