jan

Computer Science enthusiast

[Jonny @JonnyLomond]

@waterloo_intern

[jan @opfleyr]

@mylifcc @york7979 10 minutes is not normal for 30k loc. Are you using a low-powered device with few cores? How much memory? On my mac m2 air with 16gb ram, compiling a huge 500k loc project with 1100 deps (split into 90 crates) takes about 6 minutes. Incremental is 30s.

[jan @opfleyr]

@sanchogodinho @voidzerodev It's not buggy slopware

[jan @opfleyr]

@Honaaz @emryixo @lobster25777 @BellularGaming They are not doing it for the players.

[jan @opfleyr]

@AlexAegis @NobeliumMonster @smarttechtakes only because some Web Devs thought it was a really great idea to give every React Component a single object with all its props, resulting in the worst case scenario for inline caching when the object shape changes constantly (=many polymorphic or even megamorphic caches)

[jan @opfleyr]

@AlexAegis @NobeliumMonster @smarttechtakes I ran it through the Chrome profiler, I don't even want to THINK about why there are so many GC events. How are you creating this much garbage so quickly? I'm scared to dig deeper. Just how much time does the JIT spend on optimizing JS only to have to deoptimize right away, /?

[jan @opfleyr]

@duoleveling @EureliNyx_mw @Blake_C137 That's true, and you're in the right when asking people to stop making these comments. But this is also problematic:

[jan @opfleyr]

@EureliNyx_mw @Blake_C137 @HoshigamiTwins @duoleveling Agreed! "especially since we (...) are streaming even longer" is rude and weird.

[フブキ👑 Letourneau🫃(librarian) @NorthernIion_LP]

all streamers need to see this clip

[SmasherMagazine @smashermagazine]

itsnotgoodenuff_

[jan @opfleyr]

@Masterfrogg @MyuVT @SilvaniKillian Once you explain to me how you think Neuro works, we can talk about why you're wrong

[jan @opfleyr]

@MyuVT @SilvaniKillian It is in fact totally fine to like Neuro. It's an interesting project, and Vedal has managed to give an AI "soul"... but you have to understand that you're contradicting yourself when you talk about dislike towards AI, but liking Neuro. It's irrational.

[jan @opfleyr]

@MyuVT @SilvaniKillian How do you think Neuro generates convincing speech? How does she recognize images? What content was used to train the models Vedal uses for Neuro? Hope it wasn't the exact same stuff as the big bad corpos.

[Max0r 🔜 OFFKAI @realMax0r]

triple the fucking shark budget

[jan @opfleyr]

@LizardOrman @_neonsunset @awesomekling Yes: https://t.co/l6HrodJs2M
It is undoubtedly a case of sabotage of R4L by the DMA maintainer, but the response here from marcan really doesn't help.