Jeroen Massar

@azonenberg@ioc.exchange what filesystem does it go over though? transfer speed is one thing, but it the filesystem or the application read in small blocks you will have slowness.

@IPngNetworks@ublog.tech new ipng tool? is it going to spit out a bad list to BGP for auto-null route? ;)

@azonenberg@ioc.exchange @whitequark@social.treehouse.systems and you cannot even own it all... what a weird setup is that, is that like a downpayment and that one has to sell it when they do not like one anymore? ;)

@rmondello@hachyderm.io exports and more in the form: how to keep them alive and accessible when I lose my credentials (eg iCloud access), thus the option to export to another tool is important, but maybe a paper edition as a backup too?

@filippo @djc thanks to Filippo for the software (and many other things) @IPngNetworks runs one: https://ipng.ch/s/ct/ (well two, one sunlight, one tesseract) continuous 250mbit though is not compatible with most consumer internet though

@mntmn @crowdsupply Looking great, those will definitely find use in many homelabs and will allow 'recycling' hardware. I am personally awaiting MNT Rack: 1U / 2U / 4U cases: place one or multiple MNT Stations in there, along with options for modular nvme/U.2/SSD/HDD blocks, PCI bus block for eg GPU or network cards.... but then MNT has to expand production rapidly as with a decent price lots of people will want those :)

@bagder more people or would I dare say LLM tools would be in the direction of an answer: triage & prioritize But yes, you need to have manpower and thus resources (time, people, money) to automate that and to have the human in the loop to actually verify that reports and their proposed processes are valid; which is especially hard as LLMs are very convincing but do not really "understand", thus might be a witchhunt; require disclosing LLM-name& version could classify how good it is; hard though

@sash Great work Sasha (as many other things you have done! :), and thank you for responsibly disclosing it and patiently working with them to properly resolve it.

@sash@hachyderm.io Nice find, work and writeup! Escaping things and having a sane CSP policy is such a hard thing for crap-coded things…. because with a proper CSP this would not have been such a big issue; of course, properly escaping things is the better route…

@cR0w@infosec.exchange @ryanc@infosec.exchange did you DIY a Osmose system in there and mess it all up? ;) Drenched wet books are going to rot and stink at one point, but a hack is a hack. I guess the the little white box is a water/leak sensor?

@cR0w@infosec.exchange @ryanc@infosec.exchange yeah, the stack of books there looked indeed like a temp fix, good use for the doorstops ;) at least a good excuse to fix it properly.

@isotopp@infosec.exchange aber gibt zu "mehr FPU, AES- und RSA-Beschleuniger" nuzt der FPU fuer wahrungen (aber *100 geht auch); der AES/RSS fuer SSL & daten-privacy fuer kunden etc -- aber ja auf eine kiste von ~2015 macht das auch ;) Was vorteilhaft sein kan ist wie du sagt: weniger strom verbrauch und warme erzeugung. consoles/gaming gibt es nicht genugend "4k spielen" oder lass stehen leuten die >1080p TV haben @/cc @nblr@chaos.social

@markmorow@infosec.exchange and that is why many folks run homelabs or whole networks "for fun", as at scale it uses many people to work on it thus one never gets to touch all the bits, and one should not experiment with the prod network (next to few getting testing networks ;) ) hence a homelab/network solves that as one has full control, for all choices, good and bad, and often one plays with it in prod as it is the home network (though some sim their real network) and one can experiment and play with it and make all the choices; keeps one involved in things too.

@ryanc@infosec.exchange also gives you an option for OOB/IPMI alike setup -- though with 2 systems and 2 power supplies having one OOB the other would do the trick too

@azonenberg@ioc.exchange do you have as a doc a table of performance expectations? eg "a xeon X can do P" or "if you want to look at N channels of Y bandwidth you need TB3/4 with a 4ghz cpu and memory M" or maybe better asked, guidelines of certain machine types that would make one happy; possibly just "apple m1 MacBook pro = good for N" "5090 with CPU C == good upto Z"? this as thunderscopes are very likely to start appearing later in the year, thus might need to think about getting a proper host for it

@azonenberg@ioc.exchange thanks; exactly that a *indicator* of what one can expect with a certain combo of hardware; I think that can be very useful for people, inc me, to know what to expect as setting expectations is IMHO always a good thing; of course the caveat being that a complex setup will need more resources; thus maybe will have to score a fat box at one point, would be a good excuse ;)

@azonenberg@ioc.exchange so, is she saying "hey why are you two not married" ? ;) though from my POV, it is just a contract for taxes and inheritance, no real need if one like eachother

@jwz@mastodon.social @gregkh@social.kernel.org email was b4 right? https://b4.docs.kernel.org/en/latest/ :) Though @monsieuricon@social.kernel.org it is a bit funny that b4 code is on .... https://github.com/mricon/b4 ;)

@monsieuricon@social.kernel.org @jwz@mastodon.social @gregkh@social.kernel.org oh my, never noticed that one there ;) Thx!

@monsieuricon@social.kernel.org @jwz@mastodon.social @gregkh@social.kernel.org I know, as the real repo is here: https://git.kernel.org/pub/scm/utils/b4/b4.git/ The search engines of the world seem to always point to a single place though; the latter also does not have issue tracking. Fortunately there are a few projects trying to address distributed Forge, I am sure we'll see more of that over the coming year.

@huitema@social.secret-wg.org @SystemsAppr@discuss.systems With EDE (RFC8914) at least the server could indicate why it is sending an error. And then one knows the difference, assuming no lying or masking, between “.de was down/broken” and “.de is being filtered by the admin”