A small tool to view real-world ActivityPub objects as JSON! Enter a URL
or username from Mastodon or a similar service below, and we'll send a
request with
the right
Accept
header
to the server to view the underlying object.
{
"@context": "https://www.w3.org/ns/activitystreams",
"type": "OrderedCollectionPage",
"orderedItems": [
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1214879148584357888",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "<a class=\"u-url mention\" href=\"https://www.minds.com/AndersStjernholm\" target=\"_blank\">@AndersStjernholm</a> kikker på retten til liv.<br /><br />Samtalen er fed, fordi begge er ærlige om deres mening, og ikke prøver at åle sig uden om de ubehagelige konsekvenser af deres holdning (det gælder desværre ikke for alle af Anders gæster: her kan der gå mange minutter med at få dem til bare at indrømme, at der er ubehagelige konsekvenser).<br /><br />Jeg tror min grænse for liv er kombinationen af Anders' mors og Anders':<br /><br />Menneskeliv kræver, at den vordende mor ønsker det. Men beslutningen skal tages, inden organismen kan overleve uden hjælp (altså trække vejret og lign - men ikke selv fange sin mad).<br /><br />Hvor er din grænse?<br /><br /><a href=\"https://www.podbean.com/media/share/pb-4fjxr-faa74a?utm_campaign=au_share_ep&utm_medium=dlink&utm_source=au_share\" target=\"_blank\">https://www.podbean.com/media/share/pb-4fjxr-faa74a?utm_campaign=au_share_ep&utm_medium=dlink&utm_source=au_share</a>",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [
{
"type": "Mention",
"href": "https://www.minds.com/api/activitypub/users/1009027715396804621",
"name": "@AndersStjernholm"
}
],
"url": "https://www.minds.com/newsfeed/1214879148584357888",
"published": "2021-03-06T10:16:00+00:00",
"source": {
"content": "@AndersStjernholm kikker på retten til liv.\n\nSamtalen er fed, fordi begge er ærlige om deres mening, og ikke prøver at åle sig uden om de ubehagelige konsekvenser af deres holdning (det gælder desværre ikke for alle af Anders gæster: her kan der gå mange minutter med at få dem til bare at indrømme, at der er ubehagelige konsekvenser).\n\nJeg tror min grænse for liv er kombinationen af Anders' mors og Anders':\n\nMenneskeliv kræver, at den vordende mor ønsker det. Men beslutningen skal tages, inden organismen kan overleve uden hjælp (altså trække vejret og lign - men ikke selv fange sin mad).\n\nHvor er din grænse?\n\nhttps://www.podbean.com/media/share/pb-4fjxr-faa74a?utm_campaign=au_share_ep&utm_medium=dlink&utm_source=au_share",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1214879148584357888/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1212135209241067520",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "Yesss... Så har jeg lige løst en tilfældig professorterning.<br /><br />Altså gå fra en terning med 6 ensfarvede sider til et tilfældigt mønster, som er magen til en anden.<br /><br />Det er overraskende meget sværere end at gå fra random til ensfarvede sider.",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1212135209241067520",
"published": "2021-02-26T20:32:34+00:00",
"source": {
"content": "Yesss... Så har jeg lige løst en tilfældig professorterning.\n\nAltså gå fra en terning med 6 ensfarvede sider til et tilfældigt mønster, som er magen til en anden.\n\nDet er overraskende meget sværere end at gå fra random til ensfarvede sider.",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1212135209241067520/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1200963064662274048",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "Du har set Blade Runner og du husker scenen, hvor Deckard zoomer ind. Og du tænker: \"Det der zoom det kan man jo ikke. Den information er ikke i billedet. Filmfup!\"<br /><a href=\"https://www.youtube.com/watch?v=hHwjceFcF2Q\" target=\"_blank\">https://www.youtube.com/watch?v=hHwjceFcF2Q</a><br /><br />Jeg har læst lidt op på GANs.<br /><br />Det er f*cking vildt!<br /><br />GAN er neurale net, der træner hinanden: Det ene net prøver at lave noget kunstigt og det andet net skal så afgøre om inputtet er kunstigt eller naturligt. Som f.eks. fotografier.<br /><br />Net 2 får et input, som enten er naturligt eller kunstigt. Hvis det gætter rigtigt, så bliver det belønnet og ellers straffet.<br /><br />Noget af inputtet til net 2 kommer fra naturlige kilder, mens resten laves af net 1. Hvis net 2 gætter, at inputtet fra net 1 er kunstigt, så bliver net 1 straffet. Ellers bliver net 1 belønnet.<br /><br />Med andre ord: Vi lærer et neuralt net at lyve og et andet net at være løgnedetektor.<br /><br />Mest kendt er nok: <a href=\"https://thispersondoesnotexist.com/\" target=\"_blank\">https://thispersondoesnotexist.com/</a> der giver et nyt computergenereret portræt hvert sekund.<br /><br />GANs kan mere end det. De kan også gætte på manglende information.<br /><br />Her er en gammel film, som er restaureret med (bl.a.) GAN:<br /><a href=\"https://www.youtube.com/watch?v=hZ1OgQL9_Cw\" target=\"_blank\">https://www.youtube.com/watch?v=hZ1OgQL9_Cw</a><br /><br />Detaljerne er løgn. Det er noget som det neurale net har gættet på.<br /><br />Hvis Deckards foto køres igennem et GAN, så er scenen fra Blade Runner pludselig inden for, hvad der måske kunne lade sig gøre.<br /><br />Men netop det, at systemet kan gætte på manglende information er fantastisk til komprimering af fotos. Se fotos på sidste side i denne artikel:<br /><a href=\"https://openaccess.thecvf.com/content_cvpr_2018_workshops/papers/w50/Agustsson_Extreme_Learned_Image_CVPR_2018_paper.pdf\" target=\"_blank\">https://openaccess.thecvf.com/content_cvpr_2018_workshops/papers/w50/Agustsson_Extreme_Learned_Image_CVPR_2018_paper.pdf</a><br /><br />Hvis du ikke ved bedre, så bruger du JPEG idag. Prøv lige at se kvalitetsforskellen. Og det er ved 0.036 bits per pixel mod JPEG eksemplets 0.248 bits per pixel (altså 7 gange større).<br /><br />Jeg ved ikke helt, hvormeget maskinkraft, som et trænet GAN skal bruge, men lur mig, om vi ikke kommer til at se det i mobiltelefoner - både til at rette op på udtværede fotos og til receiver end noise cancellation/høreapparater (hvor GAN'et fremhæver samtalepartnerens tale fra anden lyd optaget i den anden ende).<br />",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1200963064662274048",
"published": "2021-01-27T00:38:27+00:00",
"source": {
"content": "Du har set Blade Runner og du husker scenen, hvor Deckard zoomer ind. Og du tænker: \"Det der zoom det kan man jo ikke. Den information er ikke i billedet. Filmfup!\"\nhttps://www.youtube.com/watch?v=hHwjceFcF2Q\n\nJeg har læst lidt op på GANs.\n\nDet er f*cking vildt!\n\nGAN er neurale net, der træner hinanden: Det ene net prøver at lave noget kunstigt og det andet net skal så afgøre om inputtet er kunstigt eller naturligt. Som f.eks. fotografier.\n\nNet 2 får et input, som enten er naturligt eller kunstigt. Hvis det gætter rigtigt, så bliver det belønnet og ellers straffet.\n\nNoget af inputtet til net 2 kommer fra naturlige kilder, mens resten laves af net 1. Hvis net 2 gætter, at inputtet fra net 1 er kunstigt, så bliver net 1 straffet. Ellers bliver net 1 belønnet.\n\nMed andre ord: Vi lærer et neuralt net at lyve og et andet net at være løgnedetektor.\n\nMest kendt er nok: https://thispersondoesnotexist.com/ der giver et nyt computergenereret portræt hvert sekund.\n\nGANs kan mere end det. De kan også gætte på manglende information.\n\nHer er en gammel film, som er restaureret med (bl.a.) GAN:\nhttps://www.youtube.com/watch?v=hZ1OgQL9_Cw\n\nDetaljerne er løgn. Det er noget som det neurale net har gættet på.\n\nHvis Deckards foto køres igennem et GAN, så er scenen fra Blade Runner pludselig inden for, hvad der måske kunne lade sig gøre.\n\nMen netop det, at systemet kan gætte på manglende information er fantastisk til komprimering af fotos. Se fotos på sidste side i denne artikel:\nhttps://openaccess.thecvf.com/content_cvpr_2018_workshops/papers/w50/Agustsson_Extreme_Learned_Image_CVPR_2018_paper.pdf\n\nHvis du ikke ved bedre, så bruger du JPEG idag. Prøv lige at se kvalitetsforskellen. Og det er ved 0.036 bits per pixel mod JPEG eksemplets 0.248 bits per pixel (altså 7 gange større).\n\nJeg ved ikke helt, hvormeget maskinkraft, som et trænet GAN skal bruge, men lur mig, om vi ikke kommer til at se det i mobiltelefoner - både til at rette op på udtværede fotos og til receiver end noise cancellation/høreapparater (hvor GAN'et fremhæver samtalepartnerens tale fra anden lyd optaget i den anden ende).\n",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1200963064662274048/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1199297878363783168",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "Det er ikke rigtigt overraskende, men alligevel lidt sjovt.<br /><br />LIGO (<a href=\"https://www.ligo.caltech.edu/page/what-is-ligo\" target=\"_blank\">https://www.ligo.caltech.edu/page/what-is-ligo</a>) er et af verdens største fysikeksperimenter. De kikker efter tyngdebølger. Bølgerne laver variationer, der svarer til ca. 1 atoms bredde i forhold til afstanden mellem solen og jorden - altså væsentligt mindre end en krillimillimeter.<br /><br />LIGO har (ikke overraskende) en masse IT til at administrere deres systemer, og et af deres værktøjer er GNU Parallel.<br /><br />Så indirekte-indirekte er jeg med til at finde tyngdebølger.<br /><br />Det kan jeg godt blive lidt høj af.<br />",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1199297878363783168",
"published": "2021-01-22T10:21:36+00:00",
"source": {
"content": "Det er ikke rigtigt overraskende, men alligevel lidt sjovt.\n\nLIGO (https://www.ligo.caltech.edu/page/what-is-ligo) er et af verdens største fysikeksperimenter. De kikker efter tyngdebølger. Bølgerne laver variationer, der svarer til ca. 1 atoms bredde i forhold til afstanden mellem solen og jorden - altså væsentligt mindre end en krillimillimeter.\n\nLIGO har (ikke overraskende) en masse IT til at administrere deres systemer, og et af deres værktøjer er GNU Parallel.\n\nSå indirekte-indirekte er jeg med til at finde tyngdebølger.\n\nDet kan jeg godt blive lidt høj af.\n",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1199297878363783168/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1194674280294645760",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "find-first-fail: fine function for finding first failing file fragment (or numeric argument)<br /><br />Warning: Shameless self-promotion.<br /><br />If you are an IT guy you have tried this: Program P works for 10, but not for 1000000. At what number does P fail?<br /><br />First you try 500000. If that works you try 750000. If that fails, you try 625000. If that works you try 687500. And you continue this binary search until you are too tired to go again and thus feel you are close enough.<br /><br />This in itself is annoying - especially if you make a mistake, but it is even worse if it takes 10 minutes to test a single value: It is so long you really do not want to monitor it, but when it finishes you also do not want to waste time on you starting the next test.<br /><br />find-first-fail automates this.<br /><br />All you need to do is to provide find-first-fail with a program that takes a single numeric value. <br /><br />It finds the critical value by first testing the start value (which defaults to 1). As long as the value succeeds, the value is doubled. When the value fails, find-first-fail does a binary search between this value and the previous value.<br /><br />If the start value fails, find-first-fail instead searches for the highest value that command fails for.<br /><br />If given a file with -f find-first-fail will find minimal section in the file the command fails for.<br /><br />Just today I used it on a program that took 20 minutes to run, and *boy* is it nice that it automatically finds the next value and starts testing it. If I had to find the value 1073741763 by hand, it would have taken 30 runs of each 20 mins + the time where it would be idle before I started next test. Now it is suddenly not a huge task: You start it when you leave work, and it is done when you get back the next day.<br /><br /><a href=\"https://gitlab.com/ole.tange/tangetools/-/tree/master/find-first-fail\" target=\"_blank\">https://gitlab.com/ole.tange/tangetools/-/tree/master/find-first-fail</a><br />",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1194674280294645760",
"published": "2021-01-09T16:09:04+00:00",
"source": {
"content": "find-first-fail: fine function for finding first failing file fragment (or numeric argument)\n\nWarning: Shameless self-promotion.\n\nIf you are an IT guy you have tried this: Program P works for 10, but not for 1000000. At what number does P fail?\n\nFirst you try 500000. If that works you try 750000. If that fails, you try 625000. If that works you try 687500. And you continue this binary search until you are too tired to go again and thus feel you are close enough.\n\nThis in itself is annoying - especially if you make a mistake, but it is even worse if it takes 10 minutes to test a single value: It is so long you really do not want to monitor it, but when it finishes you also do not want to waste time on you starting the next test.\n\nfind-first-fail automates this.\n\nAll you need to do is to provide find-first-fail with a program that takes a single numeric value. \n\nIt finds the critical value by first testing the start value (which defaults to 1). As long as the value succeeds, the value is doubled. When the value fails, find-first-fail does a binary search between this value and the previous value.\n\nIf the start value fails, find-first-fail instead searches for the highest value that command fails for.\n\nIf given a file with -f find-first-fail will find minimal section in the file the command fails for.\n\nJust today I used it on a program that took 20 minutes to run, and *boy* is it nice that it automatically finds the next value and starts testing it. If I had to find the value 1073741763 by hand, it would have taken 30 runs of each 20 mins + the time where it would be idle before I started next test. Now it is suddenly not a huge task: You start it when you leave work, and it is done when you get back the next day.\n\nhttps://gitlab.com/ole.tange/tangetools/-/tree/master/find-first-fail\n",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1194674280294645760/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1160600140718608384",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "Radius har flere gange skrevet, at _nu_ ville de gerne sende en tekniker til at skifte min privatlivsbeskyttende elmåler til en overvågningsmåler. De brugte ikke præcis de ord, men det var indholdet.<br /><br />Jeg ringede til Radius og sagde, at de ikke kunne få adgang til at skifte min måler, da deres nye måler ikke overholdt GDPR.<br /><br />Det havde de aldrig hørt om. Og det mente de naturligvis at den gjorde. Jeg fik dog en god snak med en tekniker, som bekræftede, at måleren sendte data om mit forbrug i kWh (ikke i DKK) flere gange i døgnet.<br /><br />Mit el-forbrug tilhører mit privatliv, for den information kan bruges imod mig. Hvis man skulle være i tvivl behøver man kun at læse <a href=\"https://www.altinget.dk/forsyning/artikel/professor-regeringen-presser-masseovervaagning-ind-i-hjemmet\" target=\"_blank\">https://www.altinget.dk/forsyning/artikel/professor-regeringen-presser-masseovervaagning-ind-i-hjemmet</a><br /><br />GDPR §25 siger, at man skal dataminimere (ikke samle flere/andre data end det, som man skal bruge), og implementere state-of-the-art databeskyttelsesprincipper.<br /><a href=\"https://www.privacy-regulation.eu/en/article-25-data-protection-by-design-and-by-default-GDPR.htm\" target=\"_blank\">https://www.privacy-regulation.eu/en/article-25-data-protection-by-design-and-by-default-GDPR.htm</a><br /><br />Over for Radius beskrev jeg ideen om en elmåler, der får prisen via nettet hver time. Den pris ganger den sammen med mit forbrug og summerer op. En gang om året rapporterer den så det samlede beløb til elselskabet. Ikke mit forbrug i kWh, men mit forbrug i DKK.<br /><br />Simpelt, fjernaflæst og privatlivsbeskyttende.<br /><br />Jeg hævdede, at en sådan måler ville være tættere på state-of-the-art mht. databeskyttelse end deres nuværende måler, og at deres måler derfor ikke levede op til kravet om state-of-the-art beskyttelse af mit privatliv.<br /><br />(Der findes endda endnu bedre løsninger. Se links på <a href=\"https://www.version2.dk/blog/fjernaflaeste-el-maalere-spionerer-mod-dig-1086035\" target=\"_blank\">https://www.version2.dk/blog/fjernaflaeste-el-maalere-spionerer-mod-dig-1086035</a> )<br /><br />Det var de uenige i, og de truede med at tvinge sig adgang med fogedens hjælp.<br /><br />Og det er der sagen er nu.<br /><br />Jeg regner med, at jeg nok får en besked fra fogeden, før han kikker forbi; og når det sker, så forventer jeg at jeg kan fremføre mine argumenter for en dommer.<br />",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1160600140718608384",
"published": "2020-10-07T15:30:36+00:00",
"source": {
"content": "Radius har flere gange skrevet, at _nu_ ville de gerne sende en tekniker til at skifte min privatlivsbeskyttende elmåler til en overvågningsmåler. De brugte ikke præcis de ord, men det var indholdet.\n\nJeg ringede til Radius og sagde, at de ikke kunne få adgang til at skifte min måler, da deres nye måler ikke overholdt GDPR.\n\nDet havde de aldrig hørt om. Og det mente de naturligvis at den gjorde. Jeg fik dog en god snak med en tekniker, som bekræftede, at måleren sendte data om mit forbrug i kWh (ikke i DKK) flere gange i døgnet.\n\nMit el-forbrug tilhører mit privatliv, for den information kan bruges imod mig. Hvis man skulle være i tvivl behøver man kun at læse https://www.altinget.dk/forsyning/artikel/professor-regeringen-presser-masseovervaagning-ind-i-hjemmet\n\nGDPR §25 siger, at man skal dataminimere (ikke samle flere/andre data end det, som man skal bruge), og implementere state-of-the-art databeskyttelsesprincipper.\nhttps://www.privacy-regulation.eu/en/article-25-data-protection-by-design-and-by-default-GDPR.htm\n\nOver for Radius beskrev jeg ideen om en elmåler, der får prisen via nettet hver time. Den pris ganger den sammen med mit forbrug og summerer op. En gang om året rapporterer den så det samlede beløb til elselskabet. Ikke mit forbrug i kWh, men mit forbrug i DKK.\n\nSimpelt, fjernaflæst og privatlivsbeskyttende.\n\nJeg hævdede, at en sådan måler ville være tættere på state-of-the-art mht. databeskyttelse end deres nuværende måler, og at deres måler derfor ikke levede op til kravet om state-of-the-art beskyttelse af mit privatliv.\n\n(Der findes endda endnu bedre løsninger. Se links på https://www.version2.dk/blog/fjernaflaeste-el-maalere-spionerer-mod-dig-1086035 )\n\nDet var de uenige i, og de truede med at tvinge sig adgang med fogedens hjælp.\n\nOg det er der sagen er nu.\n\nJeg regner med, at jeg nok får en besked fra fogeden, før han kikker forbi; og når det sker, så forventer jeg at jeg kan fremføre mine argumenter for en dommer.\n",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1160600140718608384/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1139656632167403520",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "Nå, men jeg testede lige min nye filserver.<br /><br />250.000.000 filer.<br /><br />Det tager overraskende lang tid at lave dem, når man kun kan lave 1000 i sekundet. Men mere overraskende tager det 3 gange så lang tid at slette dem.<br /><br />Så de næste 10 døgn(!) ved min filserver godt hvad den skal lave.<br /><br />Og giver det så nogen forskel om man sletter dem i parallel? Nej det gør det ikke. Faktisk er der en svag indikation af, at det er hurtigst at slette dem i seriel (men det er inden for måleunøjagtigheden).<br /><br />Øv.<br /><br />Jeg er lidt irriteret over, at ZFS ikke kan bruge Slog-disken (som er SSD) til lige at cache alle ændringerne og så skrive dem tilbage af een gang. Men ak, det ser ud til, at IO alene foregår på de magnetiske diske.<br />",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1139656632167403520",
"published": "2020-08-10T20:28:34+00:00",
"source": {
"content": "Nå, men jeg testede lige min nye filserver.\n\n250.000.000 filer.\n\nDet tager overraskende lang tid at lave dem, når man kun kan lave 1000 i sekundet. Men mere overraskende tager det 3 gange så lang tid at slette dem.\n\nSå de næste 10 døgn(!) ved min filserver godt hvad den skal lave.\n\nOg giver det så nogen forskel om man sletter dem i parallel? Nej det gør det ikke. Faktisk er der en svag indikation af, at det er hurtigst at slette dem i seriel (men det er inden for måleunøjagtigheden).\n\nØv.\n\nJeg er lidt irriteret over, at ZFS ikke kan bruge Slog-disken (som er SSD) til lige at cache alle ændringerne og så skrive dem tilbage af een gang. Men ak, det ser ud til, at IO alene foregår på de magnetiske diske.\n",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1139656632167403520/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1134066995921907712",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "I have watched \"Coded Bias\".<br /><br />I agree with the general problem, but I have issues with the deceptive way it is presented.<br /><br />E.g. It is seen as a problem that facial recognition works worse if you are not a white male. As a white male myself I wish the opposite was true: As it is now it is easier to track me through facial recognition than, say, a colored woman. And my guess is that _if_ the opposite had been true the movie would have found that even worse: \"Why is it optimized for finding black people?\"<br /><br />The problem is not the accuracy of the system. The problem is the system.<br /><br />E.g. Why is it a problem if someone attacks your science? If your science is sound, then it holds up to scrutiny. The movie makes it sound as if the reason why her science was attacked was because she was a woman of color.<br /><br />But ALL science is being scrutinized, and that is a good thing: That is how we weed out bad science from good science.<br /><br />E.g. The Stanislav Petrov incident. Here the movie gives the impression that Petrov did not \"press the button\" because he did not want the Americans to also die. That was not the reason. The reason was, that he (correctly) assumed the system was malfunctioning.<br /><br />The conclusion remains, however, unchanged: If Petrov had been a robot, it would have been bad.<br /><br />These were just the 3 I caught on a first viewing without even looking for them, and it makes me feel that there are many more.<br /><br />This harms the trustworthiness of the movie, and that is a shame, because I do not think the flow in the movie would suffer greatly, if the deception was removed. It is also really a shame because the general debate about AI and surveillance is a very important debate.<br /><br />As it stands I have a hard time recommending it. Maybe to people who knows absolutely nothing about the issue.<br /><br />But if the deception was cut from the movie, I might change my mind.<br /><br /><a href=\"https://www.imdb.com/title/tt11394170/\" target=\"_blank\">https://www.imdb.com/title/tt11394170/</a><br /><br /><a href=\"https://www.minds.com/search?f=top&t=all&q=CodedBias\" title=\"#CodedBias\" class=\"u-url hashtag\" target=\"_blank\">#CodedBias</a> ",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1134066995921907712",
"published": "2020-07-26T10:17:21+00:00",
"source": {
"content": "I have watched \"Coded Bias\".\n\nI agree with the general problem, but I have issues with the deceptive way it is presented.\n\nE.g. It is seen as a problem that facial recognition works worse if you are not a white male. As a white male myself I wish the opposite was true: As it is now it is easier to track me through facial recognition than, say, a colored woman. And my guess is that _if_ the opposite had been true the movie would have found that even worse: \"Why is it optimized for finding black people?\"\n\nThe problem is not the accuracy of the system. The problem is the system.\n\nE.g. Why is it a problem if someone attacks your science? If your science is sound, then it holds up to scrutiny. The movie makes it sound as if the reason why her science was attacked was because she was a woman of color.\n\nBut ALL science is being scrutinized, and that is a good thing: That is how we weed out bad science from good science.\n\nE.g. The Stanislav Petrov incident. Here the movie gives the impression that Petrov did not \"press the button\" because he did not want the Americans to also die. That was not the reason. The reason was, that he (correctly) assumed the system was malfunctioning.\n\nThe conclusion remains, however, unchanged: If Petrov had been a robot, it would have been bad.\n\nThese were just the 3 I caught on a first viewing without even looking for them, and it makes me feel that there are many more.\n\nThis harms the trustworthiness of the movie, and that is a shame, because I do not think the flow in the movie would suffer greatly, if the deception was removed. It is also really a shame because the general debate about AI and surveillance is a very important debate.\n\nAs it stands I have a hard time recommending it. Maybe to people who knows absolutely nothing about the issue.\n\nBut if the deception was cut from the movie, I might change my mind.\n\nhttps://www.imdb.com/title/tt11394170/\n\n#CodedBias ",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1134066995921907712/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1125760124238856192",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "\"Det kan være fordi man ikke vil konfirmeres for gavernes skyld [at man ikke ønsker at bliver konfirmeret],\" siger Biskob Peter Fischer Møller.<br /><br />Jep, PFM. Det kunne helt sikkert være af den grund.<br /><br /><a href=\"https://www.radio4.dk/program/gud-fader-bevares/\" target=\"_blank\">https://www.radio4.dk/program/gud-fader-bevares/</a>",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1125760124238856192",
"published": "2020-07-03T12:08:49+00:00",
"source": {
"content": "\"Det kan være fordi man ikke vil konfirmeres for gavernes skyld [at man ikke ønsker at bliver konfirmeret],\" siger Biskob Peter Fischer Møller.\n\nJep, PFM. Det kunne helt sikkert være af den grund.\n\nhttps://www.radio4.dk/program/gud-fader-bevares/",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1125760124238856192/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1112827514168336384",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "Jeg har brugt UNIX dagligt siden 1991 (hold da op, det er 29 år!).<br /><br />Jeg har nu for første gang brugt:<br /><br /> exec 21<file<br /><br />Jeg tænker, det ikke er funktionalitet, som er *super* vigtig.<br /><br />Har du nogensinde brugt 'exec 21<file'?<br /><br /><a href=\"https://gitlab.com/ole.tange/tangetools/-/blob/master/neno/neno\" target=\"_blank\">https://gitlab.com/ole.tange/tangetools/-/blob/master/neno/neno</a>",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1112827514168336384",
"published": "2020-05-28T19:39:14+00:00",
"source": {
"content": "Jeg har brugt UNIX dagligt siden 1991 (hold da op, det er 29 år!).\n\nJeg har nu for første gang brugt:\n\n exec 21<file\n\nJeg tænker, det ikke er funktionalitet, som er *super* vigtig.\n\nHar du nogensinde brugt 'exec 21<file'?\n\nhttps://gitlab.com/ole.tange/tangetools/-/blob/master/neno/neno",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1112827514168336384/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1101521768409948160",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "Nemmeste måde at få 100-1000 virtuelle maskiner i luften?<br /><br />Lad os antage, at du har en 0.5 TB 48 CPU maskine til rådighed med Ubuntu 18.04, og du gerne vil have den til at køre 100-1000 virtuelle maskiner. De skal alle kunne nås via SSH, men må gerne have private IP-adresser.<br /><br />Hvad er den nemmeste måde at gøre det på? Kan det klares på 10 linjer kode og lidt ventetid?<br />",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1101521768409948160",
"published": "2020-04-27T14:54:14+00:00",
"source": {
"content": "Nemmeste måde at få 100-1000 virtuelle maskiner i luften?\n\nLad os antage, at du har en 0.5 TB 48 CPU maskine til rådighed med Ubuntu 18.04, og du gerne vil have den til at køre 100-1000 virtuelle maskiner. De skal alle kunne nås via SSH, men må gerne have private IP-adresser.\n\nHvad er den nemmeste måde at gøre det på? Kan det klares på 10 linjer kode og lidt ventetid?\n",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1101521768409948160/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1099836723213217792",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "Hvorfor er Zoom er dårlig ide, og hvorfor er en self-hosted Jitsi-server en bedre ide:<br /><br /><a href=\"https://citizenlab.ca/2020/04/move-fast-roll-your-own-crypto-a-quick-look-at-the-confidentiality-of-zoom-meetings/\" target=\"_blank\">https://citizenlab.ca/2020/04/move-fast-roll-your-own-crypto-a-quick-look-at-the-confidentiality-of-zoom-meetings/</a>",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1099836723213217792",
"published": "2020-04-22T23:18:28+00:00",
"source": {
"content": "Hvorfor er Zoom er dårlig ide, og hvorfor er en self-hosted Jitsi-server en bedre ide:\n\nhttps://citizenlab.ca/2020/04/move-fast-roll-your-own-crypto-a-quick-look-at-the-confidentiality-of-zoom-meetings/",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1099836723213217792/activity"
},
{
"type": "Create",
"actor": "https://www.minds.com/api/activitypub/users/955127973680259077",
"object": {
"type": "Note",
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1099816289864364032",
"attributedTo": "https://www.minds.com/api/activitypub/users/955127973680259077",
"content": "10 years anniversary release of GNU Parallel: GNU Parallel 20200422 ('10years') released [stable]<br />pi.dk/2",
"to": [
"https://www.w3.org/ns/activitystreams#Public"
],
"cc": [
"https://www.minds.com/api/activitypub/users/955127973680259077/followers"
],
"tag": [],
"url": "https://www.minds.com/newsfeed/1099816289864364032",
"published": "2020-04-22T21:57:17+00:00",
"source": {
"content": "10 years anniversary release of GNU Parallel: GNU Parallel 20200422 ('10years') released [stable]\npi.dk/2",
"mediaType": "text/plain"
}
},
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/entities/urn:activity:1099816289864364032/activity"
}
],
"id": "https://www.minds.com/api/activitypub/users/955127973680259077/outbox",
"partOf": "https://www.minds.com/api/activitypub/users/955127973680259077/outboxoutbox"
}