Day changed to 2021-08-01
[03:01] jfw: some further detail on the recent lulz worth highlighting for future reference: https://www.reuters.com/world/us/majority-covid-19-cases-large-public-events-were-among-vaccinated-us-cdc-study-2021-07-30/ ( https://archive.is/Oe0ci )
[03:04] jfw: to be fair & balanced, this correlation doesn't necessarily prove that the genetherapyvaccine (genocine?) was the cause of the disease; they could both be caused by a common factor such as for example being from Massachusetts.
[03:13] jfw: Medical experts were last seen spinning themselves into basketball-sized balls of yarn over how to scare people enough to mask back up while simultaneously maintaining the stature of their miracle cure.
[03:16] jfw: Fortunately evolution has well equipped the human brain to jump back and forth between logic and madness without noticing or skipping a beat.
[23:42] whaack: manual feedbot: http://ztkfg.com/2021/08/blockexplorer-web-interface-and-the-updated-todo-list/
[23:50] whaack: mom was giving me a speech on how i would need to get the vaccine if i went to visit the fam. i don't think i'd be barred from visiting without it, but there'd be an attempted guilt trip every other meal
Day changed to 2021-08-02
[00:19] jfw: whaack: doesn't sound like much fun.
[00:21] jfw: whaack: broken link in http://ztkfg.com/2021/07/warning-bitcoins-stored-in-segwit-addresses-are-not-safe/#comment-610
[00:27] jfw: whaack: indeed sqlite isn't the best for structural changes like renaming columns, but the docs give some relatively reasonable approaches (which will essentially involve copying the data to a new table, so would still take a while but not like the full scan)
[00:31] jfw: 1/2 TB - sounds not too bad considering what it is
[00:50] jfw: "For unknown scriptPubkeys, the hex representation of the scriptPubkeys is displayed instead." - seems the cleaner & less surprising approach there would be two separate fields one of which may be empty/null.
[00:50] jfw: output fields that is.
[00:53] jfw: anyway, loads & works fine for me. it's a CLI prog dressed in web delivery rather than much of a web interface, as you note, but looks like a good start.
[00:56] jfw: trips up on excess whitespace.
[00:58] jfw: the block view perhaps needs to say how many transactions to provide a bound for the get-txn coordinates, or otherwise something to list the txns in a given block.
[01:00] jfw: I like the hash or coordinates to specify a txn, and yeah, would be nice to DWIM for any identifier.
[01:01] jfw: should be pretty easy too, 160bits = address, 256bits = txn or block - try both and see what works, then could optimize by noting that leading zeros probably means block.
[01:06] jfw: did you find how to plug python into apache without importing all sorts?
[16:10] whaack: jfw: thanks for trying it out and providing suggestions
[16:12] whaack: jfw: re python and apache, no, i sinned in order to get this online
Day changed to 2021-08-05
[20:13] whaack: dorion: package arrived at my local mail office, shipping price ended up being $80. I haven't had the chance yet to pick it up, will likely do so tomorrow
Day changed to 2021-08-06
[03:30] jfw: whaack: not too bad I suppose if it's arrived undamaged - good luck!
[03:31] jfw: in random glimpses of how far things have decayed in the mines, I am paid to quickly select tools and never invent any ... If I don't host a new company project on GitHub I'll eventually be fired and replaced with somebody who will.
[03:32] jfw: apparently apt-get installing a git daemon is considered inventing your own tools now or something?
[17:32] jfw: rereading the above quote a bit deeper, it strikes me as self-contradictory. First it's that she's paid for the valuable work of picking the right tool for the job from the modern deluge & correctly identifying when inventing new ones is actually a business necessity (there was that exception to the "never" in the elided part). But then it's that Github is a foregone conclusion, and not based on
[17:32] jfw: her own judgement but that of the higher-ups. But exam-taking can't possibly be worth paying salary for - the answer key is cheaper.
[17:35] jfw: She's presumably paid for something, but perhaps has not correctly identified what. And not knowing any more about the situation I'd be ill equipped to guess.
Day changed to 2021-08-07
[01:08] whaack: package received, looks good on first pass! gonna do a more thorough inspection in the next 3 days
[11:37] dorion: whaack, sounds good. 3 usb-ttl cables are also on the way (you may already have some from fg usage). I'll also gpg you more details about the install.
[14:16] whaack: dorion: ah when did you send them?
[17:28] dorion: whaack, tracking number in pm
[17:31] dorion waves farewell from vermont, about to hop on a train and be in Panama tomorrow morning.
[17:39] whaack: dorion: which company did you ship with? tracking number did not show up in ups or fedex's site.
[18:07] dorion: ups, correction in pm.
[18:09] jfw: possibly "full tracking URL" is the proper way to do this, though it wouldn't surprise me if most of the tracking services use enough java technology as to make working URLs an impossibility.
[18:17] whaack: new one works, thx
[18:18] whaack: from my experience typing a tracking id in goolag search engine will find you the link one needs
Day changed to 2021-08-08
[19:51] dorion waves from el istmo de panama
[22:17] whaack: dorion: congrats!!
Day changed to 2021-08-09
[02:04] jfw: dorion: splendid. we're a multinational again!
[02:05] jfw: perhaps for the first time, even.
[19:17] whaack: jfw: atm i plan to have my block explorer return the data in some "awk friendly" format, it will look something like what i currently have with a few minor adjustments. then i plan to release a corresponding client python script that grabs the data and organizes it. i have a sense that i may be doing 'the wrong thing' by not using a 'standard' data format such as json, xml, or sexprs, but otoh
[19:17] whaack: with this method i avoid writing extra code for handling different types of requests (human readable vs api mode) and the overall LoC of my program (when i include imports) likely goes down dramatically for not requiring a tool such as json
[19:55] jfw: whaack: ASCII with delimiters seems pretty standard to me... like I said though it depends on the data, or more specifically how complex it is - character set, list or table vs tree structure
[19:55] sourcerer: 2021-05-24 21:44:50 (#jwrd) jfw: whaack: depends on the data, if it's just a table of strings with controlled range of characters it may be simplest to just pick a delimiter character.
[19:56] jfw: personally I would not use XML for anything, given the choice (and possibly even otherwise)
[19:59] jfw: if you want a friendly web interface you're going to want to html-wrap the data in one way or another, so separate marked up and raw views is perhaps unavoidable.
[19:59] jfw: what sort of organizing would that client script be required for?
[20:05] jfw: fwiw, printing a tree as JSON isn't that difficult, and even parsing isn't too bad especially if you ignore unicode.
[20:07] jfw: but it seems overkill to me for this application; the most tree-like thing is the transaction listings, but even that is a uniform structure only a couple levels deep, it's basically just making explicit the grouping of input/output records in the table.
[20:09] whaack: alright, i'll go without using json etc. then, and just have a public spec for how the data is outputed with an example client that reads it
[20:10] jfw: (if you want an example of a widely used standard that works just about this way: HTTP headers!)
[20:17] whaack: jfw: heh, i just looked it up, and looks like my format is 90% of the way there
[20:39] jfw: to be clear, I don't mean it as something to aim for - it has some excess baggage like wrapped lines and parenthesis matching, iirc. it's just an example of simple text formats working fine and being widely implemented.
Day changed to 2021-08-24
[03:21] jfw: On a fresh reread of djb's /package scheme, parts of it still appeal to me, like using the filesystem to provide a package-to-file mapping, and having stable install paths no matter the distribution and status as "user" or
[03:21] jfw: "local-user" or "system" or whatever arbitrary distinction.
[03:23] jfw: the most dubious part is the "globally allocated" namespace, without the republican machinery eg strong identity, punishment gazette for dealing with such Gavins as would defile the namespace.
[03:25] jfw: funny he illustrates with some relatively minor instances of namespace collision inconveniences but missed the kind that matters most, i.e. the presumed maintainer going rogue.
[03:28] jfw: his own registry, based on metadata from my archives, was last updated 2012 at the latest - not counting this March where he substituted his own links with https.
[03:34] jfw: it's got 80 entries pointing to 11 domains, of which 16 never got a link, and a good portion from a small sample of the links are turning up dead.
[03:35] jfw: it's all supposedly "discussed on the mailing list" for which he apparently never saw any need to publish archives.
[03:38] jfw: he never updated his own pre-daemontools works, i.e. most of the good stuff, to use the scheme.
[03:44] jfw: (that was back on /package rather than the namespacing per se.) then the docs are woefully lacking in the "okay so how do I implement this for my shiny new program?" department, various details left unspecified, at best implicit by trying to emulate his examples, and unenforced, such as the internal layout, how to install, and metadata in the <package-name>/package/ subdir.
[03:53] jfw: In this light, I figure I'll stick with parts of it but continue happily not registering my names with the authorities, and /command has to go. Yet another place to search for binaries and find perpetually missing from $PATH, that ~nothing else uses? /bin and /usr/bin would be way costlier to snip due to all the #! scripts at the very least; /usr/local can perfectly well be ignored, not like FHS
[03:53] jfw: fanatic distributors can stop you using "the system's" namespace when you're distributing directly. There's some argument to avoiding /bin for systems where it's distinct from /usr/bin, eg readonly root filesystem, recovery environment with basic tools only or whatnot.
[03:56] jfw: They're not distinct on Gales, as I don't find those things convincing; but people have in the past so such systems exist.
[03:59] jfw: So my approach will be: tarball extracts (or tree presses or whatever) to /package/<packagename>-<version> as I've used before, compiles in place, and executables get symlinked in /usr/bin.
[04:00] jfw: Undecided whether to stick with README and "install" script in the "package" subdir or move them to top-level but leaning toward the latter.
[04:03] jfw: (huh, now that coulda made a fine mini-blog article, and here I thought I had just a couple-line update to express)
[04:10] jfw: and I suppose patching daemontools on Gales to switch from /command would be in order too.
[15:41] dorion: http://fixpoint.welshcomputing.com/2021/jwrd-logs-for-Aug-2021/#2869 -- makes sense to me to get rid of /command and just go with /usr/bin
[15:41] sourcerer: 2021-08-24 03:53:02 (#jwrd) jfw: In this light, I figure I'll stick with parts of it but continue happily not registering my names with the authorities, and /command has to go. Yet another place to search for binaries and find perpetually missing from $PATH, that ~nothing else uses? /bin and /usr/bin would be way costlier to snip due to all the #! scripts at the very least; /usr/local can perfectly well be ignored, not like FHS
[15:42] dorion: http://fixpoint.welshcomputing.com/2021/jwrd-logs-for-Aug-2021/#2874 -- it still can become a blog article, probably good to turn this thread into one to make it easier to reference. nothing wrong with first draft in log and final draft as article.
[15:42] sourcerer: 2021-08-24 04:03:55 (#jwrd) jfw: (huh, now that coulda made a fine mini-blog article, and here I thought I had just a couple-line update to express)
[15:44] dorion: http://fixpoint.welshcomputing.com/2021/jwrd-logs-for-Aug-2021/#2867 -- calls to mind the lack of response to fg/phuctor emails alf reported, iirc.
[15:44] sourcerer: 2021-08-24 03:38:17 (#jwrd) jfw: he never updated his own pre-daemontools works, i.e. most of the good stuff, to use the scheme.
[21:06] jfw: there was the notion of the old vs. new & "improved" djb; though seeing as how I don't quite buy into demonic possession I'm tending to think the old wasn't as solid as it might have appeared.
[21:17] jfw: technically brilliant but politically inept perhaps, got grabbed by the dark side, not because the masses have any use for people like him but so as to keep him safely busy and out of further trouble.
[21:21] jfw: maybe-2003 djb fighting for the unknowing & uncaring masses
[23:03] dorion: in better news : http://ossasepia.com/2021/08/23/its-called-vamp/
Day changed to 2021-08-30
[19:34] jfw: Before the month's log closes, I'd like to point out since no one else did publicly that I may have dunning-krugered it a bit with http://fixpoint.welshcomputing.com/2021/jwrd-logs-for-Aug-2021/#2806 , in the sense of concluding much of anything at all based on such (expectedly) flimsy science reporting.
[19:34] sourcerer: 2021-08-01 03:01:54 (#jwrd) jfw: some further detail on the recent lulz worth highlighting for future reference: https://www.reuters.com/world/us/majority-covid-19-cases-large-public-events-were-among-vaccinated-us-cdc-study-2021-07-30/ ( https://archive.is/Oe0ci )
[19:39] jfw: it was an attempt to at least realize some kind of value from the time spend reading the stuff; but perhaps the chief lesson is more along the lines of, don't buy into what any statistics seem to be saying if you can't bring actual statistical expertise to bear, given how many ways there are to go wrong.
[19:42] jfw: also "gene therapy" isn't quite an accurate label, as the mRNA vaccines don't reportedly work by altering cellular DNA, more just altering their function to produce the viral spike protein - presumably only temporarily while the mRNA supply lasts.
[19:42] jfw: *altering cellular function
[19:44] jfw: AFAIK it still stands that it's a novel type of treatment, markedly different from prior notions of what vaccines are, and while in development for some time, had not been previously tested at population scale.
[03:01] jfw: some further detail on the recent lulz worth highlighting for future reference: https://www.reuters.com/world/us/majority-covid-19-cases-large-public-events-were-among-vaccinated-us-cdc-study-2021-07-30/ ( https://archive.is/Oe0ci )
[03:04] jfw: to be fair & balanced, this correlation doesn't necessarily prove that the genetherapyvaccine (genocine?) was the cause of the disease; they could both be caused by a common factor such as for example being from Massachusetts.
[03:13] jfw: Medical experts were last seen spinning themselves into basketball-sized balls of yarn over how to scare people enough to mask back up while simultaneously maintaining the stature of their miracle cure.
[03:16] jfw: Fortunately evolution has well equipped the human brain to jump back and forth between logic and madness without noticing or skipping a beat.
[23:42] whaack: manual feedbot: http://ztkfg.com/2021/08/blockexplorer-web-interface-and-the-updated-todo-list/
[23:50] whaack: mom was giving me a speech on how i would need to get the vaccine if i went to visit the fam. i don't think i'd be barred from visiting without it, but there'd be an attempted guilt trip every other meal
Day changed to 2021-08-02
[00:19] jfw: whaack: doesn't sound like much fun.
[00:21] jfw: whaack: broken link in http://ztkfg.com/2021/07/warning-bitcoins-stored-in-segwit-addresses-are-not-safe/#comment-610
[00:27] jfw: whaack: indeed sqlite isn't the best for structural changes like renaming columns, but the docs give some relatively reasonable approaches (which will essentially involve copying the data to a new table, so would still take a while but not like the full scan)
[00:31] jfw: 1/2 TB - sounds not too bad considering what it is
[00:50] jfw: "For unknown scriptPubkeys, the hex representation of the scriptPubkeys is displayed instead." - seems the cleaner & less surprising approach there would be two separate fields one of which may be empty/null.
[00:50] jfw: output fields that is.
[00:53] jfw: anyway, loads & works fine for me. it's a CLI prog dressed in web delivery rather than much of a web interface, as you note, but looks like a good start.
[00:56] jfw: trips up on excess whitespace.
[00:58] jfw: the block view perhaps needs to say how many transactions to provide a bound for the get-txn coordinates, or otherwise something to list the txns in a given block.
[01:00] jfw: I like the hash or coordinates to specify a txn, and yeah, would be nice to DWIM for any identifier.
[01:01] jfw: should be pretty easy too, 160bits = address, 256bits = txn or block - try both and see what works, then could optimize by noting that leading zeros probably means block.
[01:06] jfw: did you find how to plug python into apache without importing all sorts?
[16:10] whaack: jfw: thanks for trying it out and providing suggestions
[16:12] whaack: jfw: re python and apache, no, i sinned in order to get this online
Day changed to 2021-08-05
[20:13] whaack: dorion: package arrived at my local mail office, shipping price ended up being $80. I haven't had the chance yet to pick it up, will likely do so tomorrow
Day changed to 2021-08-06
[03:30] jfw: whaack: not too bad I suppose if it's arrived undamaged - good luck!
[03:31] jfw: in random glimpses of how far things have decayed in the mines, I am paid to quickly select tools and never invent any ... If I don't host a new company project on GitHub I'll eventually be fired and replaced with somebody who will.
[03:32] jfw: apparently apt-get installing a git daemon is considered inventing your own tools now or something?
[17:32] jfw: rereading the above quote a bit deeper, it strikes me as self-contradictory. First it's that she's paid for the valuable work of picking the right tool for the job from the modern deluge & correctly identifying when inventing new ones is actually a business necessity (there was that exception to the "never" in the elided part). But then it's that Github is a foregone conclusion, and not based on
[17:32] jfw: her own judgement but that of the higher-ups. But exam-taking can't possibly be worth paying salary for - the answer key is cheaper.
[17:35] jfw: She's presumably paid for something, but perhaps has not correctly identified what. And not knowing any more about the situation I'd be ill equipped to guess.
Day changed to 2021-08-07
[01:08] whaack: package received, looks good on first pass! gonna do a more thorough inspection in the next 3 days
[11:37] dorion: whaack, sounds good. 3 usb-ttl cables are also on the way (you may already have some from fg usage). I'll also gpg you more details about the install.
[14:16] whaack: dorion: ah when did you send them?
[17:28] dorion: whaack, tracking number in pm
[17:31] dorion waves farewell from vermont, about to hop on a train and be in Panama tomorrow morning.
[17:39] whaack: dorion: which company did you ship with? tracking number did not show up in ups or fedex's site.
[18:07] dorion: ups, correction in pm.
[18:09] jfw: possibly "full tracking URL" is the proper way to do this, though it wouldn't surprise me if most of the tracking services use enough java technology as to make working URLs an impossibility.
[18:17] whaack: new one works, thx
[18:18] whaack: from my experience typing a tracking id in goolag search engine will find you the link one needs
Day changed to 2021-08-08
[19:51] dorion waves from el istmo de panama
[22:17] whaack: dorion: congrats!!
Day changed to 2021-08-09
[02:04] jfw: dorion: splendid. we're a multinational again!
[02:05] jfw: perhaps for the first time, even.
[19:17] whaack: jfw: atm i plan to have my block explorer return the data in some "awk friendly" format, it will look something like what i currently have with a few minor adjustments. then i plan to release a corresponding client python script that grabs the data and organizes it. i have a sense that i may be doing 'the wrong thing' by not using a 'standard' data format such as json, xml, or sexprs, but otoh
[19:17] whaack: with this method i avoid writing extra code for handling different types of requests (human readable vs api mode) and the overall LoC of my program (when i include imports) likely goes down dramatically for not requiring a tool such as json
[19:55] jfw: whaack: ASCII with delimiters seems pretty standard to me... like I said though it depends on the data, or more specifically how complex it is - character set, list or table vs tree structure
[19:55] sourcerer: 2021-05-24 21:44:50 (#jwrd) jfw: whaack: depends on the data, if it's just a table of strings with controlled range of characters it may be simplest to just pick a delimiter character.
[19:56] jfw: personally I would not use XML for anything, given the choice (and possibly even otherwise)
[19:59] jfw: if you want a friendly web interface you're going to want to html-wrap the data in one way or another, so separate marked up and raw views is perhaps unavoidable.
[19:59] jfw: what sort of organizing would that client script be required for?
[20:05] jfw: fwiw, printing a tree as JSON isn't that difficult, and even parsing isn't too bad especially if you ignore unicode.
[20:07] jfw: but it seems overkill to me for this application; the most tree-like thing is the transaction listings, but even that is a uniform structure only a couple levels deep, it's basically just making explicit the grouping of input/output records in the table.
[20:09] whaack: alright, i'll go without using json etc. then, and just have a public spec for how the data is outputed with an example client that reads it
[20:10] jfw: (if you want an example of a widely used standard that works just about this way: HTTP headers!)
[20:17] whaack: jfw: heh, i just looked it up, and looks like my format is 90% of the way there
[20:39] jfw: to be clear, I don't mean it as something to aim for - it has some excess baggage like wrapped lines and parenthesis matching, iirc. it's just an example of simple text formats working fine and being widely implemented.
Day changed to 2021-08-24
[03:21] jfw: On a fresh reread of djb's /package scheme, parts of it still appeal to me, like using the filesystem to provide a package-to-file mapping, and having stable install paths no matter the distribution and status as "user" or
[03:21] jfw: "local-user" or "system" or whatever arbitrary distinction.
[03:23] jfw: the most dubious part is the "globally allocated" namespace, without the republican machinery eg strong identity, punishment gazette for dealing with such Gavins as would defile the namespace.
[03:25] jfw: funny he illustrates with some relatively minor instances of namespace collision inconveniences but missed the kind that matters most, i.e. the presumed maintainer going rogue.
[03:28] jfw: his own registry, based on metadata from my archives, was last updated 2012 at the latest - not counting this March where he substituted his own links with https.
[03:34] jfw: it's got 80 entries pointing to 11 domains, of which 16 never got a link, and a good portion from a small sample of the links are turning up dead.
[03:35] jfw: it's all supposedly "discussed on the mailing list" for which he apparently never saw any need to publish archives.
[03:38] jfw: he never updated his own pre-daemontools works, i.e. most of the good stuff, to use the scheme.
[03:44] jfw: (that was back on /package rather than the namespacing per se.) then the docs are woefully lacking in the "okay so how do I implement this for my shiny new program?" department, various details left unspecified, at best implicit by trying to emulate his examples, and unenforced, such as the internal layout, how to install, and metadata in the <package-name>/package/ subdir.
[03:53] jfw: In this light, I figure I'll stick with parts of it but continue happily not registering my names with the authorities, and /command has to go. Yet another place to search for binaries and find perpetually missing from $PATH, that ~nothing else uses? /bin and /usr/bin would be way costlier to snip due to all the #! scripts at the very least; /usr/local can perfectly well be ignored, not like FHS
[03:53] jfw: fanatic distributors can stop you using "the system's" namespace when you're distributing directly. There's some argument to avoiding /bin for systems where it's distinct from /usr/bin, eg readonly root filesystem, recovery environment with basic tools only or whatnot.
[03:56] jfw: They're not distinct on Gales, as I don't find those things convincing; but people have in the past so such systems exist.
[03:59] jfw: So my approach will be: tarball extracts (or tree presses or whatever) to /package/<packagename>-<version> as I've used before, compiles in place, and executables get symlinked in /usr/bin.
[04:00] jfw: Undecided whether to stick with README and "install" script in the "package" subdir or move them to top-level but leaning toward the latter.
[04:03] jfw: (huh, now that coulda made a fine mini-blog article, and here I thought I had just a couple-line update to express)
[04:10] jfw: and I suppose patching daemontools on Gales to switch from /command would be in order too.
[15:41] dorion: http://fixpoint.welshcomputing.com/2021/jwrd-logs-for-Aug-2021/#2869 -- makes sense to me to get rid of /command and just go with /usr/bin
[15:41] sourcerer: 2021-08-24 03:53:02 (#jwrd) jfw: In this light, I figure I'll stick with parts of it but continue happily not registering my names with the authorities, and /command has to go. Yet another place to search for binaries and find perpetually missing from $PATH, that ~nothing else uses? /bin and /usr/bin would be way costlier to snip due to all the #! scripts at the very least; /usr/local can perfectly well be ignored, not like FHS
[15:42] dorion: http://fixpoint.welshcomputing.com/2021/jwrd-logs-for-Aug-2021/#2874 -- it still can become a blog article, probably good to turn this thread into one to make it easier to reference. nothing wrong with first draft in log and final draft as article.
[15:42] sourcerer: 2021-08-24 04:03:55 (#jwrd) jfw: (huh, now that coulda made a fine mini-blog article, and here I thought I had just a couple-line update to express)
[15:44] dorion: http://fixpoint.welshcomputing.com/2021/jwrd-logs-for-Aug-2021/#2867 -- calls to mind the lack of response to fg/phuctor emails alf reported, iirc.
[15:44] sourcerer: 2021-08-24 03:38:17 (#jwrd) jfw: he never updated his own pre-daemontools works, i.e. most of the good stuff, to use the scheme.
[21:06] jfw: there was the notion of the old vs. new & "improved" djb; though seeing as how I don't quite buy into demonic possession I'm tending to think the old wasn't as solid as it might have appeared.
[21:17] jfw: technically brilliant but politically inept perhaps, got grabbed by the dark side, not because the masses have any use for people like him but so as to keep him safely busy and out of further trouble.
[21:21] jfw: maybe-2003 djb fighting for the unknowing & uncaring masses
[23:03] dorion: in better news : http://ossasepia.com/2021/08/23/its-called-vamp/
Day changed to 2021-08-30
[19:34] jfw: Before the month's log closes, I'd like to point out since no one else did publicly that I may have dunning-krugered it a bit with http://fixpoint.welshcomputing.com/2021/jwrd-logs-for-Aug-2021/#2806 , in the sense of concluding much of anything at all based on such (expectedly) flimsy science reporting.
[19:34] sourcerer: 2021-08-01 03:01:54 (#jwrd) jfw: some further detail on the recent lulz worth highlighting for future reference: https://www.reuters.com/world/us/majority-covid-19-cases-large-public-events-were-among-vaccinated-us-cdc-study-2021-07-30/ ( https://archive.is/Oe0ci )
[19:39] jfw: it was an attempt to at least realize some kind of value from the time spend reading the stuff; but perhaps the chief lesson is more along the lines of, don't buy into what any statistics seem to be saying if you can't bring actual statistical expertise to bear, given how many ways there are to go wrong.
[19:42] jfw: also "gene therapy" isn't quite an accurate label, as the mRNA vaccines don't reportedly work by altering cellular DNA, more just altering their function to produce the viral spike protein - presumably only temporarily while the mRNA supply lasts.
[19:42] jfw: *altering cellular function
[19:44] jfw: AFAIK it still stands that it's a novel type of treatment, markedly different from prior notions of what vaccines are, and while in development for some time, had not been previously tested at population scale.