#ossasepia | 2020-04-20

http://younghands.club/2020/04/20/jfw-plan-week-of-20-apr-2020/ << Young Hands Club -- JFW plan, week of 20 Apr 2020

~ 39 minutes ~

03:09

http://ossasepia.com/2020/04/20/ossasepia-logs-for-20-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 20 Jul 2019

03:20

http://ossasepia.com/2020/04/20/ossasepia-logs-for-21-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 21 Jul 2019

03:29

http://ossasepia.com/2020/04/20/ossasepia-logs-for-22-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 22 Jul 2019

03:41

http://ossasepia.com/2020/04/20/ossasepia-logs-for-23-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 23 Jul 2019

03:49

http://ossasepia.com/2020/04/20/ossasepia-logs-for-24-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 24 Jul 2019

04:00

http://ossasepia.com/2020/04/20/ossasepia-logs-for-25-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 25 Jul 2019

04:09

http://ossasepia.com/2020/04/20/ossasepia-logs-for-26-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 26 Jul 2019

04:20

http://ossasepia.com/2020/04/20/ossasepia-logs-for-27-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 27 Jul 2019

04:30

http://ossasepia.com/2020/04/20/ossasepia-logs-for-28-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 28 Jul 2019

04:41

http://ossasepia.com/2020/04/20/ossasepia-logs-for-29-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 29 Jul 2019

04:50

http://ossasepia.com/2020/04/20/ossasepia-logs-for-30-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 30 Jul 2019

05:01

http://ossasepia.com/2020/04/20/ossasepia-logs-for-31-Jul-2019/ << Ossa Sepia -- #ossasepia Logs for 31 Jul 2019

05:10

http://ossasepia.com/2020/04/20/ossasepia-logs-for-01-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 01 Sep 2019

05:19

http://ossasepia.com/2020/04/20/ossasepia-logs-for-02-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 02 Sep 2019

05:30

http://ossasepia.com/2020/04/20/ossasepia-logs-for-03-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 03 Sep 2019

05:40

http://ossasepia.com/2020/04/20/ossasepia-logs-for-04-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 04 Sep 2019

05:51

http://ossasepia.com/2020/04/20/ossasepia-logs-for-05-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 05 Sep 2019

06:00

http://ossasepia.com/2020/04/20/ossasepia-logs-for-06-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 06 Sep 2019

06:11

http://ossasepia.com/2020/04/20/ossasepia-logs-for-07-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 07 Sep 2019

06:20

http://ossasepia.com/2020/04/20/ossasepia-logs-for-08-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 08 Sep 2019

06:31

http://ossasepia.com/2020/04/20/ossasepia-logs-for-09-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 09 Sep 2019

06:41

http://ossasepia.com/2020/04/20/ossasepia-logs-for-10-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 10 Sep 2019

06:48

http://thetarpit.org/2020/on-the-hellpits-of-abstraction << The Tar Pit -- On the hellpits of abstraction

06:50

http://thetarpit.org/2020/work-plan-for-m2-2020 << The Tar Pit -- Work plan for M2 2020

06:52

http://ossasepia.com/2020/04/20/ossasepia-logs-for-11-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 11 Sep 2019

06:53

http://thetarpit.org/2020/on-the-hellpits-of-abstraction << The Tar Pit -- On the hellpits of abstraction

07:01

http://ossasepia.com/2020/04/20/ossasepia-logs-for-12-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 12 Sep 2019

07:07

http://thetarpit.org/2020/again-on-general-purpose-tools << The Tar Pit -- Again on "general-purpose" tools

07:12

http://ossasepia.com/2020/04/20/ossasepia-logs-for-13-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 13 Sep 2019

07:21

http://ossasepia.com/2020/04/20/ossasepia-logs-for-14-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 14 Sep 2019

07:31

http://ossasepia.com/2020/04/20/ossasepia-logs-for-15-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 15 Sep 2019

07:43

http://ossasepia.com/2020/04/20/ossasepia-logs-for-16-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 16 Sep 2019

07:52

http://ossasepia.com/2020/04/20/ossasepia-logs-for-17-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 17 Sep 2019

08:03

http://ossasepia.com/2020/04/20/ossasepia-logs-for-18-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 18 Sep 2019

08:12

http://ossasepia.com/2020/04/20/ossasepia-logs-for-19-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 19 Sep 2019

08:21

http://ossasepia.com/2020/04/20/ossasepia-logs-for-20-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 20 Sep 2019

08:33

http://ossasepia.com/2020/04/20/ossasepia-logs-for-21-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 21 Sep 2019

08:43

http://ossasepia.com/2020/04/20/ossasepia-logs-for-22-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 22 Sep 2019

08:54

http://ossasepia.com/2020/04/20/ossasepia-logs-for-23-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 23 Sep 2019

09:04

http://ossasepia.com/2020/04/20/ossasepia-logs-for-24-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 24 Sep 2019

09:13

http://ossasepia.com/2020/04/20/ossasepia-logs-for-25-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 25 Sep 2019

09:15

http://trilema.com/2020/the-puppyphant-and-other-stories/ << Trilema -- The puppyphant and other stories

09:24

http://ossasepia.com/2020/04/20/ossasepia-logs-for-26-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 26 Sep 2019

09:33

http://ossasepia.com/2020/04/20/ossasepia-logs-for-27-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 27 Sep 2019

09:42

http://ossasepia.com/2020/04/20/ossasepia-logs-for-29-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 29 Sep 2019

09:54

http://ossasepia.com/2020/04/20/ossasepia-logs-for-30-Sep-2019/ << Ossa Sepia -- #ossasepia Logs for 30 Sep 2019

10:03

http://ossasepia.com/2020/04/20/ossasepia-logs-for-01-Oct-2019/ << Ossa Sepia -- #ossasepia Logs for 01 Oct 2019

10:15

http://ossasepia.com/2020/04/20/ossasepia-logs-for-02-Oct-2019/ << Ossa Sepia -- #ossasepia Logs for 02 Oct 2019

10:24

http://ossasepia.com/2020/04/20/ossasepia-logs-for-03-Oct-2019/ << Ossa Sepia -- #ossasepia Logs for 03 Oct 2019

10:33

http://ossasepia.com/2020/04/20/ossasepia-logs-for-04-Oct-2019/ << Ossa Sepia -- #ossasepia Logs for 04 Oct 2019

10:44

http://ossasepia.com/2020/04/20/ossasepia-logs-for-05-Oct-2019/ << Ossa Sepia -- #ossasepia Logs for 05 Oct 2019

10:55

http://ossasepia.com/2020/04/20/ossasepia-logs-for-06-Oct-2019/ << Ossa Sepia -- #ossasepia Logs for 06 Oct 2019

11:04

http://ossasepia.com/2020/04/20/ossasepia-logs-for-07-Oct-2019/ << Ossa Sepia -- #ossasepia Logs for 07 Oct 2019

~ 3 hours 58 minutes ~

15:02

I devoiced feedbot in here for now, so the logs-flood doesn't get in the way of discussion.

15:06

(says /me the ever optimist re discussion.)

15:09

diana_coman, thanks for the feedback on http://younghands.club/2020/03/22/dg-week-1-plan-mar-23-29-2020/#comment-648 ! I tried replying with html formatting, but it seems one can't do that in comments?

15:12

cruciform: hm, html formatting does (should!) work in comments - that's how I did mine for instance (using <blockquotes>); what html tags did you try to use?

15:12

and you're welcome; looking forward to those articles promised for tomorrow evening, too!

15:13

correction: the tag I uses is <blockquote>

15:13

used* ; gah.

15:14

dorion: you around and up for jwrd priorities discussion?

15:16

I was doing it via an online editor; <p style="padding-left: 30px;" align="left"> in an attempt to indent quotes - this works for articles posted on YH, but not in comments

15:18

ah, you can use tags but not the full set of parameters and then anyway - how did you come with the 30px exactly? lolz; there's a reduced set of tags that will work inside comments.

15:18

dorion: and one question coming to mind from while you were out was revisiting wallet or payments as a service. Could also include sweeping / recovery from various formats.

15:19

diana_coman, aha! the 30px is just the default indent the online editor used via their GUI (https://html-online.com/editor/)

15:21

cruciform: sounds more like a compiler than an editor really - emits unreadable garbage you'd never type out on your own

15:22

cruciform: aham, "default" whatevers will rarely be what you actually need for ...your specific situation; kind of the thing with "defaults" like with "average person" - it is an average but one that fits no actual, real person; anyways, there's the blockquote tag that does what you need there, so...use it;

15:22

jfw: maybe that's the point! I can see the appeal there - why have it emit what you could type yourself? Let it emit original works! ART!

15:23

jfw, diana_coman thanks! I guess I... "just wanted to"TM have something that "worked" without having to do any work!

15:24

aha; it works ..to make more work!

15:25

lol

15:25

diana_coman: makes sense of why all gui html exports seem to look like this; otherwise they'd have much less supposed reason to exist

15:25

jfw, dorion, I've not forgotten about http://logs.ossasepia.com/log/ossasepia/2020-03-28#1023165 ; I’m still interested, but don't want to commit until I’ve actually managed a week or two of self-directed work

15:25

Logged on 2020-03-28 22:23:48 jfw: cruciform, your sample class outline: http://fixpoint.welshcomputing.com/wp-content/uploads/2020/03/jwrd-sample-01-initial-unix-commands.txt

15:26

jfw: exactly so.

15:27

cruciform: cool, and welcome to the lots of talk about commitments and reviews party!

15:27

Logged on 2020-04-18 05:14:16 lru: I've been trying to understand what this channel is about... so far seeing lots of talk about commitments and reviews :-)

15:27

cruciform: what are you keeping busy with otherwise nowadays? (can't *quite* believe you clean that flat to obsession all the remaining one-week-minus-half-day-time)

15:28

cruciform: we haven't forgot about a quote broken down by module either; I think it got swallowed up into a bigger sales article that's been dragging out.

15:28

jfw: ahaha, the talk's the party and the doing's the ...afterparty?

15:29

apparently sometimes the hangover.

15:29

diana_coman, mostly wasting time on anti-social media platforms, a la http://trilema.com/2019/no-platforms/ . I stopped ' em all last week, and didn't actually miss them at all!

15:29

lolz, I was trying to avoid that word.

15:29

jfw: ^

15:30

http://logs.ossasepia.com/log/ossasepia/2020-04-20#1024801 looking forward to seeing it!

15:30

Logged on 2020-04-20 16:49:54 jfw: cruciform: we haven't forgot about a quote broken down by module either; I think it got swallowed up into a bigger sales article that's been dragging out.

15:31

cruciform: it can be amazing how much free time one finds, once dropping various timewasters that are never missed otherwise, certainly.

15:32

diana_coman: heh, the word didn't like be avoided then

15:33

diana_coman, quite; and how obvious in retrospect it is that these timewasters aren't just zero-value, but -EV in and of themselves: antisocial media is pure noise

15:34

well, that's pretty much their whole model really - in exchange for farming *you*, they provide the easy way to *feel* like doing something; all day long; without work!

15:39

cruciform: btw, re following proofs - there's no "shortcut" as such, sure, but there's practice with it that can give "speed up"; no idea how used you are to that though because myeah, unless you took the "pure Maths" modules at A level, there's precious little to none proof work as such in UK schools as far as I'm aware.

15:41

diana_coman, I did A-level maths, and yea; can't remember doing any proof work. I expected a steep learning curve, and haven't been disappointed!

15:42

heh, enjoy it then!

15:45

"Enjoy" might be too strong a word at present, but I'll try!

~ 26 minutes ~

16:12

diana_coman: re bvt's vtools work, I got the idea that it had some shortcomings like lack of leafs command and a non-tree-like patch tree view, and would require further work from me either on GNAT builds or making v.sh do without. Whereas v.pl was supposed to be the known-working historical tool, and a relatively smaller investment there would be adequate for my present needs.

16:12

jfw: he fixed those issues/lacks

16:13

it does require gnat though, that's the price for using ada, true; and yes, in principle v.pl is the smaller investment route indeed

16:13

but yeah, it has its own quirks for sure.

16:14

diana_coman: good to know you consider them fixed now.

16:15

jfw: what are you using for v-pressing/work anyway?

16:16

(and yes re vtools; I have it on my list for this week to also mirror it all + publish my sigs for it, as they accumulated but I never set down to publish any of them.)

16:16

diana_coman: v.pl for pressing. For diffing, the ill-fated vdiff on top of GNU diff presently.

16:17

Which itself is not ideal for sure.

16:18

ah, I had the impression you found v.pl inadequate for what you needed, in the end.

16:18

eh, ~nothing's ever ideal, if it gets to that.

16:19

hm, no I was happy with v.pl after the work I did on it, at least until finding this exp-time issue

16:20

then it's just a matter of deciding whether it's something you can live with in exchange for not-having-to-get-gnat-too or not, pretty much; anyways, it doesn't sound like a priority-problem atm.

16:21

ah but I did discover another shortfall of busybox patch - it doesn't delete empty directories after deleting the last file with the -E flag, that is, a presser built on it can never delete directories. This nudges me further toward embracing GNAT so as to get phf's vpatch program. But yeah, not seeing the priority quite yet.

16:23

jfw: so give it some time then.

16:23

will do.

16:26

in other things, since I am looking again at the annoying issue with log lines and numbers and whatnots - my current thinking is to ditch that source of trouble entirely and simply use hashes for each line to identify it uniquely AND across whatever bots

↖

16:26

(trinque) 2020-04-02 diana_coman: re bots - when/if they are in sync, they have the exact same numbers for all lines so in principle any links can be ported easily from one to another; but it is a very fragile thing for sure and no, I don't like it at all either.

16:26

if anyone has a better idea, now's the time to speak up on it

16:26

diana_coman: does that mean one person can never repeat the same line?

16:27

the ordering is anyway a local matter for each bot and otherwise recoverable from any existing record if one cares enough.

16:27

jfw: at the same time too?

16:27

ah, you mean as repeat the content

16:27

hm

16:27

hm

16:28

lol, there is that, with the begging the question as to ..how are those lines different then?

16:28

diana_coman: and I recall you didn't want to include timestamps in that, indeed multiple bots will see them differently

16:28

ie why would you want to identify them differently when they *are* the same?

16:28

because they came at different times / replies to different context?

16:29

ah, you mean as same content that just happened across the whole log rather than as above

16:29

yes

16:29

*

diana_coman shall ponder this some more.

16:30

the obvious thing would be ofc to use context but hm.

16:30

so just saying "yes" like that would be user-error... lolz

16:31

eh, not user-error, no; in this context there is no user error possible really

16:31

if we reinvented irc so each client would send its own unique ID with the message, would be easy...

16:33

(though someone could temporarily squat a nick and repeat IDs to possibly cause trouble.)

16:33

jfw: well, even there, it can be the same client saying the same line with the same id at different times so in that sense still potential clash; what it's not yet clear to me is whether this sort of clash is in itself a problem though, hm.

16:35

basically either there's sync on some external counter/similar or on the content+context, not much else as option that I can see at all.

16:38

I haven't yet tackled the old gossipd discussions but maybe there's something of use in there?

16:38

the thing with external sync-token of whatever sort is exactly that it's ...external; an addition that needs justification and so far I'm not sure I can still see the full case for it.

16:39

I'll probably end up re-reading around that, yes.

16:42

hash does seem the reasonable line identifier. it'd be interesting to simply allow for multiple orderings of messages. if my bot had lines yours didn't, and mine communicated this to yours, there's no reason your bot couldn't present its ordering as "what I saw" and also have some way of displaying what others claim was seen.

16:43

trinque: the point jfw makes above is that there can be lines that repeat eg on different days, aka not just a matter of "which one made it in first"

16:43

sure, and this is no problem.

16:44

consider for a sec how git works. you have a hash db, and atop that a tree of references.

16:44

your "log" is a link list of references, and there's no problem with referencing the same hash in multiple places along the linked list

16:45

that's pretty much why I was saying earlier that I'm not that sure that it is a problem in itself; still, I don't really recall the reason /rationale why exactly was this linenumber mechanism pushed into current logbots.

16:45

Logged on 2020-04-20 17:55:11 diana_coman: jfw: well, even there, it can be the same client saying the same line with the same id at different times so in that sense still potential clash; what it's not yet clear to me is whether this sort of clash is in itself a problem though, hm.

16:46

I don't see a reason for it, other than a political THERE SHALL BE ONE TRUTH

16:46

which sure, there's one truth, but many denotations

16:46

Are we talking about strong or weak hashes btw?

16:46

jfw: what do you mean?

16:47

all hashes are voodoo anyway

16:47

do references end up like http://logsite/4d752030a579980bae0c2012f70227b6f9436fd3154b995219cde0656b92cade126828b089b43881bea81e737b44df3189e800c82b330f31634e02a206829287 ?

16:48

at that point I wonder what it gains over quoting the message in full.

16:48

one could both have that as a canonical reference, and also have relative reference mechanisms that are human-friendlier.

16:48

one never hand-types the smaller int identifiers either, ftr

16:48

trinque: I am trying to make sure it's not just my bias that makes me not see there some reason other than exactly that "ONE TRUTH AND NO DENOTATIONS ALLOWED TO EXIST EVEN IN PRINCIPLE"

16:50

jfw: your question there needs to include context to have some meaning; ie "strong" hashes exist; this doesn't mean they have to be used everywhere as such

16:51

it'd be pretty interesting if a hash worth considering had collisions that were also coherent natural language.

16:51

should be doable mad-libs style

16:52

trinque: well, atm "hash" means keccak as far as I'm concerned; not sure why would I introduce another one.

16:52

yep

16:53

alright, and keccak has infinite output stream so indeed lends itself to considering particular truncations for different uses

16:54

it is interesting to think of alternate addressing systems that could handle line insertions.

16:54

jfw: re links, that's no issue really because that is already presentation of the content so you can do it whichever way you want and as a separate thing from the logging itself anyway; eg the lines end up in blog articles anyway, so can simply have there counters per article, there's the whole context given anyway, not like one needs to drop the hash in the article itself; bot can recover the rest when/as needed; after all, that's ...

16:54

... one thing at which machinery is supposedly useful.

16:55

trinque: what do you mean re addressing systems?

16:55

"Fifth-and-a-half street"

16:58

what jfw alludes to, can have some amount of wiggle room per unit in your line enumeration, such that references are stable but there's room to insert

16:58

diana_coman: then the blog in question becomes that external provider of line identity?

16:58

diana_coman: thinking about it, imo it's interesting to quote a line in the context, not just the line itself; so the address should include a bit of context as well

16:59

bvt: question there is what exactly do you pick then as "context"

17:00

jfw: for some...context! there's no way you can somehow *mandate* that such thing is not possible, how would that be?

17:00

sure, you can choose whether you use that presentation or not but that's about it all

17:02

i.e. http://paste.deedbot.org/?id=rCkQ -> the address for line4 could be '48a7-8d32-57d248cff05137eb3a2eaf2352b838350186ae95-346c-d134', and when resolving the link, would attempt to maximize the number of matching context hashes, ignoring the order of non-central elements

17:02

trinque: if one simply uses truncated keccak hashes, in principle can insert anywhere and nothing changes anyway, the enumeration is separate entirely.

17:02

diana_coman: not following, was I mandating something?

17:03

jfw: when you say that the blog becomes that external provider of line identity; how does it become? it doesn't; it provides an enumeration and nothing more.

17:04

diana_coman: is that any different from what the separate bots do now through their sequence numbers?

17:04

I admit I'm not quite sure the problem to be solved here.

17:07

bvt: so you mean previous lines as seen by one logger

17:07

previous n lines

17:08

jfw: the core question I had was why does logger bot bother exactly with line numbers to start with

17:08

yes, but ignoring the order, and mandating, for example, at least 2 or 3 out of 4 context matches

17:09

note that this is at logging time, not at presentation time; hence the difference re blog is that well, blog does presentation not logging, to start with.

17:10

bvt: that's a pretty interesting idea.

17:10

bvt: that would be about the most straightforward definition of context of a line available, indeed; picking a number out of hat anyway re how many previous lines.

17:11

but I guess it can even be flexible really

17:11

ie can request as many or few as one wants at one time or another

17:13

yes, it straightforward; well, also could pick the context lines not consecutively, but pick e.g. 4 context hashes out of range of +-10 lines for the url

17:14

so overall it seems to me to tend towards store lines as they are seen locally; identify by mixed-hashes within the group; can even take lines after the one sought, etc.

17:14

but all of it goes back to - it's enough to have the hash for the line; can then build on top of that whatever strategy to identify uniquely a line with as much precision as one considers needed.

17:15

(and now I'm even more puzzled than before re why linenumbers currently, ugh)

17:25