Show Idle (>14 d.) Chans


← 2020-04-17 | 2020-04-19 →
03:27 diana_coman jfw: heh, so now you know the length of your maybe half hours!
~ 23 minutes ~
03:51 diana_coman !!up lru
03:51 deedbot You may not !!up yourself.
03:51 diana_coman !!up #ossasepia lru
03:51 deedbot lru voiced for 30 minutes.
03:52 diana_coman hi lru, it's enough to ask for voice - and then use it reasonably, that's about it all.
03:52 lru that's fair enough
03:52 diana_coman what brings you here?
03:53 lru I've been trying to understand what this channel is about... so far seeing lots of talk about commitments and reviews :-)
03:53 lru general bitcoin interest, and the fact that the trilema channel closed, and this one appeared to be something of a continuation
03:54 lru oh... I'm also curious if the bots you folks use is open source or not
03:54 lru for irc
03:55 diana_coman it's about learning, working efficiently and effectively with others and generally building up what you do so it grows day by day, sustainably, rather than going nowhere in a year or less.
03:55 diana_coman lru: do you have a site/blog/write anywhere?
03:57 lru I used to, but not anymore... I was writing more opinions than a log of projects, and it didn't serve the use I originally wanted, so took it down
03:57 diana_coman this chan started in fact before trilema closed, it was part of tmsr; it's not a continuation of #trilema itself, no; there's the article linked from the topic giving some more details and otherwise lots to read around
03:57 diana_coman lru: why take it down anyway? (and in the process break the links of anyone and everyone who linked to your writing, for that matter)
04:00 diana_coman lru: you know, in general, ~nothing serves "the use I originally wanted" - mainly because you can't really tell upfront like that *what* something will end up useful for; but that's a feature, not a bug!
04:01 lru normally I'd agree, but this was about religious and political topics, and leaving it there would have done more harm than good, I believe
04:02 diana_coman lru: how/why?
04:03 lru it would have been the equivalent of someone posting a howto of how to use grep with examples that didn't work :-)
04:03 lru better to just read the primary material
04:04 diana_coman lru: so you add a big-red warning upfront "meanwhile I realised this is bullshit" or similar , sure; but leave it there, there's no harm to it, even better to have it in fact as reminder that yeah, usually that's how things go - not that straightforward as it seems otherwise.
04:05 diana_coman so link to the primary material and send readers there, what
04:06 diana_coman the thing is this - whether it's still there or not, you still wrote it at that time and it's still part of your history; just "taking it down" doesn't win for you anything but it does come at a cost - all those broken links + no visible history at all + "that's the guy who vanishes" etc
04:06 diana_coman can even keep it as "archive" or whatever, but no win from not keeping it, that's about it all.
04:07 diana_coman lru: anyways, what are you working on otherwise, if you don't mind saying?
04:08 lru true.. and I think it's still stored somewhere on a disk, just not live... it was not a decision I took lightly at the time... probably 9 years or so ago now
04:08 lru systems admin and web service development at the moment... personally, my pet project is making the timewarrior tool work for me
04:09 diana_coman lru: what's the timewarrior tool?
04:09 lru it's a command line time tracking software
04:09 lru it was 90% of what I needed, just had to polish the rest
04:10 diana_coman lru: how big is it and what does it do beyond "keep track time per task" ?
04:11 lru not very big, only found it a few days ago... it's one of those unix-like 'do one thing well' tools... it makes tracking and editing fairly easy, focuses on categories first and descriptions second (which is a paradigm shift from other tools), shows charts by day/week/month, totals, etc, and stores the raw data in text files
04:12 lru written in C++, python for extensions (or anything really, it uses stdin/stdout)
04:13 diana_coman ahahah, just when I was typing that it sounds potentially interesting and/or at least worth writing up a review of it and publishing on your fresh blog, you drop the c++ & python, heh
04:13 diana_coman next I'll hear that "not very big" means it's ~only 100k loc!
04:13 lru is C++ and python verboten around here? :-)
04:13 diana_coman not a million yet
04:14 diana_coman lru: nothing is verboten around here
04:14 lru yep, not a million lines yet :-)
04:14 diana_coman it's just inefficient and not something I want to import anywhere for the basic task of keeping track of time on tasks, ffs; awk and bash are more than enough for that.
04:15 diana_coman anyways, if it won you as a champion for it, still nothing wrong with writing it up and publishing that; but that's the thing - do speak up publicly for the things (and people!) you like, what.
04:15 lru yep... I figured it was worth finding what already existed instead of starting from scratch, and was happy to find it
04:16 diana_coman well, there exist a LOT of things, that's the modern trouble; it's not "it doesn't exist" but "there's a pile to sort through and a life to sink into figuring out which of them is least broken"
04:16 lru I did notice a penchant for publishing when I was reading the channel over the last few days :-) my default mode is quiet... is there a philosophical reason for writing?
04:18 lru oh yeah... "which one is least broken" is indeed a task....I limited my search to things that were already installable from the Debian repos... still had to junk a lot
04:18 diana_coman so it's not even enough to say "finding what already existing" really; it's more "finding what is already used and maintained by people I want to work with, long term" - because you'll need to work with others if it's to add up to anything really
04:20 diana_coman lru: yes, there's lots of reasons for writing and of all sorts really; it starts with the fact that writing IS thinking - it forces you to structure better and it quickly shows where/what you don't master that well, in a word it's the most basic tool of intellectual life anyway
04:20 diana_coman and in addition to that, it allows others - *useful* others - to interact with you and *help*
04:21 diana_coman !!up #ossasepia lru
04:21 deedbot lru voiced for 30 minutes.
04:25 lru true, part of a bigger picture of cooperation and potentially business too
04:26 lru is your irc bot open source btw?
04:26 diana_coman lru: sure, the alternative - and even the *most usual and most common* alternative - is the ~autistic "will work on my own, no need to talk and waste all that time" and/or with added various flavours of "won't change, no matter what" ; it never really gets one anywhere, there's plenty of examples to look at, but it can *feel* very virtuous or whatever else, sure.
04:26 lru or perhaps based on one that is?
04:27 diana_coman lru: which one?
04:27 lru the one that does the logging
04:27 lru and the web frontend
04:27 diana_coman ah, there is the code published, sure; though I want to ditch that shit, ffs
04:27 diana_coman yeah, see on my Reference Code Shelf from my blog
04:27 lru thanks, I'll take a peek
04:28 diana_coman lru: do you know about V for version control?
04:28 diana_coman because hm, "open source" is a rather horribly failed experiment as such
04:28 lru no, only had experience with cvs, rcs, svn, git, and a microsoft one from the past
04:28 lru source safe I think it was called
04:29 diana_coman lru: here's an introduction/overview : http://ossasepia.com/2019/12/13/a-walk-among-the-trees-of-v/
04:29 lru thanks
04:29 diana_coman fwiw I did use cvs, svn, git , even a few others and gah, I hope I never have to have one of those on a computer of mine, lol
04:30 lru I've pretty much settled on git these days... never heard of V before
04:31 diana_coman lru: you'll need to at least run a V for those vpatches of the logbot; there's a starter pack http://ossasepia.com/2019/11/10/v-tree-and-v-starter-v2/
04:31 diana_coman lru: for how long have you been around #trilema?
04:31 diana_coman did you read it at all?
04:32 diana_coman and yeah, of the bunch and before realising that I can ditch it, it was git I had uncomfortably settled with, indeed
04:32 lru I joined the channel a few times, but didn't keep up with it... it was interesting in a sense of "half the conversation seems to be missing" most of the time :-) I mostly read the blog,and read the logs if they were mentioned in a blog post
04:33 diana_coman lru: re my "code control with V",here's the article: http://ossasepia.com/2020/02/05/the-v-tree-nursery-or-code-control-with-v/
04:33 diana_coman lru: kind of weird you didn't pick up on V then anyway, as it was quite central, hm (and yes, introduced on trilema.com)
04:34 diana_coman anyway, you've got already quite the reading list there
04:34 lru maybe I read "V" and had no idea what it referred to :-)
04:34 lru who wrote V?
04:35 diana_coman re logbot though, I'll underlign that the current thing uses python and moreover it imports the nightmare called "flask" ; it forces also postgres and it ensures that you are locked in with it because yeah, if you change, you'll break everyone's links
04:35 lru I found most people don't like git until they read and understand the book "Git from the bottom up"
04:35 lru but that's newbies
04:35 diana_coman so no, I don't like it and I am still using it atm as I didn't get around to sink some time to excavate myself out of its shit
04:36 diana_coman will do it too because it's becoming unbearable but myeah, not yet done
04:36 * lru chuckles "nightmare called flask"
04:37 lru isn't flask based on sqlalchemy under the hood? that's supposed to abstract the DB
04:37 diana_coman lru: I read the git book years ago, and I liked git better than the rest of monsters; but as that v-tree nursery article says in the very start, it's just bringing in on my systems a lot of things I have no need for, ie it's asking for too much from me for what it does
04:38 diana_coman lru: here, I have an article for that question too: http://ossasepia.com/2019/09/02/ossabot-and-its-flask-of-python-27-on-centos-6/ (and see? that's why it's best to publish - don't need to re-tell the same thing every time one asks)
04:38 lru :-)
04:38 lru I'm loading up the tabs here
04:38 diana_coman see, it's enough to start talking for a bit and then... it explodes!!
04:38 diana_coman lolz
04:38 lru lol
04:39 diana_coman anyways, happy reading then; I'll try to get around to give you a rating too and then you should be able to !!up yourself
04:39 lru thanks very much!
04:40 diana_coman or hm, trinque - is this working now?
04:40 ossabot Logged on 2020-04-13 16:17:43 diana_coman: cruciform: try it in chan please
04:40 diana_coman lru: anyways, if it's not working/you don't have voice, just ask me in pm and I'll sort it out one way or another
04:41 lru no worries... as I said, my default mode is quiet, so not like I have urgent things to say anyway :-)
04:42 diana_coman better find them! heh; (but seriously, ONE good thing I learnt in all those years at uni was really this - to ask questions; HAD TO ASK questions, even! because yes, initially I sucked at this, totally)
04:43 lru true, especially when time is of the essence
04:43 diana_coman lru: you'll need to register a key with deedbot before I can rate you though, lol
04:44 lru does deedbot have a manual?
04:44 diana_coman !!help
04:44 deedbot http://deedbot.org/help.html
04:44 lru thanks
04:45 * lru wonders idly if 'deed' in deedbot enjoys both meanings
04:45 diana_coman trinque: btw, deedbot isn't answering to my pm !!ledger commands (sent one on the 16th and still not answered; one 2 minutes ago and still not answered)
04:45 diana_coman lru: there's reading on that too! :D
04:46 diana_coman (won't add it now to the pile, you can find it anyway)
04:46 lru the web of trust document? or something else?
04:46 lru I did read MP's web of trust blog page
04:46 diana_coman more really, since it's (just like V) part of a wider whole, of course
04:46 lru ahh ok
04:47 diana_coman anyways, I'll be away from keyboard for a while, laters!
04:47 lru cheers
~ 3 hours 52 minutes ~
08:39 feedbot http://trilema.com/2020/the-nude-wedding/ << Trilema -- The Nude Wedding
~ 3 hours 31 minutes ~
12:10 trinque hm interesting, will look diana_coman
12:13 trinque huh! the command's gone. I wonder if I deployed an old wallet service.
12:13 trinque should be easy to fix, will let ya know.
12:26 trinque yep confirmed. the command's there now, but I need to fix something with the number formatting, currently being displayed as rationals, which is quite a pain!
12:27 trinque will get that decimalized this afternoon. I recall I encountered this in the past.
12:27 trinque in other news, jfw and dorion, I've been running some experiments locally on minimal-footprint gcc bootstrap.
12:28 trinque it began as a refactoring of the musl-gcc scripts ave1 found and modified.
12:28 trinque I'm working with gcc-4.7.4 per your indications that it's the last to bootstrap with C compiler alone.
12:32 trinque so far I've only got binutils, busybox, gcc, linux kernel, make, mawk, and musl as dependencies.
12:33 trinque rather, + mpc, mpfr, and gmp, which I've merged into the gcc src.
12:33 trinque and I understand that the same can be done with binutils, which seems entirely appropriate.
12:34 trinque I had the whole thing building (and self-building) last week on gcc-4.9.4, but pivoted to gcc-4.7.4
12:35 trinque once I get the 4.7.4 build sorted (there's some include-path fuckery that appears to be a known issue with 4.7.4) this should be an exemplar of what I was driving at with the OS series so far.
12:36 trinque from there, I have some notions on how to allow bigger dependency chains to be brought in (to build eulora, for example) without having them engulf the system, and suddenly "oh you need clang on every box because mesa"
12:36 trinque at any rate, haven't forgotten about any of this, been working on all of it, but I wanted to prove out some of my claims before going forward.
~ 2 hours 29 minutes ~
15:06 diana_coman trinque: good to hear from you; I'll give another go to deedbot's ledger later today then; is self-voicing working now for those I rated 1 or more?
15:19 jfw trinque: well, I did get a gcc-4.7.4/musl build pretty well documented, including those dependencies (minus mawk, busybox awk seems to have worked fine) and a couple patches, and not using those musl-cross magic scripts
15:26 jfw speaking of awk though, I've been finding it pretty obnoxious trying to stick to the standard (or let's say busybox-compatible, lacking a formal standard) and avoiding gnuism
15:26 jfw bb awk: 3k lines; gawk: 80k
15:30 diana_coman jfw: myeah, there is that re awk & gawk (on both line count and pain if trying to stick to standard; there are also some things that pretty much require gawk; I still prefer it to python & co)
15:31 jfw but, how can it claim to be a string-oriented language while not supporting NUL separators, on unix where NUL is the _only_ reliable separator for file paths?! I'm supposed to what, base64 them? but awk doesn't understand that either. Or even hex/octal numbers.
15:32 jfw diana_coman: yeah I could see it as gawk compared to python
15:37 jfw I imagine gawk must implement garbage collection, since it allows e.g. arrays containing arrays. But, well, my Scheme does that, and sockets, subprocesses, bignums... all under 10k lines, so what's their excuse!
15:38 diana_coman read and find out! lol
15:41 jfw diana_coman: I suppose I'll have to; a linux needs an awk after all and seems it's not a clear decision yet on which
15:45 diana_coman as usual, it's a matter of what you are fine to do without, really
15:45 diana_coman sockets are in gawk only though afaik
15:46 diana_coman ie directly, as I used for sonofawitch
15:46 diana_coman that being said, I don't need that on eg. an offline machine
15:46 diana_coman (though I would still want awk there, lol)
15:48 jfw there's inetd style tools to pass a client socket down to a child on particular file descriptors... ofc awk can't read/write arbitrary FDs either
15:51 diana_coman well, that's the main addition there, indeed; read/write from/to arbitrary fds
15:53 jfw well I suppose it can if you count print | "cat >&3" and such
15:58 jfw http://logs.ossasepia.com/log/ossasepia/2020-04-18#1024485 - indeed.
15:58 ossabot Logged on 2020-04-18 04:48:53 diana_coman: jfw: heh, so now you know the length of your maybe half hours!
~ 18 minutes ~
16:17 jfw diana_coman: http://logs.ossasepia.com/log/ossasepia/2020-04-18#1024565 - there something wrong with postgres, as DBs go? and why would it be hard to change the web view without breaking links?
16:17 ossabot Logged on 2020-04-18 05:56:19 diana_coman: re logbot though, I'll underlign that the current thing uses python and moreover it imports the nightmare called "flask" ; it forces also postgres and it ensures that you are locked in with it because yeah, if you change, you'll break everyone's links
16:19 diana_coman jfw: it's not wrong per se, no; but I resent being *tied to it for no good reason really* (other than "so it happened"); and given that otherwise publishing means mpwp which uses currently mysql, I see no reason to add postgres to the pile; fwiw I used both postgres and mysql otherwise, I know of die-hard supporters for both, I still don't care that much to go to war for either.
16:20 jfw I recall full-text search mentioned as an advantage of postgres for the application
16:21 jfw though wordpress normally has a search feature so dunno how substantial that is.
16:22 diana_coman jfw: the search thing is always a contention for this to start with ie on one hand if you want to get the raw data can get it and do whatever search you want and on the other hand if it's published, I much rather it was integrated with the blog otherwise and not require me to maintain 2 things; and yes, exactly re wordpress; really, I can't see a reason for it being *yet another thing*
~ 30 minutes ~
16:53 trinque jfw: interesting, I will have to revisit what brought mawk in. I'd thought gcc demanded it. did you perhaps alias mawk -> awk?
16:54 trinque diana_coman: sitting down to another round of deedbot tweaking now, will make sure the voice params are also changed.
16:56 jfw trinque: POSIX awk should suffice per https://gcc.gnu.org/install/prerequisites.html
16:56 diana_coman trinque: cool, thank!
16:56 diana_coman trinque: cool, thanks!
16:56 jfw (and no I haven't made such an alias)
16:56 trinque jfw: yep, loudly didn't though in my case, so I'll figure out why.
16:56 jfw autoconf most likely :(
16:59 trinque diana_coman: to confirm, your rating of 1, or your L1's rating of 2 (or more) == voice?
17:04 jfw lru: http://logs.ossasepia.com/log/ossasepia/2020-04-18#1024571 - no, in fact pretty much Flask's only claim to being "micro" is that it doesn't mandate an ORM like, say, Django.
17:04 ossabot Logged on 2020-04-18 05:58:22 lru: isn't flask based on sqlalchemy under the hood? that's supposed to abstract the DB
17:05 diana_coman trinque: what I wanted was this: anyone with positive ratings (>=1) from me to be able to self-voice; those with rating >2 from me to act as "lordship" or whatever ie *their* l1 to be able to self-voice too; if this is too much/complicated, I'm open to any proposal really - my idea was to give people an easy way to self-voice but not necessarily get together with any newcomer all their newcomers too; otoh can just get all and if they ...
17:05 diana_coman ... bring in idiots, will negrate and it will still work I guess.
17:06 jfw lru: I've used sqlalchemy but don't care for it these days. You think it gets you out of database-picking politics, but this comes at the cost of quite a volume of code (and not even pure python!)
~ 1 hours 29 minutes ~
18:36 diana_coman !s hi
18:36 sonofawitch Hello there, diana_coman
18:42 diana_coman !o uptime
18:42 ossabot diana_coman: time since my last reconnect : 15d 10h 21m
18:51 diana_coman !o help
18:51 ossabot diana_coman: my valid commands are: src, uptime, seen-anywhere, help, s, version, seen
18:54 diana_coman !o version
18:54 ossabot I am bot version 597858.
~ 2 hours 1 minutes ~
20:55 trinque diana_coman: ledger fixed, found the patch I was missing.
20:56 trinque re: your request on voice thresholds, makes sense. I need to simplify the involved code a bit, will circle back tomorrow and oughta have it working as described sometime then.
← 2020-04-17 | 2020-04-19 →