#soylent | Logs for 2023-05-26
« return
[00:02:29] -!- kolie__ [kolie__!~kolie@kkpeqz-44-538-06-820.lsan.ca.frontiernet.net] has joined #soylent
[00:05:15] -!- kolie has quit [Ping timeout: 256 seconds]
[00:11:47] -!- kolie [kolie!~kolie@2620:13c:nprx::vr] has joined #soylent
[00:14:19] -!- kolie__ has quit [Ping timeout: 256 seconds]
[00:29:08] -!- kolie__ [kolie__!~kolie@2620:13c:nprx::vr] has joined #soylent
[00:31:53] -!- AzumaHazuki [AzumaHazuki!~hazuki@the.end.of.time] has joined #soylent
[00:32:27] -!- kolie has quit [Ping timeout: 256 seconds]
[04:56:23] -!- AzumaHazuki has quit [Remote host closed the connection]
[05:50:21] <mechanicjay> Jeez, I was just going through my old subs on the site -- I subb'ed a couple fun ones over the years! :) :*(
[11:40:10] -!- audioguy [audioguy!~audioguy@Soylent/Staff/Developer/audioguy] has joined #soylent
[11:40:10] -!- mode/#soylent [+v audioguy] by Imogen
[14:22:49] -!- AzumaHazuki [AzumaHazuki!~hazuki@the.end.of.time] has joined #soylent
[15:42:26] <chromas> https://www.youtube.com
[15:42:28] <systemd> ^ 03China's Largest Factory - Living Where You Work (Full Documentary)
[16:03:06] -!- halibut has quit [Quit: leaving]
[16:08:50] -!- halibut [halibut!~halibut@CanHazVHOST/halibut] has joined #soylent
[16:17:10] <Deucalion> https://logs.sylnt.us
[16:17:33] <Deucalion> It's in the topic of this very channel :P
[16:30:53] <chromas> https://stats.sylnt.us
[16:30:54] <systemd> ^ 03#soylent @ irc.soylentnews.org stats by Deucalion
[16:30:58] <chromas> that's there too :D
[16:31:47] <Deucalion> That's fairly meaningless - it covers a 3029 day period! :D :D
[16:33:48] <kolie_> https://soylentnews.org
[16:33:49] <systemd> ^ 03Journal of NCommander (2)
[16:45:04] <mechanicjay> ....a 7K page document. Why use 3 words, when 30 will do!
[16:45:17] <mechanicjay> s/page/word
[16:52:27] <fab23> mechanicjay: it could be worst and a 7 kB Word Document :)
[16:53:05] <mechanicjay> With any luck he's written it in TeX!
[16:53:41] <fab23> can you convert that to .doc?
[16:53:53] <fab23> .oO( dump of code )
[17:00:07] <bryan> isn't a completely blank word document larger than 7kb?
[17:00:47] <ted-ious> Are you including macro viruses or no?
[17:01:06] <bryan> hehe
[17:05:45] <fab23> just keep LibreOffice updates and all will be fine. :)
[17:05:55] <fab23> s/updates/updated/
[17:23:13] <janrinok> mechanicjay, I still cannot see the document. Where are we looking?
[17:23:52] <mechanicjay> We don't get the document, just NC's claim of big it is in his journal post
[17:24:19] <janrinok> OK - I've seen the journal post.
[17:26:05] <janrinok> If I read that correctly we didn't approach him as 'human'. Oh, ok, it's my fault again.
[17:40:56] * Deucalion has patience and motivation level zero at this point.
[17:41:20] <janrinok> join us on #staff
[18:32:48] <kolie__> Mechanicjay
[18:33:03] <mechanicjay> That's me
[18:33:05] <kolie__> I've seen previews of it
[18:33:37] <kolie__> Hes still working on it.
[18:33:54] <kolie__> I've told jan a little bit.
[18:34:43] <kolie__> But he's going to release a document he sees as basically a potential way forward and a list of things that in his opinion are good aspirations to check off the list for a healthy growing community and platform.
[18:35:16] <kolie__> A lot of those items I've basically said I will commit to doing and working with him and a contract with the PBC
[19:32:57] <mechanicjay> in other news I'm using my 12inch powerbook G4 right now as second screen to reference documentation.
[19:34:10] <janrinok> are you happy with the powerbook?
[19:37:30] <mechanicjay> I mean it's a fun little machine -- not terribly useful at this point in it's life.
[19:37:50] <janrinok> they say that about me too......
[19:38:50] <fab23> /o\
[19:40:07] <mechanicjay> hey now!
[20:37:25] <kolie__> What's the full TMB story?
[20:47:27] <progo> well, we're not sure how life started on Earth exactly
[20:48:34] <kolie__> The buzzard or the egg?
[20:49:41] <progo> I took today off. I stayed up too late last night working on work stuff that HAD to be done, and didn't have enough time to sleep before going hiking as planned
[20:50:06] <progo> slept 10 hours, got diner breakfast. got a headache and don't feel like working on my personal project to build a new CMS
[20:50:10] * progo sighs lightly
[20:50:21] <progo> I think it's a pollen headache
[20:50:58] <mechanicjay> TMB left on good terms due to a philosophical difference over moderation and what is needed to build a healthy community.
[20:51:13] <kolie__> I remember the public side of that
[20:51:17] <kolie__> vaguely.
[20:52:16] <kolie__> What was his philosophical stake on moderation and health communities?
[20:53:00] <mechanicjay> I mean, the private side of that was A LOT of debate. The end result that we shut out AC's from the front-page as a regrettable, yet pragmatic thing.
[20:53:25] <Deucalion> kolie__, what are you so interested in TMB?
[20:53:28] <kolie__> And he was of the mind that AC's are integral to what he saw as the core of the site.
[20:54:29] <kolie__> "So interested" a bit of a reach but I for sure am interested in people that understand the operations and functions of SN - in his case specifically it's a short list of people who do perl , even shorter that can grok rehash.
[20:55:14] <kolie__> My interest and I bring it up here - was to contact him and pick his brain - I can't contact the fellow to even inquire if that's something I should pursue.
[20:55:56] <kolie__> At the very least I have a few questions he could answer that would help my pursuits.
[20:56:27] <progo> maybe someone who has a personal connection can tell TMB that kolie__ wants to chat
[20:56:30] <kolie__> Just getting a better understanding of history as well.
[20:58:20] <Deucalion> So you intend to attempt to continue with rehash?
[20:59:15] <progo> my recommendation: adopt Django, ProcessWire, or Flarum, and plugin lots of plugins. it would probably be a lot less of a nightmare
[20:59:19] <kolie__> Uh - if its my shot to call what would I do ? - rehash needs to be stabilized - automated/best practice cacoon built around it - and then promptly functionally replaced.
[21:00:26] <Deucalion> Why would it need automating if it is to be terminated? Wouldn't that be a waste of effort for what would be effectively a dead code base?
[21:00:47] <kolie__> No. It would greatly assist the operations and development and testing.
[21:01:12] <kolie__> And it makes it portable and able to be moved out of its current hosting environment / leadership,.
[21:01:19] <kolie__> And repeatable for DR purposes.
[21:01:33] <kolie__> It's more than just rehash - it's all auxillary services running at large.
[21:03:04] <Deucalion> What hosting environment would yo want it in?
[21:03:24] <kolie__> Not under the lock and key of ncommander.
[21:04:03] <Deucalion> Just transfer the Linode instances. Just a few clicks in the Linode account webpages.
[21:04:54] <Deucalion> And what development and testing if it's to be promptly functionally replaced?
[21:05:04] <kolie__> The instances need to be rebuilt. Who knows what's in them.
[21:05:30] <mechanicjay> are you worried about backdoor or something?
[21:05:35] <kolie__> That doesn't happen over night. And they need to be standing in parallel.
[21:05:40] <Deucalion> NC does. He nuked them all from orbit just 6 months ago and kinda put some of it back together remember.
[21:05:47] <kolie__> Best practice is theres a tons of hands, unknown config, open access to the world.
[21:06:18] <kolie__> THe automation/scripts/config e.g. terraform/ansible is a win win type of move.
[21:07:08] <kolie__> It addresses a lot of the issues NC brings up and ultimately get's him detached and comfortable in doing so.
[21:07:55] <Deucalion> OK. I'll wait and see what his latest vision doc says.
[21:08:37] <Deucalion> Where is this open access to the world?? That's a bit concerning
[21:08:59] <kolie__> EOL VMs - years out - on the web is what I've heard.
[21:09:42] <Deucalion> what do you mean VMs "on the web" ?
[21:10:14] <kolie__> Machines plugged into the wan with no security support or patches humming along.
[21:10:50] <kolie__> With unknown configuration running ancient software stacks.
[21:10:59] <kolie__> You got zero days? these could have +500 days.
[21:11:01] <Deucalion> Well there are plenty of those in the world, sure. Not sure you can fix all of those though.
[21:11:24] <kolie__> You can follow and practice published "best practices" of it infrastructure and operations.
[21:11:44] <kolie__> From what I've heard the current sytem is far from that for various reasons. Don't care - want to address it.
[21:12:41] <Deucalion> If you mean the SN VMs, as I say they were all nuked 6 months ago and repaved by NC. And there aren't that many of them. The only thing archaic is rehash and the apache + mod_perl it sits on.
[21:13:21] <kolie__> I can't speak to actuals because I've yet to finalize the contract being contemplated between myself and the PBC.
[21:13:37] <kolie__> But I've heard from numerous sources basically the same.
[21:14:35] <mechanicjay> I think "open to the world" is a GROSS exaggeration and misrepresentation.
[21:15:06] <kolie__> Perhaps but in the context of security I was using that term generally to describe the configuration I've been lead to believe is in place.
[21:15:08] <Deucalion> You can't speak to actuals kolie__ because you don't know what the actuals are.
[21:15:20] <kolie__> I've been told by staff and NC these things.
[21:15:52] <kolie__> But yea I don't know what they are as fact because I haven't seen them with my eyes.
[21:16:07] <kolie__> Just as they have been described by those close to the issue who would've seen them.
[21:16:22] <kolie__> And summarizing what I've been told.
[21:16:25] <mechanicjay> Well, regardless, what's left of the internal infrastrucre sounds like it's been updated to something more recent.
[21:16:35] <Deucalion> Perhaps you should livestream fixing them in real time :D
[21:16:56] <kolie__> If you want I can look into that. Probably going to be pretty boring.
[21:17:16] <kolie__> I don't have to be the only one doing it but I'm on the hook to make sure it happens.
[21:17:52] <Deucalion> mechanicjay, NC nuked all the servers in November and rebuilt with current OS
[21:18:08] <mechanicjay> Right, exactly.
[21:18:21] <kolie__> let me clarify with him on the os point.
[21:19:15] <Deucalion> IIRC they are all Ubuntu 22.04 now
[21:19:46] <kolie__> The os specifically yes he says most are up to date - some of the software versions are very old and the configs are who knows what.
[21:19:48] <Deucalion> Can't 100% confirm that as my ssh access has been killed.
[21:19:49] <progo> I remember from main page blog posts they're definitely on Ubuntu 22.04
[21:20:01] <kolie__> there was a 16.x system sometime but i think he finally got that out.
[21:21:13] <kolie__> Largely the configs are "mystery meat" of these systems.
[21:21:51] <kolie__> Previously a lot of stuff was open publically ports wise. I think some of its been trapped down now - there is a lot of questions because its not entirely sure what everything is configured or setup to protect.
[21:21:53] <Deucalion> Yes, some of the software versions are very old.... because rehash won't run on anything newer. The ancillary stuff like mail and IRC is all current
[21:22:24] <kolie__> NC knows a bit more - but even he can't answer a lot. Infrastructure by configuration can clarify that and be repeatable.
[21:22:57] <kolie__> The stuff that is immovable stones, rehash, i understand its containered/jailed a bit.
[21:22:58] <mechanicjay> "Previously a lot of stuff was open publically ports wise." Yes, for services that were being run.
[21:23:16] <kolie__> Mysql was public until some time - it still is but ip restricted.
[21:23:27] <kolie__> Interesting that a private linode network isnt used for that to the website.
[21:23:29] <mechanicjay> irc, http(s), imap(s), smtp
[21:23:31] <kolie__> But thats what it is.
[21:23:38] <kolie__> Yea ofcourse those things are necessary.
[21:23:40] <mechanicjay> ...it was -- it was all internal
[21:23:48] <mechanicjay> mysql was never fucking open to the public
[21:23:59] <kolie__> Again I don't know what is actual but just what was said.
[21:24:18] <Deucalion> Just pontificating then really.
[21:24:21] <mechanicjay> Well, if that's the line of shit he's feeding you, he's more of an asshole than I previously though
[21:24:26] <mechanicjay> I'm going for a walk.
[21:24:42] <kolie__> thats not all NC
[21:26:00] <kolie__> I'm going to let the past be what it is because I wasnt there. An automated publically documented build for the site seems like a very reasonable sane goal that should greatly help contributions, testing, and longevity of the site.
[21:26:08] <Deucalion> As I said, I'll wait to see his vision doc. Then I'll look forward to your implementation plan for that vision.
[21:28:27] <kolie__> Sounds like best practice, I like to follow those, and NC agrees that if this is the route forward he wont stand in the way. This is how I run my own production systems, these are what I teach my team, and this is what I've seen work well for long term operations and my colleagues in industry do it similarly.
[21:28:35] <Deucalion> As an aside, I'm intrigued who you think is going to be documenting build and deployment of rehash.
[21:29:20] <kolie__> If no one does its a task ill take on. Building old software with no dev environment or notes ive done enough. Not easy, never fun, but you get it done.
[21:30:11] <kolie__> It's just been moved as you said from wherever to a new ubuntu node - so we know its not futile.
[21:33:03] <Deucalion> ? that's just the OS. rehash basically transferred in a bubble of deprecated pinned apache and deprecated pinned mod_perl. Sure you can containerise that deprecated tower of software and deploy it wherever you like. But wherever and however you deploy it, it's still publically facing deprecated apache and mod_perl. It can't not be publically facing being a website.
[21:33:16] <Deucalion> So back to having to refactor the whole of rehash
[21:35:18] <kolie__> cacoon rehash and replace yea basically the line of thinking I'm going with rn.
[21:36:11] <kolie__> perl is well perl and given that janrinok was diligent but empty handed in finding anyone who does it and may help - NC not reliable - who else do we know that could feasibly do major hacking on perlk
[21:36:35] <kolie__> I'm comfortable isolating it as it is and putting it infront of an LB sure.
[21:37:11] <Deucalion> Why would it be janrinok's task to find someone to wrangle Perl? He's editor in chief
[21:37:41] <kolie__> Not my business - I was told he had been searching for years - as were others - and they were empty handed.
[21:37:49] <kolie__> Whos stepped up?
[21:37:56] <kolie__> Whos hacking the perl today?
[21:38:25] <kolie__> No ones commented that they feel they could refactor the whole of rehash as you put it.
[21:39:06] <kolie__> I don't care who did or didn't find someone - largely irrelvant to me and beyond the goal of my actions.
[21:39:22] <Deucalion> I think you mean ~behind~ an LB. As we used to have with nginx frontends sat in front with memcached and varnish in the mix.
[21:39:41] <kolie__> Yea three tier I was told put up.
[21:40:18] <Deucalion> I didn't say anyone had stepped up. I asked why you thought it would be janrinok's responsibility to find a perl wrangler
[21:40:41] <kolie__> I don't know whos responsbility it was - i know he was one person who was looking for a bit.
[21:41:26] <kolie__> I assume he is capable of doing so given that my interactions with him seem to be a pretty stand up sharp fellow. I don't think he came up empty handed because of lack of effort or intent.
[21:41:43] <kolie__> Just seems to be a shortage of willing perl hackers.,
[21:43:14] <kolie__> If there was an adequate candidate - is that the best option? One person with the knowledge? Perl is pretty esoteric at this point. Existing code base has inertia but - devils advocate - modern language is more accessible and widens the candidate pool. Could be an opportunity to grow the community to by inviting fresh bodies into the community via FOSS outreach.
[21:44:47] <kolie__> Lot of ways to skin this cat.
[21:50:27] <fliptop> FWIW I've been working on setting up a small dev server running the latest Rocky Linux OS w/ up-to-date perl, apache and mod_perl and trying to get rehash to play nice w/ it all.
[21:51:12] <Deucalion> My take on it is that the remaining Perl wranglers who could do it don't work on a volunteer basis. Like COBOL wranglers, they may be niche and sparse but they can command a hefty salary.
[21:51:13] <fliptop> However, I'm leaving tomorrow on a long road trip for a couple of weeks and may not have time to wrestle w/ it while I'm on the road. I'm bringing my laptop but I can't make any promises to commit time until I get back home.
[21:51:50] <Deucalion> Enjoy the roadtrip fliptop - best of luck not having two left feet :D
[21:52:22] <fliptop> Heh, that's not until end of July (officially) we're just practicing this coming week.
[21:52:45] <kolie__> Sure that's probably the case. So if it was decided that people wanted a rehash rework - that would have to be figured out and ultimately if its for contract then funded.
[21:53:39] <kolie__> If I was a perl wizard I'd throw my hat in but as it is not one of the things I've spent a great deal of my programming career on.
[21:54:47] <fliptop> I've been programming and maintaining perl code for more than 20 years, I can probably handle it but for the next couple of weeks my time is limited.
[21:58:42] <Deucalion> Who would fund it kolie__ ?
[22:01:19] <Deucalion> And have you spoken with the other board member Matt Angel at all?
[22:07:10] <kolie__> Who wants rehash refactored?
[22:07:39] <kolie__> People with those thoughts might have thoughts on how to accomplish such a feat.
[22:11:24] <Deucalion> I was asking you what your thoughts are. Do you know anyone who wants rehash refactored? And if so do they have any thoughts on how to fund it?
[22:12:26] <Deucalion> If not, it's all a bit hand-wavey isn't it? Some people (unknown) will want rehash refactored and they'll in turn be able to come up with ideas on how to fund it.
[22:13:35] <Deucalion> In which case, if it is these people (unknown) who want rehash refactored and will bring the funding to fruition, what exactly do you bring to the table in that scenario?
[22:13:46] <kolie__> "Like COBOL wranglers, they may be niche and sparse but they can command a hefty salary." I don't need to figure out that problem - I'm not aware of anyone who wants to keep perl around. If that's an opinion held by the community or anyone its unknown to me.
[22:25:55] <AzumaHazuki> you know, all the borderline-bikeshedding over technical implementation details is moot if the community sucks
[22:26:19] <AzumaHazuki> we need some actual rules. SN attempted Musk-style "free speech absolutism" and you see what it's gotten us
[22:28:04] <kolie__> well we should probably figure out some kind of community governance and get people involved and an understanding put together as a result of talking through some hard open questions.
[22:30:26] <Deucalion> Put AzumaHazuki in charge :D
[22:30:43] <AzumaHazuki> please, no
[22:33:55] <Deucalion> Just for a day?
[22:34:27] <AzumaHazuki> please, NO
[22:34:54] <Deucalion> OK. How about we recall EF and make him king for a day?
[22:36:13] <AzumaHazuki> that's an even worse idea
[22:37:31] <kolie_> Over here for now headed out of the office.
[22:53:48] <progo> I'd vote for AzumaHazuki
[22:54:33] <AzumaHazuki> please, no. i don't want any part of this
[22:56:18] <progo> I do agree about the bikeshedding comment though
[22:56:50] <progo> and sorry - I was just joking with the adding on endorsing you as supreme leader
[22:57:13] <progo> I haven't got a stake anyway. my contribution is like 2 or 3 main page posts, so I have no say :^)
[22:58:09] <AzumaHazuki> we need to ban toxic people. this libertarian horseshit was unsustainable from the start
[22:58:09] <kolie_> Even long time lurkers can contribute and participate right?
[22:58:47] <progo> sure
[22:59:09] <progo> basically I know I don't have the energy to take any responsibility for what happens this summer. just watching
[22:59:32] <kolie_> So it's the woke that one extreme and the libertarians on the other right? Rock and hard place. Azuma will guide us
[23:00:01] <progo> unfortunately you can't conscript a good leader
[23:00:03] <kolie_> NC got canceled by both sides already
[23:00:23] <kolie_> That's a dog pile everyone just loves jumping on
[23:05:39] -!- norayr has quit [Ping timeout: 256 seconds]
[23:10:42] * mechanicjay setsup a vm to try and get rehash built/installed
[23:25:25] -!- norayr [norayr!~norayr@37.252.sq.ypj] has joined #soylent
[23:32:06] <Deucalion> How, when and where did NC get cancelled by anyone?
[23:37:55] <kolie_> Tongue in cheek mate