#dev | Logs for 2018-04-05

« return
[00:00:41] -!- fungus [fungus!~ben@66.171.jo.vll] has joined #dev
[00:01:50] -!- fungus has quit [Remote host closed the connection]
[00:02:06] -!- fungus [fungus!~ben@66.171.jo.vll] has joined #dev
[00:02:54] -!- fungus has quit [Remote host closed the connection]
[00:03:10] -!- fungus [fungus!~ben@66.171.jo.vll] has joined #dev
[00:03:31] -!- fungus has quit [Remote host closed the connection]
[00:04:20] -!- fungus [fungus!~ben@66.171.jo.vll] has joined #dev
[00:04:43] -!- fungus has quit [Remote host closed the connection]
[00:05:23] -!- fungus [fungus!~ben@66.171.jo.vll] has joined #dev
[00:05:54] -!- fungus has quit [Remote host closed the connection]
[00:06:41] -!- fungus [fungus!~ben@66.171.jo.vll] has joined #dev
[11:11:52] <TheMightyBuzzard> fyngyrz, since it looks like you got the pull request sorted i'm just gonna go with routine stuff.
[11:12:41] <TheMightyBuzzard> for starters, go to your rehash directory and do: git remote add upstream https://github.com
[11:13:52] <TheMightyBuzzard> and every time you want to do a new batch of fixes, run the following...
[11:14:46] <TheMightyBuzzard> git checkout master && git pull upstream master && git push origin master && git checkout -b <new branch name>
[11:16:10] <TheMightyBuzzard> that'll keep your master branch up to date both locally and on github to make sure you're always starting from the most current position you can be.
[12:55:25] <fyngyrz> TheMIghtyBuzzard, I have not set up rehash locally at all. I have simply forked it on Github, edited it there, then done the pull. Once you let me know you've merged the pull, I delete the fork entirely. That way, I always fork a brand new, up-to-date rehash for working with. That should accomplish the same thing without me having to have rehash locally, I think. Yes?
[13:08:48] <TheMightyBuzzard> fyngyrz, not exactly making the most of git or github like that but i suppose it'll work if you feel the need to do it that way.
[13:09:45] <TheMightyBuzzard> oh, speaking of, your latest pull request...
[13:10:12] <fyngyrz> yes?
[13:10:23] <TheMightyBuzzard> doesn't need the <a> tags.
[13:10:58] <TheMightyBuzzard> use id=whatever inside the <h3> tags and you get the same result.
[13:11:01] <fyngyrz> The point of the <a name="thing"></a> is so that a link such as faq.html#thing will work
[13:11:35] <fyngyrz> I think that's from a much later HTML standard
[13:11:43] <fyngyrz> the id=thing
[13:11:52] <TheMightyBuzzard> like <h3 id="wits">
[13:12:26] <fyngyrz> right, lemme see what level of HTML that is. I try to go with HTML4 if possible, as tht covers the most browsers and phones
[13:12:31] <TheMightyBuzzard> possibly. it's how we're doing "send you back to the comment you just made" though so consistency would be preferred.
[13:13:02] <fyngyrz> okay. give me a second here so I can learn
[13:15:10] <TheMightyBuzzard> i did the same thing a year or two back and paulej72 said do it with id= instead or i wouldn't have said anything.
[13:16:05] <fyngyrz> okay, it's part of HTML 4.01; I'll use that
[13:16:14] <fyngyrz> so dump the pull, and I'll fix it
[13:16:40] <TheMightyBuzzard> nah, just modify that branch and the pull will update itself.
[13:16:52] <fyngyrz> unfortunately, that means it's harder to do, because I still need the hint.
[13:17:04] <fyngyrz> oh it will? That' sinteresting
[13:17:27] <TheMightyBuzzard> yup. right handy cause you can request changes without having to make a new pull request.
[13:17:28] <fyngyrz> So I dont' even need to make a new pull?
[13:17:35] <TheMightyBuzzard> ninja'd
[13:17:41] <fyngyrz> that seems... weird
[13:17:59] * fyngyrz gets after it
[13:18:45] <TheMightyBuzzard> cheers. i'll be in and out the next few hours. heading fishing shortly before noon i think though.
[13:31:22] <fyngyrz> okay, I'm done; can you please verify that you see the changes?
[13:50:16] <TheMightyBuzzard> they show up just fine. don't really need the [#wits] lines though. users aren't going to care if they can see the id of the link, just that the link works.
[13:57:11] <fyngyrz> The point is that *WE* need to see the [#wits] so that we can assemble the link properly.
[13:59:14] <fyngyrz> Since this isn't really an HTML page, just a template, there's no way to test it, either, which is annoying.
[13:59:57] <fyngyrz> no easy way, anyway
[14:01:16] <fyngyrz> You follow what I'm saying about assembling the link?
[14:02:50] <TheMightyBuzzard> if you want access to dev that can be arranged. you do gotta get comfortable using command line git for that if you're not though.
[14:04:21] <fyngyrz> oh, I'm comfortable enough with command line git, I use it all the time. I just didn't see a reason to load up my HD with the entire rehash project for tiny little edits like this.
[14:04:32] <TheMightyBuzzard> but yeah, i feel where you're coming from. they can be removed or <!-- commented out --> while we're creating the links just as well as they can now.
[14:04:38] <fyngyrz> especially since the process on github is so fast
[14:05:35] <fyngyrz> not following you on removal or commenting
[14:06:13] <TheMightyBuzzard> heh, all 34MB of it? you must got a smaller hdd than mine but whatever blows your skirt up.
[14:07:14] <TheMightyBuzzard> oh, putting <!-- [#whatever] --> will leave it in the template and on the page but it won't get rendered by the browser.
[14:08:03] <TheMightyBuzzard> you'd have to view source on the page to see it
[14:08:35] <fyngyrz> Right, but then you can't see it on the FA page, and you have to view source to figure out what the link is. The entire point of this evolution is to make it so you or I can go to the FAQ page, locate the answer, copy the link from the browser bar, then add the #wits part which we can *see* on the page, and then provide the link to the person who needs it.
[14:08:49] <fyngyrz> If you hide it, you just make it harder to do
[14:08:56] <fyngyrz> a lot harder, in point of fact
[14:10:23] <TheMightyBuzzard> nod nod. what i mean is while the links are being created, comment the one you just linked to out or just remove it. keeps you from losing your place but leaves the ones you haven't done yet visible.
[14:10:27] <fyngyrz> The entire reason for this is to make it easier for someone to point at a specific item in the FAQ. If you have to view source to do that, you're not making it easier, you're making it harder.
[14:10:46] <TheMightyBuzzard> ahhh, i get ya now.
[14:11:21] <TheMightyBuzzard> not for us but for whoever wants to link to it later. roger.
[14:11:37] <chromas> The faq is generated by a template? Does it change that often?
[14:11:44] <fyngyrz> right - exactly. I was trying to improve the actual usability of the FAQ
[14:12:10] <fyngyrz> The FAQ contains various things that will change depending on the project
[14:12:18] <fyngyrz> so they use a template to generate it
[14:12:31] <TheMightyBuzzard> chromas, only when someone complains about it being out of date loudly enough that it's less annoying to fix it than to listen to them. =P
[14:12:45] * chromas whistles nonchalantly
[14:13:27] <fyngyrz> chromas: https://github.com
[14:13:32] <TheMightyBuzzard> fyngyrz, okey doke. double checking and merging then.
[14:13:35] <chromas> But using anchors tag is the 'correct' way to do the fragment links
[14:13:43] <fyngyrz> lol
[14:13:47] <TheMightyBuzzard> #smake chromas
[14:13:47] * MrPlow smakes chromas upside the head with tp
[14:14:00] <fyngyrz> that's how I initially did it. But TMB wants id=, so there it is
[14:14:52] <fyngyrz> as long as it works, it seems reasonable. Support appeared in HTML 4.01 (I checked) so it's reasonably back-compatible
[14:15:12] <fyngyrz> support for id= I mean
[14:15:12] <TheMightyBuzzard> i don't personally care really. boss men up the chain prefer it. i'm still an HTML 2.0 kinda guy for preference.
[14:15:36] <chromas> At the very least, stick a link next to the title
[14:15:52] <chromas> With a little icon and a tooltip that says "permalink"
[14:16:30] <chromas> Anything less is racist
[14:16:44] <fyngyrz> and extremely right/left wing, too
[14:16:58] <TheMightyBuzzard> fyngyrz, "How can I support SoylentNews?" needs tagged too. inside the IF of course.
[14:17:27] * TheMightyBuzzard continues proofreading
[14:17:59] <chromas> While you're at it, you should make the faq be generated from a table in the db
[14:19:14] <TheMightyBuzzard> it is, actually. templates are slurped into the db at apache start and read from there.
[14:19:31] <chromas> Jesus, dude
[14:19:53] <TheMightyBuzzard> beats reading them from disk
[14:20:04] <chromas> But
[14:20:15] <chromas> The db is on a disk
[14:20:26] <chromas> Or does it use ram storage?
[14:20:29] <TheMightyBuzzard> nope. ndbd tables are held in ram.
[14:21:08] <TheMightyBuzzard> well on disk as well but our poor db servers cry every time we start them up.
[14:21:19] <chromas> I wonder why
[14:22:55] <fyngyrz> This is why I use static site generation; all the pages are pure HTML and both servers and browsers rip right through them. Can't do that on the story pages, obviously - comments are far too dynamic - but for something like the FAQ, I would definitely.
[14:23:12] <TheMightyBuzzard> 141M free on helium and all it does is serve the ndbd tables to the mysql instances on hydrogen and fluorine.
[14:23:53] <fyngyrz> But soylent is old, old code
[14:24:12] <chromas> Not that old. They should've known better
[14:24:14] <TheMightyBuzzard> fyngyrz, yeah, prolly the logical way to do that but i despair at finding all the places it'd need to be changed.
[14:24:20] <fyngyrz> I hear you
[14:25:11] <fyngyrz> I have an advantage they didn't, anyway - I have a generator that is optimized for exactly this
[14:25:31] <TheMightyBuzzard> okay, adding that one link is the only issue i see
[14:25:32] <fyngyrz> chromas, that's what aa_macro was actually designed to do
[14:26:35] <TheMightyBuzzard> nicotine break
[14:26:43] <fyngyrz> change has been made
[14:27:38] <TheMightyBuzzard> merged
[14:27:40] <fyngyrz> I'd have that entire FAQ heading thing in a style like {faqh3 Where did it come from?} and everything would generate from there
[14:28:13] <TheMightyBuzzard> now really nicotine break
[14:28:32] * fyngyrz loves his macro language and documentation generation system
[14:28:50] <chromas> Find a way to shove it into rehash
[14:29:43] <fyngyrz> Unfortunately, rehash is so badly designed it's nearly impossible to set up a test installation. Far too many dependancies
[14:30:10] <fyngyrz> again, not anyone's fault here - this is legacy awfulness from slashdot
[14:31:39] <chromas> Jesus wept
[14:31:52] <fyngyrz> What I get tempted to do now and then is actually create a slashdot/soylent work-alike written to be an atomic project
[14:32:03] <fyngyrz> in Python, of course. :)
[14:32:09] <chromas> Well there's pipecode, but that's php
[14:32:16] <fyngyrz> of fuck me not php
[14:32:22] <fyngyrz> that's even worse than perl
[14:32:39] <chromas> Probably easier to port to another lang than rehash though
[14:32:55] <fyngyrz> I'd like to do it black box - not port it
[14:33:51] <chromas> You mean push data into it without looking at the source?
[14:34:38] <fyngyrz> I mean write it from scratch by looking at what soylent does, rather than at the perl source code
[14:34:56] <chromas> Oh, I think that's what pipecode is
[14:35:01] <chromas> but in php :D
[14:35:13] <fyngyrz> fresh perspective. We know what we want from the site; it's hardly all that complicated
[14:35:35] <chromas> We could switch to wordpress
[14:35:50] <fyngyrz> also php
[14:35:54] <fyngyrz> and very, very slow
[14:36:24] <chromas> It's got more bugs though. More is better, right?
[14:36:35] <fyngyrz> Not that soylent is a busy site (unfortunately) but something like this needs to be scalable, too
[14:36:59] <fyngyrz> you're yanking on my chain. nm.
[14:37:17] * chromas enjoys yanking it
[14:37:23] * fyngyrz sighs
[14:37:56] <fyngyrz> anyway. perl is what we have, and templates too. So there you have it.
[14:38:43] <chromas> I wonder if they'd consider another software if it was easy to import the current data
[14:38:45] <TheMightyBuzzard> perl's better than python for what we're doing, really. it is hands down the king of scripting languages when it comes to text jiggery pokery.
[14:38:56] <fyngyrz> In order to do the nice iconic permalink thing, more needs done than just editing the FAQ file. This will work as is, and perhaps at some point It'll go that way
[14:39:01] <chromas> What about rust?
[14:39:22] <TheMightyBuzzard> rust would take about 5x as many lines of code
[14:39:42] <chromas> But the speed
[14:39:51] <fyngyrz> TMB, disagree entirely.
[14:40:37] <chromas> How much text jiggery pokery is there though? Isn't it just template filling?
[14:40:50] <TheMightyBuzzard> fyngyrz, you're entitled to be wrong =P
[14:40:52] <fyngyrz> Python's not only good at it, it's good at doing it _clearly_
[14:41:12] <chromas> Python has Whitespace Privilege though
[14:41:22] <fyngyrz> while regular expressions are powerful, they're also opaque as hell
[14:41:30] <TheMightyBuzzard> perl text manipulation's clear as anything to me.
[14:41:51] <fyngyrz> which is terrible for maintenance and not particularly good for creation, either
[14:42:02] <TheMightyBuzzard> python uses the exact same regexes though, it just takes more lines to create them.
[14:42:34] <fyngyrz> yes, I'm not a fan of using regex to manipulate text, because of that exact reason: regexes get opaque very quickly
[14:43:01] <TheMightyBuzzard> can. we try not to get too funky with the cheese whiz though.
[14:43:02] <fyngyrz> Python can be written to be nearly as opaque as perl. But you have to really try had. :)
[14:43:04] <fyngyrz> hard
[14:43:32] <fyngyrz> anyway, doesn't matter, perl's what we have
[14:43:33] <TheMightyBuzzard> mostly because we know we're going to have to be able to read them later.
[14:44:07] <TheMightyBuzzard> ya. mod_perl not necessarily though. i'd love to get rid of it in favor of fcgi or fast cgi or whatever
[14:44:49] <chromas> Then you could switch ti nginx
[14:44:55] <TheMightyBuzzard> nod nod
[14:45:05] <chromas> and c#
[14:45:53] <TheMightyBuzzard> c# is ancient. did you mean go?
[14:46:19] <chromas> Go has a better regex implementation according to google
[14:46:23] <chromas> than perl
[14:47:01] <TheMightyBuzzard> nothing has a better regex implementation than perl. except maybe awk.
[14:47:25] <chromas> It uses the awk implementation
[14:47:38] <TheMightyBuzzard> that's fine then
[14:47:48] <fyngyrz> Nothing is better than NOT using regex at all. :)
[14:48:05] <chromas> https://swtch.com
[14:48:44] <TheMightyBuzzard> true but sometimes it's insanely easier to write a regex than to write a specific text transform.
[14:50:39] <TheMightyBuzzard> like there's no way in hell i'd ever try the single regex it took to limit the number of unicode diacritic marks per character as a proper algorithm.
[14:50:56] <fyngyrz> oh, I know. it's easy to do a lot of things the shortcut way. But it's not a good idea in terms of maintainance and sharing
[14:53:27] <TheMightyBuzzard> sorry, three lines of perl for that. requires zero maintenance on our part unless perl's core libraries catch a bug.
[14:54:55] <fyngyrz> chromas, great link, thank you
[14:55:40] <TheMightyBuzzard> considering how many regexes we run every page load through and our rendering speed, you gotta give perl props for regex optimization.
[14:58:05] <fyngyrz> considering that is the design, yes. It's a terrible design, though.
[14:58:19] <chromas> If you used rust or something, you could do compile-time regex optimization
[15:00:32] <fyngyrz> were I to approach this, every page would be a static page from the POV of the webserver and the end user. The pages would be changed in the background, completely transparently
[15:00:50] <fyngyrz> doing this at page rendering time is incredibly costly
[15:01:52] <fyngyrz> soylent has the enviable feature of not having to shoehorn ads onto the page, and so has no actual need of dynamic content, multi-server or single server
[15:02:44] <TheMightyBuzzard> it'd have to be stored on-disk then and serving on-disk content is slooooooow.
[15:03:00] <fyngyrz> no. they can be stored in a ramdisk. Fast as hell.
[15:03:32] <TheMightyBuzzard> nope, not even gzipped. wouldn't fit in an affordable server with our current load.
[15:03:42] <chromas> Even ads don't need dynamic pages because javascript
[15:05:11] <chromas> Is reading a user page from disk slower than merging db content with a template?
[15:05:23] <fyngyrz> well, it's kind of a different conversation, but tl;:dr of it is that a responsible ad system serves only from local storage anyway, both to make sure that the user experience doesn't bog down, and so that advertisers can't change the content without the consent of the site
[15:06:06] <fyngyrz> TMB, your current load, near as I can tell, should be very small
[15:06:08] <chromas> Is that a thing that exists, though
[15:06:22] <fyngyrz> if it isn't, that's a design error. soylent pages are very small
[15:06:23] <TheMightyBuzzard> chromas, no. generating it would take exactly as long as serving it though and with disk access overhead added on.
[15:06:40] <chromas> Aren't all the disks ssds?
[15:06:54] <TheMightyBuzzard> fyngyrz, yes but you'd need every single page rendered for every single user. or at least all the ones accessed recently.
[15:07:14] <fyngyrz> nope
[15:07:16] <chromas> Generate, store to disk, then cache as you do already
[15:07:37] <fyngyrz> pages only need generated when they change, and that can (and should) be done outside the webserver
[15:07:40] <TheMightyBuzzard> fraid so. there is a LOT of dynamic content on every single comment.
[15:07:57] <fyngyrz> example:
[15:07:59] <fyngyrz> ?
[15:08:17] <chromas> The comments make sense I guess because there's different threading styles
[15:08:28] <chromas> And thresholds and stuff
[15:08:51] <fyngyrz> This can all be built into the page
[15:09:01] <TheMightyBuzzard> comment scores and friendship are the big ones but yeah, rendering style means a whole lot of different versions as well.
[15:09:16] <TheMightyBuzzard> it could but we're not interested in putting the javascript necessary to do so.
[15:09:30] <fyngyrz> most of it (perhaps even all of it) could be css.
[15:09:58] <fyngyrz> javascript is, IMHO, a last resort
[15:10:01] <TheMightyBuzzard> dynamic comment arrangement can't, unless you know something i don't.
[15:10:22] <fyngyrz> by arrangement, you mean.... ?
[15:10:40] <TheMightyBuzzard> oldest first, newest first, threading model or flat
[15:10:59] <fyngyrz> those should be different pages, I would think
[15:11:01] <TheMightyBuzzard> threshold, breakthrough
[15:11:20] <TheMightyBuzzard> that'd be quite a lot of different pages because you'd have to make a copy of each for each user.
[15:11:32] <TheMightyBuzzard> karma adjustments
[15:11:52] <fyngyrz> what do karma adjustments do?
[15:12:13] <fyngyrz> in a dynamic s4ense, I mean
[15:12:13] <chromas> Optionally give a score bonus
[15:12:24] <TheMightyBuzzard> let the user say "ACs suck, -5 all their comments" or "Funny doesn't interest me, -1 all of them"
[15:12:43] <TheMightyBuzzard> or a bonus, yeah
[15:13:03] <fyngyrz> wouldn't it make more sense, performance-wise, to put that load on the individual browser, rather than the server?
[15:13:10] <TheMightyBuzzard> so comment scores have to be calculated at render time
[15:13:15] <chromas> Can you do that without js though?
[15:13:17] <fyngyrz> but not by the server
[15:13:29] <fyngyrz> even if it rrequires js, what's the problem?
[15:13:34] <chromas> js
[15:13:37] <TheMightyBuzzard> sure but web 1.0 is kinda our thing. folks would riot if we did it via js.
[15:14:11] <fyngyrz> well, there you go then, you're stuck with a terrible design, like I said. :)
[15:14:15] <TheMightyBuzzard> "fuck beta" would be uttered.
[15:14:19] <fyngyrz> no doubt
[15:14:27] <chromas> The design could still be better though
[15:14:54] <fyngyrz> however, again, were I doing this from scratch, it would be done differently, and it would be fast, scalable, and atomic.
[15:15:01] <TheMightyBuzzard> it's okay though, we render faster than most of the heavily trafficked sites and we do most all of it server-side.
[15:15:24] <fyngyrz> yes, it definitely works as-is
[15:16:04] <TheMightyBuzzard> and i mean it held up under /.'s load at peak traffic, so we've got plenty of wiggle room.
[15:16:44] <fyngyrz> right, had that in mind
[15:17:10] <fyngyrz> I just wish it was easier to set up and use/test
[15:17:44] <chromas> That's probably not something they had to deal with
[15:18:40] <TheMightyBuzzard> working on that. almost have the install procedure down to two lines or so.
[15:18:48] <fyngyrz> what it does - IMHO - is retards development. One of the things you see with slashdot is that they basically can't change anything. The site's been devoid of needed changes forever, because actually modifying the code is a nightmare
[15:19:03] <TheMightyBuzzard> actually i think i do i just don't have it all thoroughly tested yet.
[15:19:46] <TheMightyBuzzard> well only a nightmare starting out. once you figure out how things are done it's fairly easy.
[15:21:16] <TheMightyBuzzard> it may be spaghetti but it's consistent spaghetti.
[15:22:51] <fyngyrz> the INSTALL document is, on a scale of 1-10 where one is "nightmare" and 10 is "copy this single file", about a 3
[15:24:12] <TheMightyBuzzard> yeah, i'm pretty sure i can simplify the docs. haven't yet though because this is going to need to go into an experimental branch since it uses the highest versions of all the deps.
[15:24:31] <TheMightyBuzzard> highest possible anyway.
[15:24:33] <fyngyrz> a ten has no deps. :)
[15:25:07] <fyngyrz> a nine includes all deps and is "copy these files"
[15:25:09] <TheMightyBuzzard> mod_perl isn't exactly rushing to keep up with the latest perl versions
[15:25:49] <TheMightyBuzzard> another reason i'd prefer fcgi or fastcgi or whatever
[15:25:56] <fyngyrz> both perl and Python have recently made some changes that actually break older programs. Insane, but there it is
[15:27:04] <chromas> port to perl 6
[15:27:24] <fyngyrz> dont' screw around, port to perl 9 and wait for them to catch up
[15:27:55] <TheMightyBuzzard> yeah, good thinkin
[15:27:58] <fyngyrz> lol
[15:28:25] * TheMightyBuzzard can't stand perl 6
[15:28:35] <TheMightyBuzzard> they fucked up regexes
[15:28:35] * fyngyrz can't stand Python 3
[15:28:47] <TheMightyBuzzard> they're all pythony now
[15:28:48] <fyngyrz> they fucked up printing and a whole raft of other things
[15:29:38] <fyngyrz> print statement, I mean
[15:29:49] <fyngyrz> stupid. Incredibly stupid (Python changes)
[15:30:02] <fyngyrz> have not been following the perl changes, just see the comments
[15:31:30] <TheMightyBuzzard> well perl 5 isn't going anywhere. six is more an entirely new language than a major version of the same language.
[15:31:46] -!- upstart [upstart!~dbot@5433-9y82-5383-3z92-9p21-f9r1-0o15-jr91.dhcp4.chtrptr.net] has joined #dev
[15:31:53] <fyngyrz> same with Python 2.7; it's not broken; I have no intent to "fix" it. :)
[15:32:26] <TheMightyBuzzard> gnawed gnawed
[15:32:38] <fyngyrz> and Python 3 is the same: it's not Python. It's something new (and incompatible.)
[15:33:23] <fyngyrz> in fact, the ONE thing I would really like Python to have, they didn't add in Python 3, so fuck 'em
[15:33:28] <TheMightyBuzzard> really, perl 6 is closer to java or python than to perl 5
[15:33:42] <fyngyrz> really?
[15:33:44] <TheMightyBuzzard> yup
[15:33:57] <fyngyrz> wow, I bet that roils a few crania
[15:34:26] <TheMightyBuzzard> i don't think anyone but the perl 6 fanboiz are touching it for more than a couple days.
[15:34:38] <fyngyrz> I mean, I hate perl, but if you LIKE perl, I wold imagine that the way it is is the way one wants it to be
[15:35:26] <TheMightyBuzzard> yar. there are some things that could use changing but not rewriting the entire language to where it's not even recognizable as perl anymore.
[15:35:31] <fyngyrz> Back in the day, I wrote some huge projects in perl. I got used to it, but I never liked it
[15:36:13] <fyngyrz> the site here is perl 5?
[15:36:19] <TheMightyBuzzard> yep
[15:36:37] <fyngyrz> have you cloned the perl5 code? :)
[15:36:50] <TheMightyBuzzard> yeah, first proper coding job i had was writing an online billing/cc system in perl.
[15:36:55] <fyngyrz> I cloned Python 2.7 here, "just in case"
[15:37:36] <TheMightyBuzzard> naw. perl 5 isn't going anywhere for a long, long time. nobody using it is interested in perl 6 really.
[15:38:12] <fyngyrz> can't say the same for Python 3. There's a pretty strong push against people who resist
[15:38:32] <fyngyrz> I can't be pushed, though. Stubborn doesn't even begin to cover it :)
[15:39:15] * TheMightyBuzzard chuckles
[15:40:18] <TheMightyBuzzard> yeah, most people who went to python from perl are the "we lurve change for the shiny and new" types. there's really not enough difference to warrant the switch.
[15:40:54] <fyngyrz> That's an interesting perspective.
[15:41:07] <TheMightyBuzzard> whereas perl types are old, crusty, and hate change unless it saves them lots of effort.
[15:41:21] <fyngyrz> it may be an introspection fail, but I don't think that applies to me at all
[15:41:41] <TheMightyBuzzard> well you probably acquired your old and crusty after the switch
[15:42:09] <fyngyrz> oh no, been old and crusty for some decades now, I assure you
[15:42:29] <TheMightyBuzzard> well yeah but perl and python have been around since dinosaurs roamed the earth
[15:42:50] <fyngyrz> I was old and crusty before there was a web
[15:43:35] <TheMightyBuzzard> not me but i was already practicing up for it
[15:44:13] <fyngyrz> ha. Perl came out after the Amiga, and when the Amiga was released - 1985 - I was already known for digging in my heels and insisting that things be done "my way."
[15:44:40] <TheMightyBuzzard> we had a blacksmith shop in my home town that the old guys used to hang out in front of all day drinking coffee. i saw em one day and said "damn, that's the life for me". so i started practicing to be a grumpy old man at 15 or 16.
[15:44:46] <fyngyrz> And considering I was able to retire based on that within just a couple years of then, I'm good with it
[15:45:59] <fyngyrz> I've been doing whatever the hell I want since 1987, but it turns out that what I want is basically to write and program, so there you go
[15:46:32] <TheMightyBuzzard> retirement's for the birds. i gotta have something to keep myself busy or i'll have nothing to do but start on my "people who need a good killing" list.
[15:47:01] <chromas> Fooshin' and shitpostin'
[15:47:06] <TheMightyBuzzard> sit on your ass retirement anyway.
[15:47:22] <TheMightyBuzzard> chromas, and writing new bugs for Bytram to find.
[15:47:42] <fyngyrz> retirement for me just meant no longer worrying about income. Basically nothing else changed; I was already the boss of me. I get up every morning and do something fun that is essentially the same sort of thing I was doing to earn
[15:48:45] <TheMightyBuzzard> guess i'm retired too then, heh. haven't done anything but what i felt like doing since... um... well since around when i could drink legally.
[15:48:53] <fyngyrz> there you go
[15:49:07] <fyngyrz> and good fo ryou
[15:49:16] <TheMightyBuzzard> self employment with modest goals is the best shit ever.
[15:49:19] <fyngyrz> as far as I'm concerned, this is the optimum state
[15:49:52] <TheMightyBuzzard> ya, some folks really don't get that and i feel sorry for them.
[15:50:24] <TheMightyBuzzard> they gotta chase them big bucks or they don't understand that you gotta have marketable skills to be a marketable employee
[15:50:31] <fyngyrz> I don't. They're thinking beings. I'm all for them stewing in their own juices.
[15:52:02] <TheMightyBuzzard> don't get me wrong. i think they're complete idiots. i pity them having to deal with a head full of idiocy though.
[15:52:28] <fyngyrz> Society is a series of pitfalls for the poor thinker. So their natural state, most of them, is to live in pits.
[15:52:43] <fyngyrz> and they spend a lot of time going "hey, this is such a nice pit"
[15:53:15] <TheMightyBuzzard> which is why i keep trying to enlighten folks about the dynamics of capitalism. if you're going to be in a system, you should probably learn that system.
[15:53:24] <fyngyrz> mortgages, credit-card debt, loud neighbors, drugs... ugh
[15:54:36] <TheMightyBuzzard> anyway, ima go check the temp and wind outside, have a smoke, and decide whether i need to make a sammich and go fishing or not.
[15:54:53] <fyngyrz> the fish are no doubt waiting
[15:55:40] <TheMightyBuzzard> peak times today were around sunrise and sunset. there's a smaller peak around noon though.
[15:58:21] <fyngyrz> where are you?
[16:02:17] <fyngyrz> more specifically, where are you fishing?
[16:03:37] <TheMightyBuzzard> western tennessee, little town north of jackson
[16:04:14] <TheMightyBuzzard> probably gibson county lake since i'd like to run jug lines and they got the best jug fishing around here.
[16:04:47] <fyngyrz> cool
[16:05:24] <fyngyrz> I live just west of Fort Peck Lake in NE Montana. I rock hunt along the shores
[16:06:01] <TheMightyBuzzard> got inspired while i was outside though. went and checked out my boat and it's full of leaves. then saw a neighbor's cat taking a nap on my wood pile. so now i'm gonna take a nap instead then go clean out the boat so it's ready for next time i take a notion to go fishing.
[16:06:54] <fyngyrz> cats lead the way
[16:23:55] <fyngyrz> so, TMB, when you get back from nap/fishing/boatCleaning:
[16:24:28] <fyngyrz> when will the FAQ changes move from the repo to the site? Wha's the procedure/custom for that?
[16:24:43] <fyngyrz> s/Wha/What/
[19:55:58] -!- fungus has quit [Read error: Connection reset by peer]
[19:56:03] -!- fyngyrz_ [fyngyrz_!~fyngyrz@66.171.jo.vll] has joined #dev
[19:58:30] -!- fyngyrz has quit [Ping timeout: 250 seconds]
[19:59:22] -!- SoyGuest13625 has quit [Ping timeout: 250 seconds]
[20:00:03] -!- cosurgi [cosurgi!~cosurgi@pskvpf01.bl.pg.gda.pl] has joined #dev
[20:00:42] -!- fungus [fungus!~ben@66.171.jo.vll] has joined #dev
[20:01:03] cosurgi is now known as SoyGuest12382
[21:14:45] <TheMightyBuzzard> fyngyrz_, generally when we get enough changes piled up to warrant a live release and Bytram has had a chance to kick all the tires.
[21:17:04] <TheMightyBuzzard> with purely cosmetic template changes there's no need to wait though since templates can be updated without touching any files. just need a few minutes testing on dev and they can get pushed live.
[21:18:15] <TheMightyBuzzard> right now i'm still fighting off that nap i just took and then working on tonight and tomorrow's dinner, eating tonight's, and probably watching a movie or something with the roomie.
[22:04:12] <fyngyrz_> okay, wasn't pushing, was just curious
[22:04:23] <fyngyrz_> hm
[22:04:29] <fyngyrz_> I seem to have an underscore
[22:05:17] -!- fyngyrz_ has quit []
[22:05:31] -!- fyngyrz [fyngyrz!~fyngyrz@66.171.jo.vll] has joined #dev
[22:05:40] <fyngyrz> thta'es better
[22:16:43] <TheMightyBuzzard> i can probably get them rolled out (dev and prod) tomorrow morning unless it's a three cup of coffee morning.
[22:20:22] * TheMightyBuzzard gets back to foods prep
[22:20:28] * Bytram has not looked at any of the changes so far...
[22:20:52] <Bytram> I have today off, and my next day off isn't until next thursday. I also have some IRL stuff to catch up on.
[22:21:41] <Bytram> I might be able to squeeze in some testing, but it would be a huge help to have dev updated to the latest (so I have a place to test) and a clear list of changes (so I know what to test) =)
[22:27:01] <TheMightyBuzzard> Bytram, nothing to test yet, just cosmetic changes to templates. enjoy your day off.
[22:27:37] <Bytram> okaaay, but my spidey sense is tingling that something non-cosmetic will work its way in w/o warning.
[22:28:02] <Bytram> if nothing else, would help to have release notes enumerating each change and what it does.
[22:28:19] <Bytram> not just for me, but for the community, too. =)
[22:29:31] <TheMightyBuzzard> aren't any code changes yet. just purely cosmetic stuff.
[22:30:24] <Bytram> and... there's no code in stories, but we still manage to make mistakes (is why we try to 2nd every story before it goes out)
[22:31:04] <TheMightyBuzzard> pshaw. how do you expect us to have stuff to fix if you won't even let typos through?
[22:31:29] * Bytram chuckles
[22:32:44] <TheMightyBuzzard> check the last four closed pull requests if you really feel the need for some QA today. you can probably do them all in half an hour or less with time for a bathroom break and a snack break.
[22:33:10] * TheMightyBuzzard wanders back to food prep
[22:34:12] <Bytram> https://www.youtube.com
[22:34:13] <upstart> ^ 03Up & Down Baxter St., Echo Park, Los Angeles, CA - YouTube
[22:43:17] -!- fungus has quit [Remote host closed the connection]
[22:43:34] -!- fungus [fungus!~ben@66.171.jo.vll] has joined #dev
[22:50:11] -!- fyngyrz has quit [Quit: Leaving...]
[22:50:19] -!- fyngyrz [fyngyrz!~fyngyrz@66.171.jo.vll] has joined #dev