Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2021-08-27

00:07:33 Quit dconrad (Remote host closed the connection)
00:19:55 Join dconrad [0] (~dconrad@208.38.228.17)
00:24:10 Quit dconrad (Ping timeout: 240 seconds)
00:26:45 Join Saijin_Naib [0] (~Saij@2603-7081-1d05-7230-b884-c4bd-58d2-14e2.res6.spectrum.com)
00:28:08 Quit pixelma (Quit: .)
00:28:08 Quit amiconn (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.)
00:30:53 Join pixelma [0] (marianne@p200300ea87293400305e95fffec66ff3.dip0.t-ipconnect.de)
00:30:53 Join amiconn [0] (jens@p200300ea87293400305e95fffec66ff3.dip0.t-ipconnect.de)
00:35:17 Join Saijin-Naib [0] (~Saij@2603-7081-1d05-7230-b884-c4bd-58d2-14e2.res6.spectrum.com)
00:36:27 Quit Saijin-Naib (Client Quit)
00:38:40 Quit Saijin_Naib (Ping timeout: 240 seconds)
00:47:30***Saving seen data "./dancer.seen"
01:00
01:19:29rb-bluebotBuild Server message: New build round started. Revision 965572705b, 303 builds, 9 clients.
01:23:11 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:b018:e136:52f:ec39)
01:27:27 Quit ZincAlloy (Ping timeout: 240 seconds)
01:29:57rb-bluebotBuild Server message: Build round completed after 627 seconds.
01:29:59rb-bluebotBuild Server message: Revision 965572705b result: All green
01:53:50 Join ZincAlloy [0] (~Adium@ip5f5abcae.dynamic.kabel-deutschland.de)
01:57:53 Quit ZincAlloy (Ping timeout: 248 seconds)
02:00
02:47:31***Saving seen data "./dancer.seen"
04:00
04:47:32***No seen item changed, no save performed.
05:00
05:01:18 Join tomato3 [0] (~tomato@user/tomato)
05:03:28 Quit yang (Ping timeout: 252 seconds)
05:03:28 Quit tomato (Ping timeout: 252 seconds)
05:03:28 Nick tomato3 is now known as tomato (~tomato@user/tomato)
05:04:21 Join yang [0] (~yang@212.129.38.130)
05:28:20 Quit yang (Changing host)
05:28:20 Join yang [0] (~yang@fsf/member/yang)
06:00
06:20:24speachy_bilgus: there are a bunch of other errors possible too
06:47:33***Saving seen data "./dancer.seen"
08:00
08:22:45 Join advcomp2019 [0] (~advcomp20@user/advcomp2019)
08:24:42 Quit advcomp2019_ (Ping timeout: 250 seconds)
08:47:34***Saving seen data "./dancer.seen"
10:00
10:10:42braewoodsincluding the dreaded "out of cheese" error :P
10:47:38***No seen item changed, no save performed.
11:00
11:09:17 Quit gsora_ (Quit: memes 'n' shit)
11:09:36 Join gsora [0] (~gsora@140.238.174.213)
11:44:29 Join ZincAlloy [0] (~Adium@ip5f5abcae.dynamic.kabel-deutschland.de)
11:48:49 Quit ZincAlloy (Ping timeout: 252 seconds)
11:49:10 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:2500:38fb:dfc1:8e50)
12:00
12:33:37 Join lebellium [0] (~lebellium@2a01:cb10:2e:2000:ed4f:e8b7:c15e:280d)
12:47:41***Saving seen data "./dancer.seen"
12:59:00 Join skipwich [0] (~skipwich@user/skipwich)
13:00
13:04:36 Quit f1refly (Ping timeout: 276 seconds)
14:00
14:47:45***Saving seen data "./dancer.seen"
16:00
16:08:09_bilguswell crap just realized my arg parser needs to count zeros in order to successfully recreate a parsed decimal DUH
16:23:13 Join f1refly [0] (~f1refly@p2e50bd38.dip0.t-ipconnect.de)
16:47:48***No seen item changed, no save performed.
16:51:43 Quit lebellium (Quit: Leaving)
17:00
17:50:32speachyhey, as a early fyi −− the server hosting the rockbox infrastructure is going to have to find a new home by the end of the year
17:51:54speachyworst case it'll get relocated to a rack at home, but that's not ideal due to the rather crappy bandwidth here.
17:52:45speachy(putting aside rockbox stuff, it hosts my entire online presence, so it _will_ remain online.
17:53:43speachysuggestions welcome, but keep in mind donations are practically nonexistent these days.
17:56:54munkisI would volunteer if not for the fact that my bandwith is probably worse than yours.
17:59:52speachythere is about 2K in the war chest
18:00
18:00:39speachy~400GB/mo bandwidth usage, and ~100G of disk.
18:01:05speachy(2/3rds of which are a month's worth of daily builds)
18:02:46speachyto make an ongoing story short, $dayjob decided to effectively shutter our entire facility, and our/my last day is new year's eve.
18:08:08speachy(happy new year)
18:08:24speachy(and, heh, happy birthday to me)
18:47:50***Saving seen data "./dancer.seen"
18:50:26 Quit ZincAlloy (Quit: Leaving.)
19:00
19:19:59munkisouch.
20:00
20:04:50speachyHmm, there are some viable VPS options, if it comes to that.
20:27:48 Join dconrad [0] (~dconrad@208.38.228.17)
20:29:30 Quit dconrad (Client Quit)
20:47:52***Saving seen data "./dancer.seen"
21:00
21:04:22Arsenspeachy: I'd be doing VPS' + bunnycdn personally
21:04:30Arsenhttps://www.hetzner.com/cloud decent provider
21:04:39Arsenbcdn can be used to lighten the load on binary downloads
21:05:51ArsenI'm very much open to helping with this, I've got some experience
21:07:08speachyok, bunnycdn would be on the order of $4-20/month with our current bandwidth loads. depending on where it's coming from. not really worth the extra cost over a standard vps
21:08:08speachyvs (eg) hetzner's CPX31 offering.
21:08:41Arsenyeah idk what kinda load you have
21:09:06Arsenoh also you don't have to use all bunny regions
21:09:27ArsenI personally only use EU as a front for a backblaze b2 bucket
21:10:02Arsenif it's not outside the possibility of going over the 40TB limit (or having very massive download traffic, enough to slow down origin), it's probably worth to get it for $4 usd
21:11:01Arsen^ the above region thing works out for me since I don't need low latency delivery usually
21:11:03speachydepending on bot activity, we're on the order of 300GB/mo these days.
21:11:09Arsenah, that's not bad
21:11:34speachydirect www serving. git and stuff like backups are more, but realistically not that much of a jump.
21:11:57Arsenif that server is still up you could check with ethtool
21:12:22speachyoh, there's plenty of non-rockbox stuff on this box
21:13:06Arsenifconfig, not ethtool, sorry
21:13:08speachy(though rockbox stuff is easily >75% of normal load these days)
21:13:45_bilgusspeachy sorry to hear that :/
21:14:27speachythe state of the infrastructure is far, far improved over what it was when I took it over; migrating elsewhere is a much more straightforward prospect now.
21:14:55speachythough this might be the impetus to finally ask for control over the domain name.
21:15:41_bilgusand why not
21:16:17_bilgusIs the forum able to be read only now?
21:16:37speachyyeah, that can be toggled at any time, why?
21:16:50speachy_bilgus: drop-dead date is end of the _year_ not the month
21:16:57_bilgusIt would be nice to host that somewhere
21:17:35_bilgusthat'll be here before we know it :p
21:17:36speachythe forum traffic is 2-3GB/mo,
21:18:04_bilguswe had discsssed moving it somewhere for posterity
21:18:35speachyit's honestly a rounding error on the overall system.
21:18:58speachythe wiki is the thing that needs to be taken out back and shot
21:19:31_bilguseh thats what I meant read only the wiki
21:20:44speachywww.rockbox.org (incluidng the wiki and tracker) is ~30-40GB/mo, 4GB of disk space, and where most of the operational headaches come from
21:20:55speachy(the wiki, I mean)
21:21:28_bilgusits a treasure trove of info
21:21:49speachymaking it read only won't change the bass-ackward insanity that is its architecture.
21:21:52_bilgusitd be nice to push it off on githubs servers
21:22:05_bilgus:p
21:22:06speachythe right thing to do is export it staticly
21:22:20Arsenone could probably write an ad-hoc parser to take the weird bbcode and convert it to something sane like asciidoc, or markdown
21:22:38Arsenthere's very many tools to convert that further
21:22:46_bilguseven if it was 10000 pdfs
21:23:03speachyyeah, I've done a few passes at exporting it into a different markdown
21:23:31speachywell, if the next few months end up going as I kinda expect, I'm going to have a _lot_ of time on my hands.
21:23:58_bilgusat least someone oes
21:24:01_bilgusdoes
21:24:20_bilgusIve been at 8-12 hr days for like 4 months now
21:24:31_bilgusand gutting a house
21:26:00 Join Moriar [0] (~moriar@107-200-193-159.lightspeed.stlsmo.sbcglobal.net)
21:26:54speachyfosiwki has a static export plugin that works pretty well, actually. I just have to put together a template to get the output to match the rest of our site.
21:27:06speachy(that was the outcome the last time this subject came up)
21:30:29speachyone thing I can do is completely wipe the cache every day.
21:30:55speachy(kinda surprised I didn't do that already..)
21:31:18_bilguswhat does that accomplish?
21:32:02speachyit'll keep the cache's cache (yes really) from eventually wedging itself
21:32:12speachyand taking down the whole www server int he process
21:33:47_bilgus... wow
21:34:19speachybut it does run the risk of edits-in-progress getting possibly lost.
21:35:11_bilgusvery random
21:35:32speachy(because the user session could get nuked)
21:37:10Arsensounds like good design...
21:37:39speachyit's not normally a problem, but when a new bot storm hits, the server has enough bandwidth to cause things to keel over
21:48:45speachyok, daily manual builds are now on the daily build dl page
21:54:06_bilgusany idea(s) on the best way to represent the fraction part of a number returned from a string (atof in an int?)
21:55:28_bilgusI could do it by multiplying so .9999 till it becomes a whole or 2 compliment I'm not sure whats the easiest for the end user
21:55:55speachyany bounds on the fractional component?
21:56:23_bilgus32 bits
21:56:35speachyI meant number of digits
21:57:11_bilgusso far the users are 2-3 but I figure 6 is prob max
21:57:53speachyso you could return a floating point number as two integers, the second one in units of 1000000?
21:58:06_bilgusI'm not sure what the upper bound is I think they could mak their own parser otherwise
21:58:17speachy1/1000000 I mean
21:58:24_bilgusyeah thats the multiplied option
21:58:44_bilgusthats how I did in lua as well but its clunky
21:58:47speachyI mean this is about as simple as a generic option can get without just passing the buck
21:59:02speachyand making all arguments strings requiring the plugin to parse out
21:59:42_bilgusI think 32 bits wouuld a;;ow 1/1000000000
22:00
22:02:03speachyyeah
22:02:35speachyin all seriousness, I think making all arguments passed as strings makes a lot of sense.
22:02:51speachyreally simplifies the ABI
22:03:12_bilgusthey are but i'm making functions for the plugin lib
22:03:28speachymake 'em hex strings to make parsing possible without division
22:04:22speachysince only our newest targets have hardware division :)
22:06:31_bilgushuh?
22:07:17speachysince atoi() requires division by 10 with each digit
22:07:33_bilgusits multiply?
22:07:37speachybut parsing hex can just shift by 4 bits for each digit
22:08:24_bilgusoh you mean for decimals
22:08:26speachyyeah
22:09:08speachysorry if I'm stating the obvious
22:09:13speachybeen a long week
22:09:33_bilgusum no sounds interesting
22:10:50speachyand yeah I meant printing not parsing. sigh. I am more tired than I thought
22:11:47speachyparsing works with multiplies, printing needs divisions (for base-10)
22:15:44_bilgusI'd like to move a lot of the stuff in core to plugins
22:16:02_bilgusbookmarks, pitch_screen
22:16:13_bilguscouple others
22:16:41_bilgusI think the smallest is 6k largtes somewhere north of 15
22:17:24_bilgusI think parsing args makes sense for that
22:18:14speachyI can imagine HDD-based players seeing a perceitpble delay when trying to use those features
22:18:18speachyif they're moved to plugins I mean
22:18:37_bilgusI was even considering adding a core plugin buffer but it might be too much added complexity
22:18:40speachyflash-based would still have a delay, just might not be as noticable
22:19:31speachy...make the entire wps a plugin, huh.
22:19:40_bilgusfor as often as they are used I can't see keeping 50+k in the main binary
22:19:53_bilgusconsidered ity :)
22:22:13_bilgusfigure the added space is more runtime so whats the main use?
22:22:30_bilgusversus setting up stuff
22:22:33 Quit Moriar (Quit: Leaving.)
22:22:46speachythat would actually make it easier to fit on modern shovelware built on (eg) the rknano platforms
22:23:04speachyheck, even the main menu/UI could be yet another plugin
22:23:10_bilgusyep
22:23:11speachyplugins all around!
22:23:21_bilgusalready experimenting with that
22:23:39_bilgusvoices in plugins opened that
22:25:24_bilgusmy only concern was the devices that are F'd and will no longer be able to limp along but..
22:25:45speachy...but time is already taking care of those
22:39:28 Join Moriar [0] (~moriar@107-200-193-159.lightspeed.stlsmo.sbcglobal.net)
22:47:55***Saving seen data "./dancer.seen"
22:55:58 Quit Moriar (Ping timeout: 252 seconds)
22:56:58 Quit emacsomancer (Read error: Connection reset by peer)
22:57:24 Join emacsomancer [0] (~emacsoman@136.60.128.68)
23:00
23:01:15 Quit emacsomancer (Read error: Connection reset by peer)
23:01:44 Join emacsomancer [0] (~emacsoman@136.60.128.68)
23:08:33 Quit emacsomancer (Read error: Connection reset by peer)
23:14:24 Join emacsomancer [0] (~emacsoman@136.60.128.68)

Previous day | Next day