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 2021-04-02

00:03:28***Saving seen data "./dancer.seen"
00:08:35speachylet's see what this publish thing does.
00:09:56speachymaking the "public" wiki purely static will be _huge_.
00:14:56braewoodsinteresting concept
00:15:22braewoodsthough not sure how easy it is to use for a user
00:19:11 Quit ac_laptop (Ping timeout: 240 seconds)
00:25:21braewoodsspeachy: i noticed bilgus was using 'register' in a few prototype patches for the H10
00:25:35braewoodsspeachy: does that actually do anything? last i heard most compilers just ignore register now.
00:25:39 Join Strife89 [0] (
00:25:48braewoodssince they have enough resources to make decisions like that themselves
00:26:18braewoodsi've only really seen it in ancient source code like for X11
00:26:31speachyyeah, it's generally pointless
00:28:00braewoodsi saw a rare H120 on ebay recently
00:28:14braewoodsit's the early design instead of the black and silver
00:35:47_bilgusspeachy nice!
00:37:10_bilgusbraewoods, I didn't look at the asm but it probably doesn't it was already there though
00:37:37braewoods_bilgus: i see
00:37:49braewoodsi bought an F20 to attempt a repair on
00:37:56braewoodsi'll let you know if it works with rockbox or not
00:38:15braewoodsit should
00:38:34braewoodssad how crappy the samsung units turns out to be
00:38:42braewoodstoo unreliable for me to recommend to anyone
00:38:54braewoodsnot rockbox, just a general design flaw with them
00:39:19braewoodsthe HDDs have a tendency of disconnecting too easily
00:39:27braewoodsso yea
00:39:31braewoodsi don't want to fuck with that mess
00:48:20braewoods_bilgus: you can markoff the H120 and H320.
00:48:28braewoodsi confirmed those working during my bootloader work, remember?
00:48:35braewoodseven recently for the H120
00:52:24_bilgus300 == 320?
00:54:50_bilgusjohnb has a bunch of sansas
00:55:05_bilguslike the fuze v1 and clip IIRC
00:55:22_bilgusmaybe an e200
00:55:54_bilgusbtw clipv1 is next on the chopping block as far as low ram devices
00:56:03_bilgusit doesn't even have a sd card
01:19:03braewoods_bilgus: it's h320/h340
01:19:11braewoodssame unit, different sizes for HDDs
01:58:58__builtinspeachy: what's going on with the wiki?
01:59:26__builtinthe edits by BaseUserMapping_### seem to be making pages require registration to view
02:03:32***Saving seen data "./dancer.seen"
02:12:50braewoodsspeachy: thanks for the help debugging the gigabeat
03:39:54 Join lebellium [0] (
04:03:35***No seen item changed, no save performed.
05:23:57 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
06:03:37***No seen item changed, no save performed.
07:29:46speachy__builtin: uh.... I dunno...? Got a specific example?
07:37:48speachySo.. at 04:08 or so, a bunch of pages were changed by 'AdminUser'
07:41:08_bilgusspeachy have a look on the frontpage under: news
07:43:48 Join massiveH [0] (
07:45:03speachymain rockbox page?
07:48:40_bilgusspeachy that user changed the paypal links
07:49:05_bilgusthey go to ?336081fec2f640eb777ae7884b0b97dd' />
07:49:49_bilgusdonate button so its a hack
07:50:50speachyI don't see it on www
07:51:11_bilgusI diffed the two
07:51:22_bilgusline 123
07:54:21speachythat's part of the template
07:54:25speachythe paypal link I mean
07:56:02_bilgusmust have been indescriminately replacing links? seems the main page still goes to the pp link
07:57:12speachyI mean, that is pulled from a template and that template has not changed
07:57:55_bilgusoh is it just a random number?
07:58:37speachyI don't know where the 'validation_key' input in that paypal form is coming from but it's probably the template engine trying to prevent multiple submissions of a given form.
07:58:51speachybut the hosted_button_id is correct, and the URL is correct.
07:59:22speachywhen I try to generate a diff between the stuff that 'AdminUser' chagned I see nothing
07:59:56_bilgusthere is a bunch but the think that matters is at the end
08:00:23_bilguswhat I did is view the page view source and used an online diff
08:01:40 Join ac_laptop [0] (~ac_laptop@
08:01:43speachyuse the wiki page diff tool... and nothing. unless the wiki's differ is broken, which is always possible
08:01:55speachyyeah, that validation_key is random
08:02:14_bilgusok good I saw that was like that sneaky fucker
08:03:10speachybut that series of changes at 04:08 -> 04:10 is highly suspicious
08:03:39***No seen item changed, no save performed.
08:04:01speachybut other than triggering permission problems..
08:05:31speachythe displayed time is GMT. so it really happened around midnight local
08:07:35speachywhich was when I was mucking with things. But at no point did I ever muck with permissions
08:07:42 Quit ac_laptop (Ping timeout: 268 seconds)
08:09:05 Join ac_laptop [0] (~ac_laptop@
08:09:15_bilguswhen I was reading through the static wiki I was reading about the incident
08:09:46_bilgusthe 'incident' I guess someone hacked the wiki and wiped out most of it
08:11:47 Join ZincAlloy [0] (
08:11:52_bilguskinda cool just how much rb history is in that wiki
08:12:23_bilgusso maybe the conversion script screwed with permissions?
08:12:29_bilgusseems unlikely
08:13:02speachyI can't seem to recreate the initial permission problem. of course I did completely blow away foswiki's cache
08:13:50speachyI'm not sure where those supposed shadow edits came from, but for now I'm going to chalk this up to the wiki engine losing its mind again
08:14:36_bilgusgonna revert them to fix it or is there something you can do on the back?
08:15:30 Join MrZeus [0] (~MrZeus@2a02:c7f:a0aa:4400:48ab:1448:3526:df42)
08:16:11speachythere's nothing to revert, as far as I can tell.
08:17:52_bilgusoh you meant by unable to repro that it was already fixed then?
08:18:35speachyI touched all of the pages that were suppoedly changed (you still have RockboxUsbHandling locked btw0
08:18:58_bilgusoh yeah viewing the history
08:19:03speachybut there's no change between the versions that I can see
08:19:30speachyall I can think of is that something tried to sneak in while I was mucking wth the config.
08:19:37speachyor just random crawling traffic
08:19:48speachyand the special snowflake got all melty
08:21:26speachyone nice thing about this publish plugin is that it can suck in all external referencs
08:22:46speachyyeah, that was triggered by the export plugin
08:22:54speachywhich accesses everything as 'AdminUser'
08:23:23speachy... I think those pages all have graphs
08:26:38_bilgusah that sounds likely
08:28:26_bilguscan we pull in the template too or set up a template for it with our logo and such or is it stuck with black/white?
08:28:55speachyI'm trying to sort that out
08:29:02_bilgusah got ya
08:29:27speachycurrent WIP
08:29:38_bilgushow big is the static btw?
08:29:44_bilgusthe static wiki
08:29:48speachy31MB of html
08:30:24speachywith all resources (css,js,images,etc) pulled externally.
08:30:37_bilgusah so another metric tonne
08:30:58speachywon't be that bad
08:32:06_bilgusI'd like to be able to have a local version of that for bettwer searching
08:32:19speachyannoyingly it doesn't distinguish between "attachments" and "external resources"
08:32:56_bilgusthat page doesn't look too bad wish it had back buttons
08:35:31_bilgusdare I say it actuaslly reads better than with the live one
08:39:38speachyhow so?
08:39:51_bilgusjust because all the wiki junk is gone
08:40:56_bilgusI assume it still has the revision date at the bottom?
08:41:49_bilgusthat always got me trying to click a link and getting pulled into 15 seconds of backing out
08:43:01_bilgusOh I guess itdoesn't save the revision dae
08:43:13_bilguslooking at the static one from LN
08:43:39speachyit's all templates
08:43:57_bilgusI guess if you gen the whole thing then it doesn't matter
08:44:36speachythat's the plan but I'd like to preserve the revision info/dates etc
08:46:10speachyI still haven't figured out how it pulls in the rockbox template crap.
08:46:21 Quit massiveH (Quit: Leaving)
08:47:06_bilguswell it surely looks like the way forward in that form we might even be able to ge someone to archive it
08:47:30speachyoh, niice.
08:47:35speachyrefresh the X3 page
08:49:29speachythat's the default "plain" template
08:50:22_bilgusnot bad
08:59:55 Join dys [0] (
09:34:22 Quit pamaury (Ping timeout: 246 seconds)
09:36:05 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
10:00:33 Quit ac_laptop (Quit: WeeChat 3.0)
10:00:53 Join ac_laptop [0] (~ac_laptop@
10:03:42***Saving seen data "./dancer.seen"
10:04:48 Quit ac_laptop (Client Quit)
10:06:04 Join ac_laptop [0] (~ac_laptop@
10:50:04_bilgusWell adding my own rempa context was the easy part My loader to place the keymap in buffer will be a bit harder but probably not too bad compared to the plugin to create these key maps ( g#3274 )
10:50:05fs-bluebotGerrit review #3274 at : Add custom action mapping to core WIP by William Wilgus
10:51:48_bilgusI'm thinking first you choose a context then it gives you a list of actions then you choose button 1 and it asks you to press the button combo
10:51:50speachywhoops, wrong room.
10:52:14speachyI've cleaned out a ton of ancient twiki-related stuff out of the wiki and disabled some heavyweight compatibility plugins
10:52:24speachyso it should subjectively feel a bit zippier
10:52:29speachylet me know if something is broken
10:52:51speachy(a side effect is that I touched pretty much every file. so recentchanges is ... useless)
10:53:40_bilgusI'll take a stroll through hopefully since I have no particular destination I'll stumble onto a BIP
10:54:26speachyI also added to the forum plea for testing post
10:54:37_bilgusold bug
10:55:08speachy_bilgus: seems to apply to #3724 :D
10:58:54_bilgusI fell my idea fits that ethos but hope the implementation does too!
10:58:58_bilgusfeel lol
10:59:52_bilguslooking through the wiki I see lots of refs to the Archos :P
11:01:40_bilgusdon't know if these were broken before
11:02:16_bilguscolor units is the only group that has broken images
11:03:26speachyit's pulling from the h300 manual
11:03:39speachybeen broken for some time actually
11:03:48speachybut interestingly, it appears that the h300 manual is broken
11:04:18_bilgus same deal as the KNOWN_ISSUES
11:09:53_bilgusTORB TORB TORB TORB
11:10:17_bilgusforgot thats where that was ^ :p
11:11:10gevaertsThat's the most important page on the wiki! :)
11:11:41_bilgusI hope these guys come back around once they hit retirement age
11:12:23_bilgusjust in awe of what it all became even if it is a bit rusty
11:17:12fs-bluebotBuild Server message: New build round started. Revision 39db911164, 298 builds, 11 clients.
11:17:23speachythat should fix the h300 manual.
11:17:40speachy(and tonight, the missing screenshots will appear)
11:25:25speachyoh! the docs/MAINTAINERS is also _very_ outdated. I'm adding myself into a few places
11:27:35fs-bluebotBuild Server message: Build round completed after 623 seconds.
11:27:37fs-bluebotBuild Server message: Revision 39db911164 result: All green
11:29:08fs-bluebotBuild Server message: New build round started. Revision 5d292d9283, 298 builds, 11 clients.
11:40:13fs-bluebotBuild Server message: Build round completed after 664 seconds.
11:40:15fs-bluebotBuild Server message: Revision 5d292d9283 result: All green
11:40:22blbro[m]gevaerts: Don't forget GoldenQuotes :)
11:48:39gevaertsAh yes :)
11:59:34speachyfixing up piles of git links in the wiki now
12:03:43***Saving seen data "./dancer.seen"
12:11:33speachythere are about 100 links to
12:12:19speachy74, to be precise
12:14:07speachynot really sure how to go about fixing that up.
12:33:15 Quit emacsomancer (Read error: Connection reset by peer)
12:33:56 Join emacsomancer [0] (
12:36:25 Quit emacsomancer (Excess Flood)
12:36:58 Join emacsomancer [0] (
12:53:41 Quit ac_laptop (Ping timeout: 240 seconds)
12:57:13 Join ac_laptop [0] (~ac_laptop@
13:16:16 Quit ecs (Changing host)
13:16:16 Join ecs [0] (esawady@sourcehut/interns/ecs)
13:38:45 Quit kakaka (Ping timeout: 240 seconds)
13:52:07 Join kakaka [0] (~koniu@gateway/tor-sasl/koniu)
14:03:44***Saving seen data "./dancer.seen"
14:29:43_bilgusgevaerts, I can never rem if its you or saratoga that has all those players
14:31:54speachyI recall saratoga having a pretty large stash
14:33:03_bilguswell yours is probably getting that way lol
14:33:54speachyonly if you include broken Fuzes
14:34:10speachy(stupid ribbon cable)
14:35:41speachyfor some reason aliexpress thinks I'm russian now
14:39:06speachyhuh, xduoo has a new x2s model now.
14:39:40speachy... and the ChinaInc copy editing is in full force
14:44:48_bilgusOh I hope its powerful enough looks like the form factor is a bit bigger and thicker than a clip
14:45:24_bilgushaha yellow line lcd too
14:46:16speachyThe earlier X2 was a Rockhip RKNano, highly likely this is the same
14:47:13speachy(IIRC dual core cortex-M3, each core with 1MB RAM)
14:47:24_bilgusiirc someone started that port
14:47:46speachythough I think the RAM is fully accessible by both cores
14:47:58_bilgusmaybe it was just to find out it wasn't gonna happen idr
14:48:21speachynon-contiguous 2MB is really pushing it for a swcodec device
14:48:35speachyand since SOMEONE nuked our hwcodec stuff from orbit...
14:48:45_bilgusclip v1 does it
14:48:59_bilgusits a good thing!
14:50:11bertrikspeachy: often thinks I'm arabic
14:50:25speachyhmm. rockchip claims it only has 1MB total.
14:50:42_bilgusoh in that case nooope
14:51:23speachy64K PMU SRAM, 320K IRAM / 256K DRAM for core0, 128/256 for core1
14:51:36_bilgusi bet you could do it but itd wipe out the flash and storage in short order
14:51:43_bilgusand slow
14:51:54braewoodsevne if it has hardware codecs we'd need to know how to use it
14:51:58speachyit can execute from flash.
14:52:06braewoodsand it would heavily limit what you can do
14:53:14speachyI'd love to do a cortex-M port of rockbox but only the highest-end microcontrollers support external DRAM
14:53:41braewoodsi honestly think we're better off with devices that can do it all in software without blinking an eye
14:53:46braewoodsmore flexibility
14:53:49speachyand when you compare the cost of that vs a RPiZero...
14:54:29braewoodsi sometimes wonder if we could use one of those prebuilt cases and just cut out the holes we need for hardware buttons or so
14:54:41speachythere's a series of cortex-m-based players on the market now; a lot of clones of a hacky dsign someone did a while ago
14:54:42braewoodsmake our own rockbox in a sense
14:55:15speachyidentifiable from their very cube-like box, keypad, and display.
14:55:37braewoodsspeachy: all those generics that are unbranded?
14:55:47 Quit J_Darnley (Ping timeout: 252 seconds)
14:55:59braewoodsnot a good choice for building a stable port from since it's hard to know if the hardware will vary or what.
14:56:53braewoodswhat would one need to make their own rockbox unit?
14:57:02braewoodssome GPIO buttons
14:57:05braewoodsa type C port
14:57:32braewoodsand some kind of board to bring it all together
14:57:52speachyit's not "hard" but it takes upfront investment
14:58:11braewoodsa prototype case might be possible from these aluminum project boxes i see floating around
14:58:29braewoodswon't be as space efficient as a polished design
14:58:58braewoodsoh, and i guess we'd need a DAC and LCD of course
14:59:10braewoodsradio and RTC optional
14:59:15speachybut why bother, when an old ipod costs less and is more portable/robust? :)
14:59:39braewoodsand the popularity of the ipods even now...
14:59:56speachyapple did sell a bajillion of them.
14:59:58braewoodsyou can find replacement shells which aren't an option on any other line
15:00:07speachyand say what you will about apple, their industrial design is top-notch.
15:00:15braewoodsit was anyway
15:00:19braewoodstoday i'm not so sure
15:01:18speachyI have a couple Allwinner V3 dev boards, plus the pinecube
15:01:35 Join J_Darnley [0] (
15:01:44 Join chris_s [0] (
15:01:46speachyboth of which are good baselines for another hosted port.. or even native
15:02:09braewoodsspeachy: what do you think of dropping the code for half-finished ports that no one has moved forward with in over 5 years?
15:02:15speachybut as the conversation with pine64 demonstrated, it's the case tooling that's the real investment
15:02:20braewoodsi see a number of those in the current GIT.
15:02:31braewoodsmost of them have no development builds even
15:03:06speachyI'm generally in favor of nuking 'em, but there's not actually much code involved.
15:03:28braewoodsyea but i don't like keeping dead weight around
15:03:39 Join jdarnley [0] (
15:03:49braewoodsi'd like to clean up the codebase of old ports that people never finished
15:04:00braewoodsor stable ports that we can't verify as working anymore
15:04:05braewoodsbut that's for later
15:04:05speachyuntil it causes pain, it's not really an issue. unlike, say, the swcodec stuff that's been a thorn in our sides for a decade
15:04:19speachyhwcodec, yeah
15:06:11 Quit J_Darnley (Ping timeout: 246 seconds)
15:07:58chris_sDoes somebody know what this comment is supposed to mean?
15:08:05chris_sI'd like to enable the QS when a dir filter is enabled (like browsing playlists), but I' wondering whether I'm missing something as to the purpose of it being disabled there:
15:09:17speachychris_s: I think that's a leftover from the Archos Reorder which had F1/F2/F3 generic keys
15:09:43chris_sah, thanks
15:14:23speachythere's a bunch that look sorta the same as this; sometimes with a mono screen, sometimes with a larger one, sometimes with a scrollwheel like this, but always the same keypad
15:14:41speachyit's an STM32 attached to a high-end DAC
15:14:56speachy(and usually a pretty good analog path)
15:22:22braewoodsspeachy: i think if i find again where dialing back optimizations makes stuff work again, i'll look at stack size first.
15:22:33braewoodsit's a quick and easy thing to verify.
15:22:50speachybraewoods: yeah, but which stack? :D
15:23:02braewoodsthat's what the debug menu is for :D
15:23:12braewoodsif there's any stacks close to overflowing
15:23:19braewoodsthat's a sign it may be respnsible
15:23:41braewoodsthat last one was kinda weird
15:23:52braewoodswho would have expected an issue due to a thread stack overflow
15:24:31braewoodsi always wondered how linux gets away with giving each program 8MB of stack
15:24:40braewoodsconsidering how many processes you got running
15:24:44braewoodsnot counting threads
15:25:12braewoodsi have 300 processes running right now
15:25:15speachytwo reasons −− 1, a _lot_ more RAM to play with, and 2, it's virtual.
15:25:22speachywell, (3) −− overcommit by design
15:25:43braewoods2.4GB just for stack, if it was actually allocated fully
15:26:12braewoodsbut truth is most programs don't actually come close to using the full stack
15:26:27braewoodsthey're <2 MB or so
15:27:12braewoodsin any case i found myself using the stack for larger string constants to reduce costs associated with malloc
15:27:25braewoodsespecially if it's short lived
15:27:38braewoodsstrdup makes sense for copying to more permanent storage
15:27:56speachybraewoods: another DAP on the same basic platform:
15:28:05braewoodsstack is the fastest way to allocate short lived memory
15:28:09speachybig color screen but those same five buttons. :)
15:28:12braewoodsonly real drawback is limited size
15:28:52braewoodswouldn't it be hard to support?
15:29:29braewoodsin fact it looks like they just hollowed out an aluminium boc
15:30:16speachystm32f475, which has a Cortex-M7 @216MHz, 512K flash, 320K RAM.
15:30:35braewoodsso how could rockbox possibly work with that?
15:30:38braewoodsthe RAM is pitiful
15:31:12braewoodsmost targets have 32MB of RAM
15:31:54speachythat design has an FPGA onboard; usually it's to take a single I2S stream and split it to two indepedent DACs
15:32:04speachybecause crosstalk is teh devil
15:32:15speachy(and then plug a shared-ground headphone into it)
15:35:15speachybut that's the state of artisinal DAPs
15:35:36speachyfor the chinese hipster equivalents
15:38:25braewoodsspeachy: i thought we called them audiophiles now?
15:39:31speachywell, audiophones are the ones buying bluetooth tube amps
15:53:06 Quit chris_s (Quit: Ping timeout (120 seconds))
15:55:30speachybraewoods: incidently, I have access to a 3D printer now. so it's feasible for me to make a custom case for prototyping purposes.
16:03:46***Saving seen data "./dancer.seen"
16:05:31 Quit Acou_Bass (Quit: ZNC 1.8.2 -
16:11:21 Join johnb7 [0] (
16:13:09johnb7_bilgus As testing current builds, is there anything in particular I should look out for? I don't really use any plugin, just audio with different codec and radio.
16:13:27johnb7*As for testing
16:14:28speachyjohnb7: fundamentally, "it boots and plays music" is more data than we have
16:14:49speachy(we've had two targets broken in dev builds for ~6 months!)
16:15:28johnb7ok, tomorrow I can test sansa e200 (v1 and v2) and fuze v1.
16:16:06 Join amachronic [0] (5284b862@
16:18:43amachronicspeachy: I can say the M3K meets the boots and plays music criteria... it appears someone on the forums successfully booted the native port too
16:19:11 Quit Strife89 (Ping timeout: 240 seconds)
16:21:44speachyamachronic: the difference there being that it wasn't previously working and nobody's going to be nbesmirching our good name for breaking someting that used to work. :)
16:22:30amachronicyeah :P just noticed the forum thread says untested
16:23:08 Join Strife89 [0] (
16:27:59speachyI'll mostly likely be spending my (limited) hacking time over the weekend railing against the wiki and other infra crap
16:28:30speachyoh, there looks to be a NIB sansa c200 on ebay, so far $35 will take it home
16:34:20speachyI updated the forum post to indicate the M3K is partially working. (native port)
16:37:11speachypretty sure the DX50/DX90 are okay, or at least they're no worse than they used to be.
16:37:40 Quit johnb7 (Quit: Nettalk6 -
16:42:03 Quit MrZeus (Ping timeout: 246 seconds)
16:42:39 Join MrZeus [0] (~MrZeus@2a02:c7f:a0aa:4400:48ab:1448:3526:df42)
16:51:46braewoodsspeachy: NIB?
16:52:14speachyno idea if it's a v1 (PP) or v2 (AMS)
16:56:35braewoodsaren't all sansas using flash?
16:56:36gevaerts_bilgus: I have a decent number of players. Most of them haven't been powered on in years though, and their battery wasn't in the best of shapes ten years ago, so who knows...
17:01:30speachyyes, all flash (some with uSD slots)
17:08:09speachythink anyone will mind if I purge the WPS gallery stuff on the wiki? It's been completely irrelevant for some time.
17:10:53speachy26 wiki pages, and about 550MB worth of theme data.
17:59:42braewoodsis that even relevant to a wiki?
17:59:52braewoodsthemes should probably be a separate site
18:03:50***Saving seen data "./dancer.seen"
18:03:53bertrikspeachy: there is a crossfeed-algorithm in rockbox!
18:13:13 Join Acou_Bass [0] (
18:28:55 Quit amachronic (Quit: Connection closed)
18:33:36 Quit pamaury (Ping timeout: 265 seconds)
18:54:01 Quit lebellium (Quit: Leaving)
19:07:31 Quit ac_laptop (Ping timeout: 268 seconds)
19:12:36 Join ZincAlloy1 [0] (~Adium@2a02:8108:943f:d824:8db1:a467:4810:c681)
19:13:14 Quit ZincAlloy (Ping timeout: 265 seconds)
19:29:36_bilgusjohnb i'm ok with boots and plays a song
19:30:13speachybraewoods: it's the themesite-before-there-was-a-themesite
19:34:25 Join ac_laptop [0] (~ac_laptop@
19:35:42 Join massiveH [0] (
19:42:12speachyit's only useful in the archaelogical sense.
20:03:52***Saving seen data "./dancer.seen"
20:14:44 Quit ac_laptop (Ping timeout: 268 seconds)
20:18:23 Join ac_laptop [0] (~ac_laptop@
20:22:50 Quit ZincAlloy1 (Quit: Leaving.)
20:24:10 Join dconrad [0] (d026e411@
20:28:32fs-bluebotBuild Server message: New build round started. Revision e09df1ce5b, 298 builds, 10 clients.
20:37:34dconradamachronic: I was trying to install the m3k native port, and I think I screwed something up
20:37:50dconradI can't get it to boot or boot over usb anymore
20:38:37dconraddo you reckon you could try to help?
20:39:30fs-bluebotBuild Server message: Build round completed after 657 seconds.
20:39:32fs-bluebotBuild Server message: Revision e09df1ce5b result: All green
20:52:18 Quit MrZeus (Ping timeout: 246 seconds)
20:56:41 Quit Acou_Bass (Quit: ZNC 1.8.2 -
21:01:16 Join Acou_Bass [0] (
21:40:50 Quit dconrad (Quit: Ping timeout (120 seconds))
21:42:53 Join dconrad [0] (d026e411@
22:03:54***Saving seen data "./dancer.seen"
22:32:18speachyjust blocked four IP ranges that were pounding the site.
22:44:51 Join f1reflyylmao [0] (
22:45:41 Quit f1refly (Ping timeout: 258 seconds)
22:45:42 Nick f1reflyylmao is now known as f1refly (
22:50:22speachycloser to a dozen now. all using random user agents, all working in concert
22:50:36speachythis is why we can't have nice things, sigh.
23:07:34speachytwo dozen
23:11:15 Quit dconrad (Quit: Ping timeout (120 seconds))
23:17:09 Join dconrad [0] (d026e411@
23:21:36_bilguswtf like a bot crawler attack?
23:27:53speachyI dunno. I suspect this bot swarm is responsible for most of the traffic the site gets
23:28:08speachycloser to 3 dozen. including several /16s
23:32:58speachycurrently crawling the tracker. again.
23:33:23speachyhit the wiki three times and the tracker at least once before, irc logs 3 or 4 times this evening so far
23:34:00speachyone more /22
23:34:52speachya /24
23:35:07speachyeach indivual IP might only hit once every 5-10 minutes
23:35:36speachybut the b/w usage is to the point where folks are going to start noticing if I can't cut some of this back
23:36:49speachybumped two /23s into a /22...
23:39:10speachyalso banned a couple of IP ranges that were trying to break into the forum
23:40:43 Join saratoga [0] (
23:41:09speachy(no legit operation would be faking their user-agents like that..)
23:41:57saratogaI was going to say that the high performance/low power switch on the AK4376 seemed silly, but looking at the datasheet, putting the DAC in "high performance" mode increases power consumption by more than the *total* power consumption of the AMSv2 Sansas
23:42:33*speachy chuckles.
23:42:59saratogai think they may not have picked the best dac for a low power portable device
23:44:32speachyit's for the street cred!
23:49:56saratogaeven in low power mode the DAC will be quantization noise limited given 16 bit data, so actually we should probably use low power mode and get that nice battery bump
23:50:06saratogaat least until someone implements 24 bit output in the driver
23:51:21 Quit ac_laptop (Ping timeout: 260 seconds)
23:51:59speachyyeah, that's not going to happen anytime soon
23:52:12saratogaalso strange, high power mode doesn't actually increase the amplifier output at all, rather it just seems to disable some filter taps in the DSP core and some other digital stuff
23:53:26_bilgussaratoga can you get some of your players out for some dev build checks?
23:53:42saratogaactually no, it does change the headphone amp power consumption
23:53:50saratoga_bilgus, yes I can test some
23:53:54saratogajust the current build?
23:54:02_bilgusI've a post on the forum fell free to fill in
23:54:57_bilgusyeah we are finding after the toolchain bump some players aren't working, also the lcd rewrite, just want to be sure we aren't 'bricking' players in the next release
23:55:17_bilgusfeel* second time ive misspelled that today
23:56:08speachybasic smoke testing is enough
23:57:00saratogaFWIW, the Creative Zen Vision M port is barely a bootloader and hasn't been touched in over a decade, so no need to test that
23:57:29_bilgusbraewoods, and I were puzzling that yesterday

Previous day | Next day