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 2020-02-08

00:00:30globalcI wonder if there were any complaints claiming firmware rights infringements, or if there was exceptionally much traffic.. but probably its not something very hard to serve (just guessing, from the services I saw open via the rockbox site)
00:01:13bagderwe're just a bit bored to host and deal with the quirks when none of us have been involved since many years
00:01:33bagderfoswiki being the biggest annoyance
00:06:52globalcfixing software issues with it, or undoing things from malicious users?
00:07:46bagderblocking bots and crawlers that sink it due to how foswiki can spend an awful amount of CPU on some of the history links
00:08:05globalci see
00:08:37gevaertsAnd of course there's the entire gerrit being half broken situation
00:08:44bagderand a general feeling people who still care for the project are better of caring for the project...
00:12:05gevaertsYes, doing stuff for a project you thought you stepped away from years ago is a chore
00:16:18gevaertsbagder: will you also announce this on the dev mailing list?
00:16:28bagdersomeone should... =)
00:16:28gevaertsPeople might miss irc stuff
00:18:36gevaertsYes. I could send an email, but I think it would be better coming from one of you haxx guys
00:18:49bagderI agree
00:19:57***Saving seen data "./dancer.seen"
00:20:39bagderI'll get it done
00:51:53 Join beencubed [0] (~beencubed@
01:01:17__builtinI remember scorche saying that SMF was getting to be a hassle is not hosted by us
01:06:03gevaertsIt's probably a good idea to investigate that too, but that's different
01:33:56 Quit petur (Quit: Leaving)
02:02:28 Join sherbets [0] (~Thunderbi@
02:12:59 Join Timetravelerr [0] (
02:20:00***Saving seen data "./dancer.seen"
02:28:48 Join massiveH [0] (
02:31:20 Nick mixfix41_ is now known as mixfix41 (~homefame@unaffiliated/mixfix41)
03:31:36 Quit Timetravelerr (Remote host closed the connection)
03:33:10 Join Timetravelerr [0] (
03:53:15 Quit enigma9o7 (Quit: - Chat comfortably. Anywhere.)
04:06:29 Quit prof_wolfff (Ping timeout: 256 seconds)
04:20:04***Saving seen data "./dancer.seen"
04:33:26 Quit sherbets (Ping timeout: 260 seconds)
05:13:26 Quit TheSilentLink (Ping timeout: 240 seconds)
05:14:09 Join TheSilentLink [0] (~TheSilent@unaffiliated/thesilentlink)
05:32:59 Join sherbets [0] (~Thunderbi@
05:58:45 Join enigma9o7 [0] (
06:06:42 Part Timetravelerr
06:14:33 Quit sherbets (Ping timeout: 256 seconds)
06:20:07***Saving seen data "./dancer.seen"
06:41:26 Quit enigma9o7 (Ping timeout: 240 seconds)
07:09:28 Join enigma9o7 [0] (
07:18:26 Join ZincAlloy [0] (
07:42:41 Quit ZincAlloy (Quit: Leaving.)
07:49:43 Join ZincAlloy [0] (
07:54:17 Quit ZincAlloy (Ping timeout: 256 seconds)
08:20:09***Saving seen data "./dancer.seen"
09:07:40 Join Timetravelerr [0] (
09:07:46 Part Timetravelerr
10:05:07 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:d8e1:b6ce:9787:9198)
10:05:10 Join lebellium [0] (
10:20:11***Saving seen data "./dancer.seen"
10:42:51 Join prof_wolfff [0] (
11:57:20 Join dys [0] (
12:11:32 Join GofidLetterkerl [0] (
12:20:12***No seen item changed, no save performed.
12:22:56 Quit dys (Ping timeout: 265 seconds)
12:23:19 Join dys [0] (
13:38:06 Join speachy [0] (d102414d@
13:40:15speachywith regards to the service migration. The proposed plan IMO is far too ambitious; it isn't a "move off haxx-hosting" so much as "retool most of the infrastructure"
13:41:11speachyretooling is a major undertaking, and to be honest, there isn't sufficient user or developer interest to warrant that level of effort from the few folks that are still active.
13:42:47speachyplus I suspect I'm not the only one who finds the notion of basing our tooling completely around a propritary service (ie github) antithetical to the Free Software ideals that are surely still relevant.
13:44:18gevaertsIf people want something else, that's perfectly fine
13:44:20speachyOne thing I didn't see in the annoucement or the migration plan is some rough hosting requirements of the existing stuff. ie how much bandwidth is typically used, disk space, performance needs, etc.
13:46:11*gevaerts isn't convinced that redoing things in one of the "modern" code hosting systems with built-in CI support will actually be more work than migrating the existing infrastructure and fixing all the issues with it
13:46:17speachySo I think a much saner approach is to simply cut-n-paste the existing infrastructure to a different host, which will be a significant challenge on its own. Then once that dust has settled, we can figure out what needs changing.
13:46:27gevaertsI mean, our current gerrit setup is just broken
13:46:28*speachy agrees with that sentiment.
13:46:54speachyI think gerrit is far, far more powerful (and useful) than github's review and ci tooling.
13:47:04gevaertsMaybe, if it's maintained.
13:47:08speachybut yes, rockbox's instance needs love.
13:47:14speachya lot of love.
13:48:22speachyin a former life, I was a local admin for my employer's instance. I wasn't involved in setting the system up, but I did a lot of work to make my team's use of it integrate into CI and other workflow needs.
13:49:43speachybut I don't think the existing tooling is a significant barrier to entry. vs the "bare metal embedded" way of thinking. :)
13:52:52speachyt's one thing to toss gerrit, flyspray, heck, even the forums and wiki. But the existing CI/build system _works_, and quite well. Recreating to the same level of functionality will be a significant undertaking.
13:54:03gevaertsYes. The build system might be the one I'd keep a bit longer. I don't think it needs a lot of resources, so someone should be able to host it
13:55:29gevaertsThe wiki as I understand it is problematic. IIUC Zagor is has done some work on migrating it to github wiki, but IIUC (I understand a lot ;) ) that's "just" markdown in a git repo, so migrating that to something else should be easy
13:55:42speachyso in the end the real question is: how much of the old data in the existing systems is worth keeping.
13:56:00gevaertsforums are on scorche's server, so out of scope for now (although again, it's basically unmaintained)
13:56:25speachyyeah, the wiki is simple enough, but the web infrastructure ties a bunch of things together.
13:59:20speachyAt least there's plenty of time to figure out a sane plan.
13:59:54gevaertsI think the first step should be to find out who's willing to do what. If we have volunteers to maintain self-hosted stuff in the long term (and the money to pay for servers), then that's an option. If not, I think one of the code hosting services is the only option
14:00:10speachyFWIW I could probably host everything entirely; my main concern are the average bandwidth numbers.
14:00:37gevaertsbagder: do you have such numbers?
14:01:30globalci would be surprised if these were considerably high, considering that big files seem to be hosted somewhere else
14:01:44globalc(but better ask than being sorry, indeed)
14:02:19globalcstorage might also be an important resource, I could imagine
14:02:32speachyglobalc: only the binaries are likely to be significant.
14:03:12speachyglobalc: but I've learned to not understimate the bandwidth that the git repo and nightly build systems eat up
14:03:25globalcthe firmware pieces I fetched were hosted externally, might besides storage requirement also be due to keep out of copyright things
14:05:05speachyon my machine in the $work lab, I have ~4.2TB free, and it's on a municipal fiber link that routinely gets >500MB/s upstream bandwidth.
14:06:16speachythat lab routinely pushes multi-gigabyte files back and forth several times a day, though as I pull more of our infrastructure in-house that will go down a bit.
14:08:19speachyIMO most of the woes of rockbox's infrastructure are basically due to a lack of maintainance.
14:09:05globalcwow, nice
14:09:48globalcindeed, looks like the application maintennance would then more likely be the important ressource
14:11:39speachyI already self-host everyhing I use; git repos, mailing lists, dns, web, email, build/ci systems, bug tracking, and so on...
14:13:16speachyin part because I needed it before service providers for those things were terribly mature, but mainly because I'm a control freak who won't be beholden to any third party for stuff I consider critical.
14:13:25gevaertsIf you're already hosting the same or very similar systems for something else, that of course helps
14:17:01speachyso yes, I'm willing (and able) to host in a general sense. I've been using rockbox since the Archos Player days, and I want to see it go on.
14:18:31globalcthe current hoster said he tries to post a reminder to the mailinglist, that might then be a good place to ask for bandwidth traffic
14:20:14***Saving seen data "./dancer.seen"
14:24:36speachyyeah, bandwidth and storage needs. And a sense of how self-contained the haxx-hosted rockbox sutff is −− eg all on one dedicated server?
14:25:15gevaertsBy the way, bagder posted some more background on his blog at
14:27:15speachyyeah, I saw that
14:34:14zagor[m]Neither bandwidth nor cpu of is any sizable burden. The only burden that matters is maintenance.
14:34:47 Join petur [0] (~petur@rockbox/developer/petur)
14:35:53zagor[m]The site was designed to be reasonably self-maintaining, which is why it still works at all. But that only works for so long and the inevitable bit rot is becoming more and more visible.
14:36:10*speachy nods.
14:37:03speachyare there any available numbers on server resources? "not sizeable" is not something easily quantified. :)
14:37:34zagor[m]I'll get some numbers.
14:37:47speachythank you
15:23:31kugelwhat will happen to emails?
15:23:35 Join cc___ [0] (~ac@2001:910:1033:1:6a05:caff:fe1c:1627)
15:24:15speachyaccording to their transition proposal, they intend to keep hosting dns, email, and mailing lists indefinitely.
15:24:51speachybut IMO those should be eventually migrated too.
15:25:35zagor[m]Well, we *offer* to keep hosting them. Partly because the mailing lists are against and not
15:26:09speachyah, that's a big distinction.
15:31:04__builtinThe IRC logbot?
15:32:19zagor[m]I consider it part of the web site.
15:37:18__builtinI'm willing to provide some manpower in the migration, but I have comparatively limited compute resources
15:39:39GofidLetterkerlHi. Anyone here experienced with reading out firmware with an ISP device? I have one of those Sansa Sport players, and found something that looks like an ISP port on the front:
15:40:57GofidLetterkerlI tried reading it out using avrdude, but didn't get it to work (unsurprisingly, since it's no Atmel chip).
15:54:12 Join blbro[m] [0] (blbrostrat@gateway/shell/
15:58:32zagor[m]GofidLetterkerl That looks like a jtag port.
16:01:08 Quit cc___ (Ping timeout: 272 seconds)
16:12:57GofidLetterkerlSo, can I do anything useful with the jtag port, given a soldering iron and an ISP programmer?
16:13:18GofidLetterkerlI have the original 1.43 firmware from SanDisk's website; my player's firmware identifies as 1.18.
16:20:17***Saving seen data "./dancer.seen"
16:46:11speachyhmm, you'll need a JTAG tool capable of speaking to MIPS processors.
16:47:14speachybut given the very limited RAM on that SoC, I don't know it'll be terribly useful to hack around with.
16:47:35speachyIIRC something like 192KB.
16:54:55 Join tchan1 [0] (
16:55:15 Join tchan2 [0] (
16:55:32zagor[m]speachy: I added some numbers here:
16:57:55 Quit tchan (Ping timeout: 268 seconds)
16:59:10GofidLetterkerlspeachy: 192K is disappointingly little. Here I was, thinking I could somehow get Rockbox running on the Clip Sport. Looks like I have to live with the inferior original firmware :-(
16:59:38 Quit tchan1 (Ping timeout: 260 seconds)
17:01:56speachyawesome. other than downloads, those numbers are lower than I'd expected.
17:04:40speachyGofidLetterkerl: It's actually not _that_ bad; that ATJ part uses hwcodecs. We'd have to execute directly out of NAND, including resources, saving the RAM only for true state and (very) minimal buffers. If the ATJ part has the ability to transparently mmap the SD card into the CPU address space it makes things even smaller.
17:05:32speachybut .. it's still a pretty hairy port to try and undertake, and even in the best case it would require significantly culling the feature set.
17:07:26zagor[m]I'm sure people would be thrilled to add another ram-limited hwcodec target :)
17:08:06__builtinThat's the fun of it!
17:15:28 Quit mixfix41 (Quit: Lost terminal)
17:24:05 Quit tchan2 (Ping timeout: 272 seconds)
17:27:48bagderspeachy: I was deliberately vague and unspecific about where or how to transition stuff away from Haxx hosted services. I just wanted to mention that there are available (free) choices to go to. In the specific case gerrit vs github, my opinion was always that gerrit was overkill and quirky and the github model would lower the bar and encourage contributions better...
17:28:12bagderbut I don't care where or how it transitions
17:29:45bagder(I never contributed anything to rockbox via gerrit, I dropped out before that)
17:36:37 Join tchan2 [0] (
18:17:11*speachy personally quite likes gerrit. It's least crappy code review tool out there...
18:20:18***Saving seen data "./dancer.seen"
18:20:34speachygranted, it's decidedly less pleasant to administer and set up.
18:21:09 Part GofidLetterkerl
18:21:14bagderI won't push it, but adopting the github way of getting contributions really lowers the bar for contributors since "everyone" already knows how to do it and has accounts. So gerrit really needs to add significant benefits for that to be worth it. And if you ask me, it rarely does
18:21:44speachyit's a fair point.
18:25:34speachyas is the whole "let github/someone_else worry about keeping the infrastructure going" perspective.
18:55:11 Join S|h|a|w|n [0] (~shawn156@unaffiliated/shawn156)
19:03:47__builtinI should note that using gerrit does not preclude us from merging the occasional github PR
19:23:33***Invited to #openocd by PaulFertser!
19:24:02-->"help" received from PaulFertser (
19:43:43 Quit prof_wolfff (Ping timeout: 265 seconds)
19:45:23 Quit vup (Quit: vup)
19:50:44 Join vup [0] (~~~~@
19:56:15scorcheI love y'all dearly, but I'd like to also throw in my services as a part of this shutdown notice. There are 11 months remaining, so I don't think it'd be too onerous...
19:56:25scorcheI'll edit the wiki page accordingly
19:57:23 Join prof_wolfff [0] (~prof_wolf@
20:01:20 Quit tchan2 (Quit: WeeChat 2.6)
20:01:40 Join tchan [0] (
20:01:40 Quit tchan (Changing host)
20:01:40 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
20:17:04 Quit prof_wolfff (Ping timeout: 265 seconds)
20:20:19***Saving seen data "./dancer.seen"
20:25:38__builtinthat's forums and themes, right?
20:32:34 Join prof_wolfff [0] (
20:33:45 Quit michaelni (Ping timeout: 268 seconds)
20:33:55*speachy does not want to end up with the forums...
20:34:16speachyjust on principle.
20:42:08scorcheforums, themes, and translate
20:42:32scorcheforums really shouldn't live on in its current form, but I'll put details on the wiki page
20:47:05 Join michaelni [0] (
20:58:58speachyscorche: if you know them, please include the server resource usage (memory, disk, bandwidth) too
21:08:57 Quit ZincAlloy (Quit: Leaving.)
21:23:45 Join Timetravelerr [0] (
21:56:30 Quit Timetravelerr (Remote host closed the connection)
21:57:13 Join Soap_ [0] (~Soap@rockbox/staff/soap)
22:08:42 Join Timetravelerr [0] (
22:15:38 Quit Timetravelerr (Ping timeout: 260 seconds)
22:20:22***Saving seen data "./dancer.seen"
22:22:24 Join Timetravelerr [0] (
22:32:26 Quit Timetravelerr (Ping timeout: 260 seconds)
22:34:56 Join Timetravelerr [0] (
22:51:34 Quit Timetravelerr (Ping timeout: 260 seconds)
23:03:15 Join Timetravelerr [0] (
23:07:28 Quit Timetravelerr (Remote host closed the connection)
23:08:46 Join Rocy [0] (
23:09:16RocyHi can anyone provide me the channel to Rockbox community?
23:09:33RocyI cant seem to get to it even when I type it im
23:15:32 Part Rocy
23:18:03 Join Rocy [0] (
23:18:15Rocyhi i need help with something
23:18:46RocyWhenever I insert my disk it appears as if it was empty as if I never inserted a disk at all
23:18:54 Part bagder ("gone")
23:19:28RocyWhen it was previously working When I try to acces the disk is says please insert disk in
23:28:10 Quit Aldem (Read error: Connection reset by peer)
23:28:21 Join reductum [0] (
23:31:38 Join Aldem [0] (~Aldem@unaffiliated/aldem)
23:34:57 Quit Rocy (Ping timeout: 260 seconds)
23:45:37speachycontext? what do you mean by disk, insert into what?
23:46:57 Join Timetravelerr [0] (
23:55:29 Quit Timetravelerr (Ping timeout: 260 seconds)

Previous day | Next day