2023-07-29.log

- XYZ (QUIT: Remote host closed the connection) (~XYZ@78-80-25-18.customers.tmcz.cz)00:50
+ XYZ (~XYZ@78-80-25-18.customers.tmcz.cz)00:52
- jacobk (QUIT: Ping timeout: 264 seconds) (~quassel@47-186-126-199.dlls.tx.frontiernet.net)01:14
minutea311d can decode 1080p h265 in software easily, i just realized01:32
- Gooberpatrol66 (QUIT: Remote host closed the connection) (~Gooberpat@user/gooberpatrol66)01:44
+ Gooberpatrol66 (~Gooberpat@user/gooberpatrol66)01:45
- klardotsh (QUIT: Quit: zzz) (~klardotsh@98.97.112.82)01:53
- mtm (QUIT: Ping timeout: 250 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net)02:02
+ sl (~sl@contrib.inri.net)02:04
- mjw (QUIT: Ping timeout: 245 seconds) (~mjw@gnu.wildebeest.org)02:14
- nsc (QUIT: Ping timeout: 244 seconds) (~nicolas@51-49-142-46.pool.kielnet.net)03:28
- vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50)03:28
+ nsc (~nicolas@66-97-142-46.pool.kielnet.net)03:30
+ mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net)04:08
ex-parrotok I'm very happy, I've got vanilla bookworm running from LUKS encrypted nvme, xorg is working with glamor acceleration, poweroff works with the lpc module04:43
ex-parrotbattery measurement works04:44
ex-parroti3, firefox, glxgears all seem good04:44
ex-parrotnot sure to what extent xorg is "not accelerated" in this scenario but it seems pretty good so far04:48
joschex-parrot: what tweaks did you do to make xorg work?06:04
- ec0 (QUIT: Ping timeout: 260 seconds) (~ec0@vps-446f4f39.vps.ovh.ca)08:44
+ ec0 (~ec0@vps-446f4f39.vps.ovh.ca)08:44
ex-parrotjosch: just this config snip from https://community.mnt.re/t/classic-x11-with-config-file/1171 in bookworm10:16
ex-parrotplus I added "SWcursor" "on"10:16
ex-parrotto stop the cursor flickering10:16
ex-parrotxorg at least claims to be accelerated10:17
ex-parrotand I get refresh rate locked FPS in glxgears which is a good sign10:17
ex-parrotI can post some log snips tomorrow10:17
- GNUmoon (QUIT: Read error: Connection reset by peer) (~GNUmoon@gateway/tor-sasl/gnumoon)11:00
+ GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon)11:00
+ Boostisbetter (4a410829d7@irc.cheogram.com)13:46
+ eery (~eery@77.137.76.80)13:48
BoostisbetterWell folks, I can officially say that I am done even half supporting Purism. I don't think everyone working for them is bad. There are many really good developers there, but I am done supporting the company. Long live MNT. Thanks for the keep it real Lukas! 13:49
Boostisbetterfor reference watch the latest Louis Rossman video if you are wondering what sealed the deal for me. 13:51
eeryIf I have some nearly dead lifepo4 cells, can I apply, say, a 5-10 volt DC current to help bring them back?13:54
BoostisbetterI think I have read other people doing that. I remember thinking it was kind of a lot of voltage or maybe amps. 13:55
BoostisbetterSeems that it can work though. 13:55
- mtm (QUIT: Ping timeout: 245 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net)14:02
eeryYeah, it'll be a while before I can get proper tools or order new batteries, so I'm thinking I'll give it a shot14:04
Boostisbetterwell for your information, in case you didn't already know, the Reform is able to run without batteries and just plugged into AC mains. 14:06
eeryOh I know -- I just like working on battery, especially because all my mains connections are ungrounded, so I get tingly zaps when I touch the metal chassis while on AC power lol14:08
Boostisbetteruh what. How are all of your mains ungrounded? Are you a mad scientist? 🙂14:09
eeryThe sockets aren't ungrounded themselves, they're just type j plugs (with a grounding pin) while my adapter only has a type f plug with the grounding contact on the ring14:14
Boostisbetterahh ok, that makes sense. 14:15
BoostisbetterIs anybody else bridging to this IRC channel via cheogram? How do you get it to stop trying to connect when there is an error?14:40
+ mjw (~mjw@gnu.wildebeest.org)15:12
+ murdock (~murdock@dhcp-68-142-56-15.greenmountainaccess.net)15:43
+ mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net)16:09
- murdock (QUIT: Read error: Connection reset by peer) (~murdock@dhcp-68-142-56-15.greenmountainaccess.net)16:14
- unixpoet (QUIT: Quit: leaving) (~unixpoet@user/unixpoet)16:38
+ unixpoet (~unixpoet@user/unixpoet)16:38
BoostisbetterDoes the Reform have an end key?16:54
+ reform23287 (~gaston@81.173.205.18)17:13
reform23287q17:14
- reform23287 (QUIT: Remote host closed the connection) (~gaston@81.173.205.18)17:14
BoostisbetterYou just hold the hyper key or something to use it?17:20
BoostisbetterOr are you trying to quit the channel?17:20
minuteBoostisbetter: no end key yet, but i want to put home, end, pgup, pgdn on hyper and arrow keys on kbdv317:59
- q66[m] (QUIT: Remote host closed the connection) (~q66matrix@2001:470:69fc:105::3104)18:04
Boostisbetterminute: ok, so there is no end key on the current Reform keyboard? Can we add a software update to assign this key to a hyper key combo, perhaps? Or is that would you meant with the kbdv3?18:09
minuteBoostisbetter: yes, no problem to put it in kbdv2 firmware18:22
minutehttps://source.mnt.re/reform/reform/-/blob/master/reform2-keyboard-fw/matrix.h#L7018:24
minuteinstead of MATRIX_DEFAULT_ROW_6, you would write here the list of keys you want to have the combination of HYPER + the keys in the last row to yield18:24
minutethis is the default row of keys 6 https://source.mnt.re/reform/reform/-/blob/master/reform2-keyboard-fw/keyboard.h#L11818:25
minuteso you copy that row over, remove the backslashes and change the keys18:25
minutefor example, instead of HID_KEYBOARD_SC_PAGE_UP sth lke HID_KEYBOARD_SC_PAGE_HOME, etc18:26
Boostisbetterexactly. Very nice. 18:37
Boostisbetterminute: is there some way I could encourage more effort into the suspend issue on the Reform and Pocket Reform? For me this is an intrigal part of making the device useful, and I would happy to contribute some money to that end. 18:38
minuteBoostisbetter: on pocket reform, i will invest a significant effort to make it work, anyway18:39
minuteBoostisbetter: for reform, i have a new patch that disables all the code that turns off the display bridge, i.e. on suspend or when reinitializing the display. it might help with the issue that the display fails to come on sometimes after suspend.18:40
minuteBoostisbetter: i had to do this because on a311d the problem was even worse: when switching between HDMI and internal display, or when exiting sway, the picture on internal display would be lost. the sn65dsi86 bridge would then complain about DP link failure. i believe this also happened with imx8mq when waking up from suspend?18:42
minutewith the patch, the display is never lost on a311d. i cannot test suspend with a311d yet though, because there is an issue with pcie on that platform--the nvme can't wake up from suspend18:42
sigridthat bridge sure is sensitive18:42
minutesigrid: yes, but in my experience the best thing to do is to turn it on and never turn it off18:43
minuteit tends to re-sync to a signal anyway18:43
sigridyeah18:43
Boostisbetterminute: ok, it is good to see it is getting some love. That is an great to hear. I hope you are able to get some time on it to see if things can be improved. The solutions sounds it would work on the imx8 for sure. 18:43
minuteBoostisbetter: i think in the next week we'll have this patch in the main kernel packages, i want to test it on imx8mq myself before rolling it out18:44
Boostisbetterminute: sounds good. So far the work around you posted about on the forums has been working enough to be ok. 18:44
BoostisbetterSince the Reform handles resuming from a crash well enough, I'm not too worried. But it would be great to see an improvement there, as it really helps you to get right back into the workflow. 18:45
BoostisbetterAnyway thanks again! 18:45
minutesure thing!18:45
minuteand thanks for the offer of monetary support. i will come back to it if it necessity should arise18:46
minuteBoostisbetter: btw i found + watched louis' video18:46
Boostisbetterminute: it is a shame, because I think what Purism says they are trying to do is a good thing. It is just the means matter, and not just the end. 18:49
BoostisbetterMNT has always been super straight forward with me, and I really appreciate. 18:49
minuteBoostisbetter: have there been any reactions from the company? i was surprised about this very open attack18:50
minutebut we already knew from others about this questionable refund practices, i.e. josch18:50
minutei don't think it would be legal in germany18:50
Boostisbetterminute: Honestly, this sort of thing is what is said about them ALL the time.18:51
Boostisbetterbut they wording they use is correct, and as Louis says is sneaky and slimey. 18:51
minuteso far we have always refunded anyone who requested a refund, but i can imagine that purism was maybe in a bad financial situation and really needed that "free loan"18:51
Boostisbetterit means that it is legal. 18:51
minuteah ok18:51
Boostisbetterminute: if they would have just been up front about it, and just said that they did it to do everything they could to keep the dream a live, it would have pissed off a lot of people, but would have been the truth, and would have won over more than it lost. 18:52
Boostisbetterminute: the other things that insiders have talked about is quality control. What personally happened to me and with others are all also all the same. 18:53
minuteah. this i don't know anything about yet. did the phones have QA issues?18:54
minutebut i'm surprised the money stuff was still an issue because they did this like $10 million public funding thing where they gave people a credit note, right?18:55
Boostisbetterthey have done that like 3 times. I don't know how true those all are. 18:55
Boostisbetterthe phone has been super solid. Well built. It is the only thing that Purism has directly built. So the only thing that they were kind of directly in charge of. 18:56
BoostisbetterTheir laptops and all are from a 3rd party that they make contracts with. 18:56
BoostisbetterThat contractors screwed them over a lot, and Purism just passed it on to the customer. 18:56
Boostisbetterthat is the kind of thing I mean. 18:56
BoostisbetterI mean all you see is bad press about Purism yet they are doing well and selling out of products. It is just crazy to me. 18:57
sknebelHQ in SF and >30 employees can go through millions quite easily, even before you go to the challenges of making hardware18:57
unixpoetI had similar experiences with pine64 support, they *really* do not like to do refunds or exchanges; I had to threatem them with a chargeback to get them to respond18:58
Boostisbetteryeah, I want to believe them, and i want to support them. But it is these things that are not getting addressed that just makes that almost impossible. 18:59
BoostisbetterThe L5 is REALLY a good product that the devs have made a lot of progress on. 18:59
BoostisbetterBut if I can't trust the company what does all of that matter?18:59
Boostisbettersknebel: the thing that kind of gets me as well, is that it is a US company but I think less than 2% of the company are actually Americans. Things like customer support, tech support, are all not Americans. 19:00
Boostisbettersknebel: then they are all like this is a US made phone. 19:00
BoostisbetterIf they are racking in millions why not pay Americans to do the menial jobs of tech and customer support? 19:00
BoostisbetterThe whole thing just smells real fishy. 19:00
Boostisbetterand don't get me wrong, I don't care about companies hiring only from the company they belong to. I'm just saying that if you are using that as marketing and a reason to buy from them, it should matter then. 19:01
Boostisbetterthe devs for Purism are awesome. 19:01
Boostisbetteranyway, back to MNT, and just desperately waiting for the Pocket. 19:02
Boostisbetterhahahaha19:02
unixpoetimo steam deck > pocket19:02
unixpoetunless you've got physical cartridges19:02
BoostisbetterI'm talking about the MNT Pocket Reform. Not the Analogue Pocket. 19:04
BoostisbetterI love the Steam Deck though. But it is great for playing games, and not doing work. 19:04
BoostisbetterAlso even though it is portable it is huge. 19:05
minuteBoostisbetter: ahh i see, thanks for clarifying about the QA19:05
unixpoetoh, oops, sorry :P too many devices19:05
minuteyeah, my partner has the steam deck now, it's quite remarkable. i'm also thankful to valve for putting all that effort into linux19:06
minutei think analogue (pocket) has some really strong design / detail orientation, but i haven't seen one IRL19:06
Boostisbetterminute: heck yes. Things run so well on it, and the average user will have no idea it is Linux. 19:06
Boostisbetterminute: I have one. It is just a Game boy of various models recreation. 19:07
minuteabout pocket reform, we are currently working on finishing the aluminum chassis, i.e. a rev 2 with that with reduced weight19:07
BoostisbetterNot a laptop or desktop in anyway. I have one. 19:07
Boostisbetterminute: awesome!!19:07
minutethen i will do some spikes about power saving and final or pre-final rev of motherboard19:10
- GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon)19:11
Boostisbetterminute: are there still a lot of things that can improve power savings, or are most things already dailed in?19:12
+ GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon)19:12
minuteBoostisbetter: currently the 2 rp2040s are always on, i think this can be improved19:12
minuteBoostisbetter: and the charging is a bit slower than i would like (compared to reform), i think it can be improved, currently it is the default setting of the charger IC.19:13
Boostisbetterminute: funny, I was going to ask if it would be possible to slow charging on the Reform. In an effort to extend their life cycles. 19:14
Boostisbetterminute: I am under the belief that the slower you charge the less heat you create and the easier on the cells charging is. That probably is a gross over simplification of the whole process. 19:14
minutesooo the lifepo4 cells in reform are made for high current19:15
minuteliion cells usually are not that OK with a lot of current19:15
Boostisbetterahh so it is all good basically. 19:16
+ jacobk (~quassel@2603-8080-b200-7b02-77cb-6304-f9db-dda1.res6.spectrum.com)20:42
sigridminute: is the middle external usb port J17 or some other one?20:52
minutesigrid: let me double check20:52
minutesigrid: middle is J620:53
minutesigrid: J17 is next to hdmi20:53
sigridok, thanks. we are having some weird issues with usb happening in 9front20:53
sigridwhere connecting a mouse to one port, disconnecting it, then connecting to a *different* port breaks internal usb devices somehow20:54
minuteJ6 and J5 go through USB hub, J17 is direct20:54
sigridthey receive usb transaction errors20:54
sigridyeah that's what I wanted to find out, for cinap_lenrek it seems to only be the case with middle port somehow20:54
minutehmmmmm maybe some EMI / voltage spike issue20:55
minutedoes the power go out for the nonfunctional devices?20:55
sigridit doesn't seem so, so then it would make sense20:55
minutehmm20:55
minutewhat if one resets the USB hub (via the gpio)?20:56
sigridcinap is going to see hub status flags, maybe there is something20:56
sigridwe only reset usb hub once during init iirc20:56
- eery (QUIT: Remote host closed the connection) (~eery@77.137.76.80)20:56
minuteok20:57
sigridthanks :)20:58
sigridI remember that internal usb devices sometimes (very rarely) would fall of in linux20:58
sigridbut here we have a very easy repro, for some reason20:58
- jacobk (QUIT: Read error: Connection reset by peer) (~quassel@2603-8080-b200-7b02-77cb-6304-f9db-dda1.res6.spectrum.com)21:09
cinap_lenrekfor me, its just the *MIDDLE* usb port21:40
cinap_lenrekthat when i plug in mouse, the front falls off on the trackball and reform keyboard21:40
cinap_lenreki added some debug stuff to our hub driver21:41
cinap_lenrekwhen this happens theres no port or hub status change on the trackball/keyboard hub/ports21:41
cinap_lenrekand everything works fine when i use a powered usb hub on that port21:41
minuteok, sounds like overcurrent event maybe21:42
cinap_lenrekyep, but theres no overcurrent event21:42
cinap_lenrekthats the strange thing :(21:42
minuteah, you can see that?21:42
cinap_lenrekbtw, the newly plugged in device works fine, regardless21:42
minuteoh ah21:42
cinap_lenrekminute: yes, the overcurrent and power good is part of hub status field21:42
cinap_lenrekas well as the port status21:42
minuteis that a special mouse? or happens with other devices as well?21:43
cinap_lenrekno bits change at all in the affected ports21:43
cinap_lenrekgood idea21:43
cinap_lenreki should try it21:43
sigridI tried the same with a usb flash drive and it did not repro21:43
cinap_lenreksame here21:45
cinap_lenrekworks fine with usb sdcard adapter21:45
minutespecial mouse21:46
minutewhat's the brand and model?21:46
cinap_lenrekasus21:46
cinap_lenrekits the tinies mouse ever :D21:46
minuteah, laptop mouse?21:46
cinap_lenrekthe crazy thing is21:46
cinap_lenrekit works fine on the other port21:46
cinap_lenreklike the one to the left21:46
minuteon both other ports?21:46
cinap_lenrekwhich is port2 on the hub21:46
cinap_lenrekmiddle one is port1 on the hub21:46
cinap_lenrekand the right one is a completely different hub21:47
minuteok hm21:47
cinap_lenrekso ????21:47
minutepuzzling21:47
cinap_lenrekmaybe its a quirk of the hub chip?21:47
cinap_lenrekman, i wonder if theres like usb compliance testing equipment for this21:47
cinap_lenrekhold on21:48
minutethere is21:48
cinap_lenrekso all i'd need is to source the 5v power lines from somewhere else no?21:48
cinap_lenrekand just connect the data lines21:48
joschex-parrot: ah indeed, we are installing that file into /usr/share/X11/xorg.conf.d with the reform-tools package. Do you have that installed?21:49
sigridthe hub itself can't really be controlled from the soc, beyond resetting via gpio, right?21:49
sigridlike, the i2c(?) of the hub isn't even wired anywhere iirc21:49
minuteno. i mean maybe the overcurrent stuff is not reported correctly or something21:49
minuteon mb2.5 i use a different (a bit more primitive) hub and got rid of this functionality...21:50
minuteU15 and U16 are the chips that can cut the power to those ports21:53
Boostisbetterhas anyone noticed the mouse no longer reall working in DosBox with the Reform?21:56
Boostisbetternever mind, I think I was sitting on a bunch of updates and hadn't rebooted. 21:57
BoostisbetterAfter a reboot the mouse is work normal again in DosBox21:57
+ jacobk (~quassel@47-186-126-199.dlls.tx.frontiernet.net)22:00
Boostisbetterminute: as for suspend and the Reform it would be good if we could power down more on the mainboard to net bigger power savings. This is not something all that crucial of course. Suspend stability is far more important. Just throwing it out there. 22:11
cinap_lenrekok22:12
cinap_lenrekcan confirm22:12
cinap_lenreki can control the power on the ports22:12
cinap_lenreksigrid: echo portpower 6 2 off > /dev/usbhubctl22:12
cinap_lenrek(for left port)22:12
cinap_lenreksigrid: echo portpower 6 1 off > /dev/usbhubctl22:13
cinap_lenrekfor middle port22:13
sigridis usbd supposed to do that when device detaches?22:13
cinap_lenrekno22:13
cinap_lenrekit doesnt do power cycling22:13
cinap_lenrekbut it kind of could22:13
cinap_lenrekthe issue is, usbd doesnt see anything wrong with the ports22:13
cinap_lenrekit doesnt see any port status change22:14
cinap_lenrekand the driver just gets transaction error and dies22:14
cinap_lenrekwell22:14
cinap_lenrekthe portreset is already handled by usbd22:14
cinap_lenrekwhen you write "reset" to the deive ctl22:15
cinap_lenrekit just sets the "reset me" flag22:15
cinap_lenrekand usbd polls it22:15
cinap_lenreksee portwantsreset()22:15
cinap_lenrekwe could have a portwantspowercycle() :D22:15
cinap_lenrektho portwantsapony() is a better name22:15
sigridheh22:17
sigrid"Device Status Register This register indicates the status of the device controller with respect to USB-related events."22:29
sigrid(C70Ch)22:30
sigridcinap_lenrek: was it that one you checked?22:30
sigridoh nvm, this is for device not host22:31
sigridPHY_CTL3_ADDR has some interesting bits22:33
sigridbut I have no clue in all that22:34
cinap_lenrekno22:35
cinap_lenreki did a standard Get Hub Status Request on the hub22:35
sigridI see22:36
cinap_lenrekas well as a Get Port Status Request and just print whenever theres a change on any port22:36
cinap_lenrekdidnt do anything on the rootports22:36
joschI watched the video you mentioned. I had the same experience except i went a bit further and complained about purism with the state attourney of california. That worked. Now purism tells me they cannot refund until my complained with the state attourney about my non-existing refund has been processed. Classic.23:18
joschBoostisbetter: ^23:18
Boostisbetterjosch: do you have a Purism product? 23:32
joschof course not23:32
joschi requested a refund23:33
BoostisbetterWas it for the L5? 23:33
joschyes23:33
BoostisbetterAnd that is good. At least you have the reliable Reform right? 23:33
joschi waited so long that since i ordered, i got married, changed my name, moved countries twice, got a phd and got my first child -- this is too long of a wait23:34
joschyes23:34
joschi cannot praise mnt enough23:34
BoostisbetterAbsolutely! 23:34
joschawsome company and awesome product23:35
joschwill defintely buy again23:35
BoostisbetterYep going to be even better with the server and pocket coming down the pipe. 23:36
+ klardotsh (~klardotsh@98.97.35.74)23:53
+ vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50)23:56

Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!