Wayland/sway binary package available
Hello Matthieu, congrats to all of you that is making Wayland true on OpenBSD.. -- Nowarez Market Dec 5, 2023 08:45:23 Matthieu Herrb : > In the latest amd64 packages snapshot there is now a #sway package (other > arches will follow). > > Installing it allows you to try #Wayland on OpenBSD.
Re: Microphone on Thinkpad X1 carbon 7th Gen
On 04/12/2023 23:12, Carsten Reith wrote: Corl3ss writes: Hi everyone, Everything but the microphone is working well on my Thinkpad X1 carbon 7th Gen (20QD). I have reviewed [1] and [2] without solution. Below some useful data. Have you enabled the appropriate sysctl ? (see: https://www.openbsd.org/faq/faq13.html#recordaudio) # sysctl kern.audio.record=1 # echo kern.audio.record=1 >> /etc/sysctl.conf Thanks Carsten for your answer ! Yes, sure (and sorry I did not mention that on my first message). Corl3ss
Re: Microphone on Thinkpad X1 carbon 7th Gen
Corl3ss wrote (2023-12-05 09:18 CET): > > > On 04/12/2023 23:12, Carsten Reith wrote: > > Corl3ss writes: > > > > > Hi everyone, > > > > > > > > > Everything but the microphone is working well on my Thinkpad X1 carbon > > > 7th Gen > > > (20QD). > > > > > > I have reviewed [1] and [2] without solution. > > > Below some useful data. > > > > > > Have you enabled the appropriate sysctl ? (see: > > https://www.openbsd.org/faq/faq13.html#recordaudio) > > > > # sysctl kern.audio.record=1 > > # echo kern.audio.record=1 >> /etc/sysctl.conf > > > Thanks Carsten for your answer ! > > Yes, sure (and sorry I did not mention that on my first message). Here is a list of platforms who's microphones won't work on OpenBSD: https://thesofproject.github.io/latest/platforms/index.html#supported-platforms Microphones in many modern Laptops don't work. Thanks to intel smart sound technology (Intel SST), microphones now need a firmware (https://github.com/thesofproject/sof) and nobody has written a driver for OpenBSD yet. You can use a headset on the headset port and the microphone there should work. It's only the builtin mic that won't do anything (even though all toggles in mixerctl / sndioctl appear operational). Fun part reg SOF: it doesn't build with clang, see FAQ: https://thesofproject.github.io/latest/introduction/index.html So - I don't think we'll see mic support for these platforms anytime soon. Best Regards, Stefan
Re: Microphone on Thinkpad X1 carbon 7th Gen
On 05/12/2023 10:11, Stefan Hagen wrote: Corl3ss wrote (2023-12-05 09:18 CET): On 04/12/2023 23:12, Carsten Reith wrote: Corl3ss writes: Hi everyone, Everything but the microphone is working well on my Thinkpad X1 carbon 7th Gen (20QD). I have reviewed [1] and [2] without solution. Below some useful data. Have you enabled the appropriate sysctl ? (see: https://www.openbsd.org/faq/faq13.html#recordaudio) # sysctl kern.audio.record=1 # echo kern.audio.record=1 >> /etc/sysctl.conf Thanks Carsten for your answer ! Yes, sure (and sorry I did not mention that on my first message). Here is a list of platforms who's microphones won't work on OpenBSD: https://thesofproject.github.io/latest/platforms/index.html#supported-platforms Microphones in many modern Laptops don't work. Thanks to intel smart sound technology (Intel SST), microphones now need a firmware (https://github.com/thesofproject/sof) and nobody has written a driver for OpenBSD yet. You can use a headset on the headset port and the microphone there should work. It's only the builtin mic that won't do anything (even though all toggles in mixerctl / sndioctl appear operational). Fun part reg SOF: it doesn't build with clang, see FAQ: https://thesofproject.github.io/latest/introduction/index.html So - I don't think we'll see mic support for these platforms anytime soon. Best Regards, Stefan Thanks a lot for your precise answer Stefan. Very kind of you. It does not seems to work neither with the headset port but will give it another try :) Thanks again ! Have a nice day Stefan, have a nice day all ! Corl3ss
Re: FAT names exceeding spec length
On Mon, Dec 4, 2023 at 12:03 PM Michael Hekeler wrote: > > > > To be honest I don't understand the problem you described. > > > > It is simple, when you come from Android (tested Android 11 tablet) with > > file names exceeding the FAT spec > > these are cut to 8.3 format in OpenBSD. > > > You mean android allows to create filenames >255 on FAT32? > Then you should report this non-compliance on android > (https://learn.microsoft.com/en-us/windows/win32/fileio/filesystem-functionality-comparison#limits) > That table is pretty vague, maybe there's a byte-vs-character bug somewhere in the chain. If "255 Unicode characters" is taken to be 255 codepoints, how are they encoded? Digging around a little more shows that they probably mean "255 UCS-2 words". Wonder if OP is actually seeing more than 255 unique "User-perceived characters" or just more than 255 bytes?
Re: FAT names exceeding spec length
Thanks for the reply. Absolutely perceived, I haven't counted them. But I counted the 8.3 format for true: all the info contained in my ebook title was lost (but experimenting it is no problem). Now I checked better. Indeed it is Opera web browser that is able to save on Android 11 exceeding the name limits: Tcl_Tk 8.5 Programming Cookbook -- Bert Wheeler -- 2011 -- Packt -- 9781849512992 -- 9ed273d2c640e4ae4761242a2c28d31c -- Capitan Cloud Ebook Repo.pdf Android as per your guessing is enforcing the name limits here and there, "enough" well. However, this the situation. Cutting to 8.3 format or being more conservative? That could be the question for you. But indeed I leave the subect to you now. == Nowarez Market Dec 5, 2023 12:29:30 Anders Andersson : > Wonder if OP is actually seeing more than 255 unique "User-perceived > characters" or just more than 255 bytes?
Re: FAT names exceeding spec length
On Tue, Dec 5, 2023 at 2:01 PM Nowarez Market wrote: > > Thanks for the reply. > > Absolutely perceived, I haven't counted them. But I counted the 8.3 format > for true: > all the info contained in my ebook title was lost (but experimenting it is no > problem). > > Now I checked better. Indeed it is Opera web browser that is able to save on > Android 11 exceeding > the name limits: > Tcl_Tk 8.5 Programming Cookbook -- Bert Wheeler -- 2011 -- Packt -- > 9781849512992 -- 9ed273d2c640e4ae4761242a2c28d31c -- Capitan Cloud Ebook > Repo.pdf $ echo "Tcl_Tk 8.5 Programming Cookbook -- Bert Wheeler -- 2011 -- Packt -- 9781849512992 -- 9ed273d2c640e4ae4761242a2c28d31c -- Capitan Cloud Ebook Repo.pdf" | wc -c -m 150 150 150 << 255 > > Android as per your guessing is enforcing the name limits here and there, > "enough" well. > > However, this the situation. Cutting to 8.3 format or being more conservative? > That could be the question for you. But indeed I leave the subect to you now. > > > == Nowarez Market > > Dec 5, 2023 12:29:30 Anders Andersson : > > > Wonder if OP is actually seeing more than 255 unique "User-perceived > > characters" or just more than 255 bytes? >
NFS Server performance
Using an OBSD 7.4 VM on VMware as an NFS server on HOST02. It is primarily used to store VMWare VM backups from HOST01, so VMWare is the NFS client. I'm seeing transfers of about 1.2 MB/s. SCP from HOST01 to OBSD VM (same filesystem) copies at 110 MB/s. Iperf3 from a VM on HOST01 to OBSD on HOST02 gives me 900+ mbps. OBSD is a stock install running -stable. NFS is using v3 (according to VMWare) and using TCP During the NFS transfer the RECV-Q on the OBSD interface runs either 64000+ or 0. I tried both em and vmx interface types. /etc/rc.conf.local: mountd_flags="" # for normal use: "" nfsd_flags="-tun 4" # Crank the 4 for a busy NFS fileserver ntpd_flags="" # enabled during install portmap_flags=""# for normal use: "" Any clues on where to look to (greatly) improve NFS performance would be appreciated. -Steve S.
Re: NFS Server performance
On Tue, Dec 05, 2023 at 02:06:44PM +, Steven Surdock wrote: > Using an OBSD 7.4 VM on VMware as an NFS server on HOST02. It is primarily > used to store VMWare VM backups from HOST01, so VMWare is the NFS client. > I'm seeing transfers of about 1.2 MB/s. > > SCP from HOST01 to OBSD VM (same filesystem) copies at 110 MB/s. > Iperf3 from a VM on HOST01 to OBSD on HOST02 gives me 900+ mbps. > OBSD is a stock install running -stable. > NFS is using v3 (according to VMWare) and using TCP > During the NFS transfer the RECV-Q on the OBSD interface runs either 64000+ > or 0. > I tried both em and vmx interface types. > > /etc/rc.conf.local: > mountd_flags="" # for normal use: "" > nfsd_flags="-tun 4" # Crank the 4 for a busy NFS fileserver > ntpd_flags="" # enabled during install > portmap_flags=""# for normal use: "" > > Any clues on where to look to (greatly) improve NFS performance would be > appreciated. Increasing write size, read size and the read-ahead count on the client has helped me. E.g., on the client's fstab: 10.17.18.10:/shared/stuff /nfs/stuff nfs rw,nodev,nosuid,intr,tcp,bg,noatime,-a=4,-r=32768,-w=32768 0 0 Cheers Zé > > -Steve S. > --
Re: FAT names exceeding spec length
:-) so what is techinal explanation of the happening and the cut off to 8.3 ? -- Nowarez Market Dec 5, 2023 14:48:49 Anders Andersson : > $ echo "Tcl_Tk 8.5 Programming Cookbook -- Bert Wheeler -- 2011 -- > Packt -- 9781849512992 -- 9ed273d2c640e4ae4761242a2c28d31c -- Capitan > Cloud Ebook Repo.pdf" | wc -c -m > 150 150 > > 150 << 255
Re: FAT names exceeding spec length
You are fantastic. Next thread I will take you also the 8.3 filename examples with screenshots of the last three chars. ( don't stop to report guys! ) -- Nowarez Market Dec 5, 2023 17:16:32 Amelia A Lewis : > Have you, perhaps, misdiagnosed > the problem? -- Nowarez Market
Re: FAT names exceeding spec length
On Tue, 5 Dec 2023 16:46:38 +0100 (GMT+01:00), Nowarez Market wrote: > so what is techinal explanation of the happening and the cut off to 8.3 ? Well, for the given title, 8.3 yields: "Tcl_Tk 8.5 P", but I imagine, for this case, that what it might do is see an eight character filename and a dot, and expect that it will see an extension after that. Since it's got not only eight characters and the start of an extension, it's got another dot way down the way that has a recognizable extension, and in the middle it's got loads of whitespace. You haven't shown what the 8.3 result of shortening is (what's the extension? pdf or 5\ P or 5pr or something else?) Does the eight character name include whitespace? (since there's one at position seven in the input filename). You've presented your conclusion, that android can handle filenames longer than 255 characters, but openbsd can't, and you've given an example of an input filename, but not of the resultant 8.3 representation, and only one example. Have you, perhaps, misdiagnosed the problem? Amy! -- Amelia A. Lewisamyzing {at} talsever.org "Oh, fuck! You did it just like I told you to!" (The manager's lament) (also the programmer's lament, directed toward a compiler)
Re: FAT names exceeding spec length
People have already explained to you that 8 + 1 + 3 is less than 255. Can you kindly shut the fuck up already? On Dec 05 13:59:11, my2...@has.im wrote: > Thanks for the reply. > > Absolutely perceived, I haven't counted them. But I counted the 8.3 format > for true: > all the info contained in my ebook title was lost (but experimenting it is no > problem). > > Now I checked better. Indeed it is Opera web browser that is able to save on > Android 11 exceeding > the name limits: > Tcl_Tk 8.5 Programming Cookbook -- Bert Wheeler -- 2011 -- Packt -- > 9781849512992 -- 9ed273d2c640e4ae4761242a2c28d31c -- Capitan Cloud Ebook > Repo.pdf > > Android as per your guessing is enforcing the name limits here and there, > "enough" well. > > However, this the situation. Cutting to 8.3 format or being more conservative? > That could be the question for you. But indeed I leave the subect to you now. > > > == Nowarez Market > > Dec 5, 2023 12:29:30 Anders Andersson : > > > Wonder if OP is actually seeing more than 255 unique "User-perceived > > characters" or just more than 255 bytes? > >
Re: FAT names exceeding spec length
Hello, Have you tried mounting with the -l flag? Otherwise I'd recommend using ext2fs instead of FAT32. Android will handle it natively, and OpenBSD can mount that read/write (on a sdXi mountpoint, like FAT32). Cheers, Noth On 05/12/2023 16:46, Nowarez Market wrote: :-) so what is techinal explanation of the happening and the cut off to 8.3 ? -- Nowarez Market Dec 5, 2023 14:48:49 Anders Andersson : $ echo "Tcl_Tk 8.5 Programming Cookbook -- Bert Wheeler -- 2011 -- Packt -- 9781849512992 -- 9ed273d2c640e4ae4761242a2c28d31c -- Capitan Cloud Ebook Repo.pdf" | wc -c -m 150 150 150 << 255
Re: cumbersome mtree (OT!)
For who is wondering (and not) I come from publishing RADXIDE ver 1.0.5 (tcl-tk, MIT license) I could appreciate anyone involvement or feedback. https://radxide.com https://github.com/par7133/RADXIDE Thnx! == Nowarez Market Nowarez Market : >I come from a couple of long white nights and indeed this night was >the longest one (for who is interested I'm playing on a barely simple >ide that at a certain point in time hopefully should enforce RAD >amenities, good practices, etc. I'm working in tcltk). > >Let me reconnect with my body and I will give a try to your suggestion.
Re: FAT names exceeding spec length
Received, thanks. I'm used to hack on three devices one OpenBSD, one Mac and the tablet so the FAT is useful and always the last resort. However yes ext2 it could be an idea. However get OpenBSD improved is always my favorite option. -- Nowarez Market Dec 5, 2023 19:25:55 Noth : > Have you tried mounting with the -l flag? Otherwise I'd recommend using > ext2fs instead of FAT32. Android will handle it natively, and OpenBSD can > mount that read/write (on a sdXi mountpoint, like FAT32).
Re: efiboot: change default partition from hd0a
On Sat, 02 Dec 2023 11:53:25 +0900 (JST) YASUOKA Masahiko wrote: > On Fri, 1 Dec 2023 11:21:22 -0800 > Johnathan Cobden-Nolan wrote: >> I have installed OpenBSD on hd0l: in my case it is for multi-booting, >> but I imagine there are other use cases where boot and/or root are >> installed on partitions other than 'a'. >> >> This is a UEFI system so I've installed the efi bootloader which I am >> able to execute. The bootloader first complains that there is no >> hd0a:/etc/boot.conf. This is expected, since my install is at hd0l. Ah, I missed you don't want to create 'a' partition. I'm sorry. In that case, you need to modify the bootloader as you are already discussing. >> Since there is no boot.conf being read, it doesn't know where to try >> booting: I am only able to boot the OS by typing "boot hd0l:/bsd". > > If you just want to avoid the typing, you can put > > set device hd0l > > on /etc/boot.conf in 'a' parition. > >> This is not the end of the world, but it feels like it should be possible >> to have a boot.conf somewhere other than the 'a' partition. Is it? Is >> it possible to have a conf file in the EFI partition alongside the >> bootloader itself? >> >> Thanks, >> Johnathan >>