home.social

Search

483 results for “zirias”

  1. @stefano BTW, this seems to come with a certain mindset even among users and admins. It starts with complaints that #FreeBSD's rc can't automatically respawn crashed services, as if this was a totally normal and safe thing to do ...

    #softwarecrisis reloaded ...

  2. @stefano BTW, this seems to come with a certain mindset even among users and admins. It starts with complaints that #FreeBSD's rc can't automatically respawn crashed services, as if this was a totally normal and safe thing to do ...

    #softwarecrisis reloaded ...

  3. @stefano BTW, this seems to come with a certain mindset even among users and admins. It starts with complaints that #FreeBSD's rc can't automatically respawn crashed services, as if this was a totally normal and safe thing to do ...

    #softwarecrisis reloaded ...

  4. @stefano BTW, this seems to come with a certain mindset even among users and admins. It starts with complaints that 's rc can't automatically respawn crashed services, as if this was a totally normal and safe thing to do ...

    reloaded ...

  5. @drscriptt I'll just dump my understanding of the issue with #kerberos, #nfsv4 and #samba (or #ActiveDirectory) now on here, so if you know better, please correct me 😉

    A kerberized service needs an SPN to prove its identity towards clients. This SPN can be attached to any account, it's just best practice to have a dedicated service account. For nfs, it must be named "nfs/<host.fqdn>". The key for this SPN must be available to the server, that's why you have to export it and add it to /etc/krb5.keytab on the server.

    To access the service, you need a "service ticket" for its SPN. This can be obtained with your kerberos TGT (ticket-granting ticket). In case of NFS, this is also needed for mounting. And here's the issue with system-wide mounts, there's no user logged into the domain, so you need a "host-based initiator". Basically a key present in the host's /etc/krb5.keytab that's used for the kinit procedure.

    [...]

  6. @drscriptt I'm still doing a lot of testing and I guess I now finally have a somewhat good understanding what's going on.

    With a samba (Windows AD) domain, you should

    * create a service account for the nfs-server and create the nfs SPN in that account, export keytab entries for just that SPN and add them to /etc/krb5.keytab on the NFS server

    * set the "host" SPN as UPN on all machine accounts that are NFS clients, including the realm in the UPN, which enables the host-based initiator to work (authenticating as the machine account) for the system-wide mount

    I'll try to write that down in some howto soon, #jailed #nfsv4 with #kerberos on #FreeBSD and a #samba domain...

    I guess it won't necessarily apply to a directory managed in OpenLDAP though ... 😶

  7. Hmm ... kerberos, for now, does *not* work.

    Unfortunately, #FreeBSD's nfs isn't all too verbose about what's failing.

    I get: "permission denied" 😎 (and nothing else in logs either).

    At least I have an idea. I use MIT #krb5 from ports together with a #samba domain controller, so I build base with WITHOUT_KERBEROS=yes (I really don't need the heimdal from base). Just read the docs again, this also disables gssapi unless you add WITH_GSSAPI=yes. Which I did now. I'm pretty sure kerberized nfs will need libgssapi. We will see....

  8. Impatiently waiting for my #ZFS backup to complete ...

    Then the next step will be to test #jailed #NFS (as introduced in #FreeBSD 13.3), to finally replace my horrible hack of redirecting NFS-related traffic with #pf (and, therefore, punching a hole for LAN machines to access the physical host located in the management segment).

    I hope to also move to #nfsv4 at the same time. And once *this* works, enable #krb5 auth and encryption. We will see 😎

  9. #FreeBSD 13.3 deployed at home. 4 times "metal", one VM, 12 #VNET #jails.

    First time ever not the vanilla version from the releng branch (with just one commit adding kernel configs on top), but pulling in extra patches from some repo on github. 🙄 Because it's unusable in my setup otherwise 😞.

    I really really hope this will stay a rare exception for FreeBSD. And of course I hope for an EN including the functionality of all these patches I now pulled in myself ... would probably be a PITA having to maintain them locally.

  10. And now, my little documentation generation tool can also produce manpages in #HTML format (this will be useful for the #Windows version of #dos2ansi) 😎

    Using simple/minimal semantic markup, the "manpage style" is entirely left to #CSS

    Screenshots:
    - Desktop unstyled
    - Desktop bright mode
    - Desktop dark mode
    - Mobile dark mode

  11. And now, my little documentation generation tool can also produce manpages in #HTML format (this will be useful for the #Windows version of #dos2ansi) 😎

    Using simple/minimal semantic markup, the "manpage style" is entirely left to #CSS

    Screenshots:
    - Desktop unstyled
    - Desktop bright mode
    - Desktop dark mode
    - Mobile dark mode

  12. And now, my little documentation generation tool can also produce manpages in #HTML format (this will be useful for the #Windows version of #dos2ansi) 😎

    Using simple/minimal semantic markup, the "manpage style" is entirely left to #CSS

    Screenshots:
    - Desktop unstyled
    - Desktop bright mode
    - Desktop dark mode
    - Mobile dark mode

  13. And now, my little documentation generation tool can also produce manpages in #HTML format (this will be useful for the #Windows version of #dos2ansi) 😎

    Using simple/minimal semantic markup, the "manpage style" is entirely left to #CSS

    Screenshots:
    - Desktop unstyled
    - Desktop bright mode
    - Desktop dark mode
    - Mobile dark mode

  14. And now, my little documentation generation tool can also produce manpages in format (this will be useful for the version of ) 😎

    Using simple/minimal semantic markup, the "manpage style" is entirely left to

    Screenshots:
    - Desktop unstyled
    - Desktop bright mode
    - Desktop dark mode
    - Mobile dark mode

  15. Now working on adding a manpage for #showansi as well. I noticed I need to add features to my "mkclidoc" tool in order to get something meaningful here ...

    First step, added support for cross-references 😎 -- next thing needed will be documenting related files...

    Next #dos2ansi release will probably just be nice and complete documentation!

  16. Now working on adding a manpage for #showansi as well. I noticed I need to add features to my "mkclidoc" tool in order to get something meaningful here ...

    First step, added support for cross-references 😎 -- next thing needed will be documenting related files...

    Next #dos2ansi release will probably just be nice and complete documentation!

  17. Now working on adding a manpage for #showansi as well. I noticed I need to add features to my "mkclidoc" tool in order to get something meaningful here ...

    First step, added support for cross-references 😎 -- next thing needed will be documenting related files...

    Next #dos2ansi release will probably just be nice and complete documentation!

  18. Now working on adding a manpage for #showansi as well. I noticed I need to add features to my "mkclidoc" tool in order to get something meaningful here ...

    First step, added support for cross-references 😎 -- next thing needed will be documenting related files...

    Next #dos2ansi release will probably just be nice and complete documentation!

  19. Now working on adding a manpage for as well. I noticed I need to add features to my "mkclidoc" tool in order to get something meaningful here ...

    First step, added support for cross-references 😎 -- next thing needed will be documenting related files...

    Next release will probably just be nice and complete documentation!

  20. Meanwhile:

    * Added a #Debian (amd64) package to #dos2ansi v1.6 release (on Github)
    * Updated the #FreeBSD port here: people.freebsd.org/~zirias/pat
    * There's of course still the #Windows binary attached to the release

    Might push the FreeBSD port to official ports later ....

    Screenshot: Random #showansi rendering ... 🙈

    #MSDOS #ANSIart #retrocomputing

  21. Just noticed the #showansi script from #dos2ansi is now "smarter" than the renderer on 16colo.rs 😂

    When there's no (#SAUCE) metadata indicating the font to use, but the canvas after rendering turns out to be exactly 80x50 or 80x43, it just *guesses* that's most likely the VGA50 or EGA43 mode and picks that font 😉

    (Of course, this "smartness" can be disabled with the "-S" flag)

    Compare to this: 16colo.rs/pack/avpack16/frg-an

    #ANSIart #MSDOS #retrocomputing

  22. @dryak Oh boy! I really just learned the "classic" Windows Console palette was even "more broken", I wasn't aware so far. This would also mean there would be a very dark "dark white" and a second "black", but no, those two were adjusted...

    Also found they only changed that in #Win10, but instead of going for a real #RGBI palette with brown adjustment, they picked pretty arbitrary values creating a new less greenish "dark yellow" 🤦‍♂️
    devblogs.microsoft.com/command

    Indeed, I remember seeing lots of very regular dithering in Windows 3.x ... so you probably have a point here 😉

  23. @dryak Oh boy! I really just learned the "classic" Windows Console palette was even "more broken", I wasn't aware so far. This would also mean there would be a very dark "dark white" and a second "black", but no, those two were adjusted...

    Also found they only changed that in #Win10, but instead of going for a real #RGBI palette with brown adjustment, they picked pretty arbitrary values creating a new less greenish "dark yellow" 🤦‍♂️
    devblogs.microsoft.com/command

    Indeed, I remember seeing lots of very regular dithering in Windows 3.x ... so you probably have a point here 😉

  24. @dryak Oh boy! I really just learned the "classic" Windows Console palette was even "more broken", I wasn't aware so far. This would also mean there would be a very dark "dark white" and a second "black", but no, those two were adjusted...

    Also found they only changed that in #Win10, but instead of going for a real #RGBI palette with brown adjustment, they picked pretty arbitrary values creating a new less greenish "dark yellow" 🤦‍♂️
    devblogs.microsoft.com/command

    Indeed, I remember seeing lots of very regular dithering in Windows 3.x ... so you probably have a point here 😉

  25. @dryak Oh boy! I really just learned the "classic" Windows Console palette was even "more broken", I wasn't aware so far. This would also mean there would be a very dark "dark white" and a second "black", but no, those two were adjusted...

    Also found they only changed that in #Win10, but instead of going for a real #RGBI palette with brown adjustment, they picked pretty arbitrary values creating a new less greenish "dark yellow" 🤦‍♂️
    devblogs.microsoft.com/command

    Indeed, I remember seeing lots of very regular dithering in Windows 3.x ... so you probably have a point here 😉