@nhbriggs installing OpenSSL 3 alongside OpenSSL 1 shouldn't have broken anything - that's the default state of Solaris 11.4.42 through 11.4.80 while we worked through converting all the software that used OpenSSL 1 to use 3 instead. But drop-in replacement can't work as they're not compatible (hence the change of major version number) - if you've done something like that, booting from a previous BE is probably your best choice, since the pkg system depends on OpenSSL.
@nhbriggs oh, once you had root access, you shouldn't need to recompile, just mount the previous BE and copy the file from there. At this point you probably want to run at least "pkg verify" to see what else was changed, and possibly "pkg fix" to fix it.
@nhbriggs oh yeah, the OpenSSL 1.0.2 removal changeset requires changes to the core OS packages to stop depending on that version- it wasn’t intended to be installed without them. We’re working on the paperwork now to get an update to the CBE out later this year to cover things like that.
@ytc1 that's the way it's been for decades - you can see the bug to document that in the man page for Solaris 9 back in 2002: https://illumos.org/opensolaris/bugdb/bug.html#!4622577 and that text is still in the 11.4 man page. Similar text is in the shadow(5) man page in the description of the *LK* value.
26 years ago today was my first day as a student intern in the Sun Desktop Software Release Engineering team (then called "Power Client Software" to distinguish it from the Thin Client orgs handling Java Stations and Sun Rays).
I became a full-time employee 6 months later, after graduating at the end of summer semester.
6 months after that, aka 25 years ago now, I moved from the release engineering team into the X11 development team.
@adilarif I believe Wayland is better designed for the requirements of the 21st century, instead of the 1980's era requirements X11 tried to meet, but since I primarily work on an OS that can only run X11 and not Wayland, I've never dug too deeply into it. (Which is why I also still do a lot of maintenance work on X11 stuff, since that's what we use and ship at work.)
In #OracleSolaris 11.4.78, the iostat command now accepts fnmatch(7)-style patterns for matching disk device names, such as "iostat -xnb c1t[0-2]d\* 1 1"" (the \ is to avoid your shell trying to expand by itself).
Also, the new option '-b' to iostat displays backend statistics for virtual devices assigned to guests (Kernel Zones or LDom guests).
The cpio command in #OracleSolaris 11.4.78 will no longer extract files to outside the current working directory unless the new -x option is provided, making it safer if archives contain files with absolute paths, or relative paths containing '../' in them.
The IPS pkg command in #OracleSolaris 11.4.78 was fixed to pick a uid over 100 when creating a user without a uid specified, to avoid conflicts with uids reserved for OS packages.
useradd(8)/groupadd(8) were changed to allocate the first free uid above 1000 when no explicit uid is given, to keep them out of the space reserved for packages.
@awb the man page sources are in XML now, using a customized DocBook DTD, much like the DTD used with SGML, from which we generate nroff, html, & PDF output, but only the nroff output is shipped with the OS now. We do the XML->nroff conversion once at build time, instead of every time a user runs the man command as was done in the SGML days.
@awb right - the original plan was to create a curses gui more like GNU info than "nroff | more" for display of the SGML man pages allowing for following links & other advanced features, but it was never completed. All that ever shipped was a pipeline to convert SGML to nroff to view the pages the traditional way when you ran the man command.
The SGML support was removed from the man command in Solaris 11.4 after we finished the SGML to XML conversion for all the man page sources.
@awb the SysV shuffling annoyed me as well when packaging FOSS for Solaris and we'd have to shuffle their man pages to fit, and when I had to implement support for it during the conversion of X11R7 from Imake to automake. So when I became tech lead for Solaris 12 (later renamed 11.4) I decided we'd undo it: https://blogs.oracle.com/solaris/post/normalizing-man-page-section-numbers-in-solaris-114
#OracleSolaris 11.4.75 adds GCC 14 to the set of available packages. Users of the C and C++ compilers may notice that GCC 14 has raised a number of issues from warnings to errors, which may impact your builds. See https://gcc.gnu.org/gcc-14/changes.html for more information on the changes in this generation of the GNU family of compilers, and https://gcc.gnu.org/gcc-14/porting_to.html for information on changes you may need to make to code and/or compiler flags in order to build it with GCC 14.
The pgrep and pkill commands in #OracleSolaris 11.4.75 also grew two new options for selecting which processes to match: -r pidlist limits matches to processes with the listed pids at the root of their process trees. -Z (capital Z) limits matches to processes in the same zone (i.e. also in the global zone, since only in the global zone can you see processes from other zones).
@lanodan heh, though even when everything's working, the way the Oracle blog pages load the content through Javascript doesn't seem to play well with the Wayback Machine. Hopefully things improve next year when they migrate the blog site to a new backend.
Solaris Engineer at Sun^H^H^HOracle (Release management, Security, X11, GNOME); former board member of X.Org Foundation & OpenSolaris. http://pronoun.is/he