[ old MirBSD™ logo ]
This is an archived weblog of the development leading to MirOS. All entries are courtesy of tg@ and historic.
In order to test my snapshot, I did a complete re-install, and I had to do a re-partitioning anyways. Unluckily I hosed the backup process, and lost the cvs repository (but not the checked out copy). Busy.
I just discovered that there has been a fix for misc/mc with regard to the terminal Fn-Key issue. It got lost, somewhen. Plus, this propolice stuff is gonna suck. Downloading CTM...
Done RCSId with a simple regexp search-replace; prepared re-upload of jupp/DOS16. Regenerated patches in sync; now trying to go and build libc, gcc and egcc propolice-enabled.
Try to fix ports - especially egcs. Need yet to test it. Add a port of Lutz Donnerhacke's pgp-2.6.3in version. Start synching with latest -current (still unusable). Update ports INDEX.
Sync, make sure it compiles (after cvs ci, maybe I've got to come back). Maybe fix md/mi lists. Need a Lynx-current port.
This one compiled fine, plus some minimal changes.
Major sync, byebye amiga. Add gcc-3.2.1 propolice diff backport from cvs. Change to /etc/profile (alias hd).
Still no lynx-current port, still not tested, but sync because xsa-the-bsdcow wants to examine my stuff. make INDEX tho.
No lynx, but make INDEX after adding just another egcs (gcc) port - this time with latest propolice from etoh@... (still not tested; not even all diffs synched)
... later the same night
That stuff (with the [http://templeofhate.com/tglaser/obsd/gcc-3.2.1p1-system.tgz] new C compiler) survived a make build release, so I got a chance to sync the lists and commit an updated set of diffs. I don't know if my bktr diff is correct, though.
... after a reboot
I just noticed I had a dupe in the diff file packaging lists. This must not happen. Never.
I also have some too much files in them, files with empty diffs. Why? Maybe some stuff was committed. Help me to spot them.
Fix the problem in obsd.tygs, make a new kernel *sigh*
Sync, prepare for libpthread changes.
Also convert most of the ports to force-flavour style instead... should fix some dependencies.
make build release, sync, commit, build new egcc.
That's how it was planned; however I had to make contbuild quite often, so I didn't make release. Do that tomorrow or so.
The system should then be ready to make a snapshot, but I'll delay that until I got the ELF patches (so I can easily go back myself too :)
- I decided to split the snXXXX.php files after each successful snapshot release, in order to ease showing you what changed since the last one.
- Fixed small bugs in regen-obsdp.sh
- Got a fix to jdk-1.3-linux by Santana
- Todd Fries joined us in IRC and I helped him finish a patch to the
kernel facilitating "ps a" and "ps e" limits. He started looking over
my diffs, finding much of them interesting, wondering why they have
not been integrated into the main trunk.
He started waking interest in me to use OpenCM. I'll wait until it's stable tho.
- I just started using ssh-agent. Had to type in the ssh key password waaay too often, since I manage(d) the OpenBSDcow Network, this page and a trash.net shell account.
- GCC 3.2 is stable, but don't try G++ or even the Jehova™ compiler.
- as usual - sync, ports/make index
I hope to be able to make a snapshot RSN - waiting for the ELF diffs. Todd Fries offered me an ELF snapshot, but I think I could only use gcc+binutils out of it when I really get the diffs. Hm, still no lynx port, lazy me.
Hmm. I could not sleep more than 6.5 hours. Weird. So I decided to finally make that damn lynx-current port. Here it is.
It needs a patch to the terminfo master file though.
19.01.2003 later the evening
CVS fucked up basically any $OpenBSD RCS Id it could find. Reverting them to content-less state.
I must be crazy and/or have too muchtime on my hands. But maybe it's just too few, because on the evenings there is only three and a half one from coming home to going into bed. Well, sync, since bind was imported I cobbled a new host(1) together, etc.pp - and I've synched the diff packaging lists. Hm. Yesterday... oh yes, I set up a freenet-project persistent node. I should do a port for that, too.
Oh well, crazy me adds contributed stuff from the web page, too...
Synched and added some new acronymes. Update fast, there are quite some fixes in it.
Sync, fix, make build, bump mc, try to switch c++ compiler to egcs-3.2 (but do not commit changes yet). make release done tomorrow.
Despite the nice weather, I fixed some gcc compiler warnings about e.g. uninitialized variables, and did a make release, synched everything. Now I've got spare time to go outside. Well, C++ is still flakey, but OTOH I don't know C++ and leave it to someone else to fix it.
Put obsd.gcc and obsd.mini -> obsd.misc
Some stylistic fixes of the web page. Update lynx port; make gcc port breakage in libestdc++ ltconfig take fewer time. Move utilities from game to base. Check in a SNAPSHOT release candidate.
Last-Minute changes include diff-packaging-list fixes, movement of host and nslookup to usr.bin, correction of the nslookup manual page (8, not 1) and the according changes in the dist set lists.
Last-Minute Fixes to Snap#2 RC3
The sets were incorrect. D'oh.
I just saw that my termtypes diff got committed (not yet in). 10x miod
The snapshot RC3 I built was a success. Expect it, affiliated diffs, my-ports.tgz and basic packages, as well as a PGP signed RIPEMD-160 hash file, to make its way to the mirrors soon.
Hope you like it. Have a lot of fun...