so the ibm boxes don't come with even one proper serial port by default?
r-a-c.de
The collected works of foetz - Page 24
ClassicHasClass wrote: they won't work for, say, modems
yeah what i meant, no proper ports then
r-a-c.de
r-a-c.de
the pet reminds me of scifi flicks from the 70s. love them
r-a-c.de
josehill wrote: an explicit, manual mkfs or mkfs_xfs step has long been a part of my personal recipe
same here and not only for irix. if you wanna rule something out do it yourself
r-a-c.de
you need to set $LIGHTWAVE to your lightwave bin dir
r-a-c.de
gijoe77 wrote: oh duh, that did the trick! Thanks again!
my pleasure
usually that should be added by the installer but if you copy the stuff over by hand or use a non standard shell it needs to be added manually of course
r-a-c.de
yeah logan's run. one of the classics i had in mind. nice pinball cover as well
r-a-c.de
you could also give ixquick a try
r-a-c.de
r-a-c.de
r-a-c.de
kshuff wrote: Got it installed but now it's hanging up booting, getting an error about the hostname not being set and can't find lboot. The fun continues.
how did you install it?
r-a-c.de
kshuff wrote: miniroot
install default
conflicts
go
okay, did it show any conflicts?
anyway no hostname at the first boot is normal. do you get to the desktop at all? or at least a single user shell?
if so just install the missing stuff. lboot for example is part of eoe1 and the required other stuff for building a kernel is part of compiler_eoe
r-a-c.de
congrats, just fiddled with mine with not so much luck
buuuuut, although the alphapcs were not considered "real" alphas by some people it should really run tru64 or vms
buuuuut, although the alphapcs were not considered "real" alphas by some people it should really run tru64 or vms
r-a-c.de
figured i could fire up my alpha again after quite a while but it wouldn't behave
at first it seems to work all fine. srm comes up, devices show as supposed to and i can boot from either cd or my old inst from disk. however unfortunately it never gets past a certain point, no matter whether i boot the installer cd or my old inst from disk. in any case it stops like this:
and that's it. when booting from cd it reboots after that and booting from disk just goes back to the srm prompt. searches on the net were not very helpful and neither is that error message.
i was a bit surprised because while i'm in srm everything is fine and booting up to that error is fine, too. so i'd guess cpu, ram and the scsi stuff are at least not too broken to get to that stage. i also tried the usual srm stuff like setting os_type to vms or unix, booting with flag 0 and so on. no joy either.
also worth mentioning the same happens no matter whether i boot tru64 or vms so i fear it's something more serious
at first it seems to work all fine. srm comes up, devices show as supposed to and i can boot from either cd or my old inst from disk. however unfortunately it never gets past a certain point, no matter whether i boot the installer cd or my old inst from disk. in any case it stops like this:
Code: Select all
using 922 buffers containing ...
CP - SAVE_TERM ...
halted cpu 0
halt code = 6
double error halt
pc = 1379c
warning -- hwrpb is invalid
and that's it. when booting from cd it reboots after that and booting from disk just goes back to the srm prompt. searches on the net were not very helpful and neither is that error message.
i was a bit surprised because while i'm in srm everything is fine and booting up to that error is fine, too. so i'd guess cpu, ram and the scsi stuff are at least not too broken to get to that stage. i also tried the usual srm stuff like setting os_type to vms or unix, booting with flag 0 and so on. no joy either.
also worth mentioning the same happens no matter whether i boot tru64 or vms so i fear it's something more serious
r-a-c.de
yeah the system is the same as before. it was running exactly like that for several years so the stuff definitely works together.
anyway if i remember right that's the point where the os actually takes over so i'd guess cpu or ram since cd and hdd both have exactly the same problem. i remember dealing with an alpha with cpu errors once but these errors were different and random which leaves the ram. would make sense because before the "crash point" ram is hardly used and it happens right after the alloc message.
scsi and ide can also be ruled out (unless the system board is junk) because both had this exact problem and it's not likely that all scsi and ide parts are broken at the same time
will see if i have some sdrams left somewhere ...
anyway if i remember right that's the point where the os actually takes over so i'd guess cpu or ram since cd and hdd both have exactly the same problem. i remember dealing with an alpha with cpu errors once but these errors were different and random which leaves the ram. would make sense because before the "crash point" ram is hardly used and it happens right after the alloc message.
scsi and ide can also be ruled out (unless the system board is junk) because both had this exact problem and it's not likely that all scsi and ide parts are broken at the same time
will see if i have some sdrams left somewhere ...
r-a-c.de
ClassicHasClass wrote: I have all the parts, but I haven't started on building the system yet. I did order a few "Alpha Powered" stickers though.
ha nice, i had them too back then
r-a-c.de
indeed that'd be the worse case but i'm optimistic because i dealt with broken alpha caches before and the problems caused by that were quite different
r-a-c.de
jpstewart wrote: all use the same "3 lines" icon for "Settings". Probably lots of other software, too. Those are just the examples I can see in front of me right now. It took me a while to figure out what the icon was the first time I encountered one.
that's supposed to be an abstract representation of a list (of options). the gfx guys these days must be more lazy than ever
r-a-c.de
we were all wrong
it was the scsi controller. strange tho because a linux test installation was no problem but the native systems wouldn't initialize. no problem tho because tru64 accepted an ide disk for which i had no use anymore anyway so it even was for a good cause. yes yes i know ide sucks but in this case it's good enough
btw i was looking for some basic freeware such as wget, tcsh and such ... just the basic stuff but except for thewrittenword i didn't find anything. maybe i should provide some packages here for tru64 if there's some interest ...
it was the scsi controller. strange tho because a linux test installation was no problem but the native systems wouldn't initialize. no problem tho because tru64 accepted an ide disk for which i had no use anymore anyway so it even was for a good cause. yes yes i know ide sucks but in this case it's good enough
btw i was looking for some basic freeware such as wget, tcsh and such ... just the basic stuff but except for thewrittenword i didn't find anything. maybe i should provide some packages here for tru64 if there's some interest ...
r-a-c.de
risking you might find me odd but is there a way to use the "desktop" used for the installation later with the real thing?
it reminds me of ultrix, just nicer. that ruby bg color and the golden wristwatch icon sort of got me
it reminds me of ultrix, just nicer. that ruby bg color and the golden wristwatch icon sort of got me
r-a-c.de
nekonoko wrote: I was a little pissed off at how the other thread was going
not your fault anyway. the phpbb devs are responsible for whatever happens
r-a-c.de
mhm, maybe i can extract the parts from the cd and hack something together
r-a-c.de
after quite some searching i found it and modified it a bit. stop xlogin and run this. you also need to copy 2 files from the inst cd: Xdefaults and mwmrc. both located in /isl. easiest way is to create a local /isl and put the files in there. otherwise edit the script accordingly.
r-a-c.de
since my disk clone script for irix was quite popular i figured i could also share the one for osf. here it comes:
Code: Select all
########################################################################
#
# R-A-C OSF/1 AdvFS system disk clone
#
#
# this script generally assumes disk id 1 and that your target disk is
# partitioned already. if not run:
# disklabel -wr -t advfs /dev/rdisk/dsk1c
#
# then partition your disk using diskconfig or disklabel.
# this script expects the following partitions:
# - a: root, rather small.
# - b: usr, the big one. everything else goes there.
# - d: swap, not necessary for the cloning.
#
# if you should not have var as part of the usr domain change the
# related parts accordingly.
# if you should use the clone with a different id than the original
# remember to change the swap entry in /etc/sysconfigtab on the clone.
#
########################################################################
echo preparing ...
rmfdmn -f clone_usr
rmfdmn -f clone_root
cd /mnt
mkdir dr du dv
echo creating clone root domain ...
mkfdmn /dev/disk/dsk1a clone_root
echo creating clone root fileset ...
mkfset clone_root root
mount_advfs clone_root#root /mnt/dr
echo creating clone usr domain ...
mkfdmn /dev/disk/dsk1b clone_usr
echo creating clone usr fileset ...
mkfset clone_usr usr
mount_advfs clone_usr#usr /mnt/du
echo putting var into the usr domain ...
mkfset clone_usr var
mount_advfs clone_usr#var /mnt/dv
echo dumping ...
vdump 0f - / | ( cd /mnt/dr ; vrestore xf - )
vdump 0f - /usr | ( cd /mnt/du ; vrestore xf - )
vdump 0f - /var | ( cd /mnt/dv ; vrestore xf - )
echo fixing domains for the clone ...
cd /mnt/dr/etc/fdmns
mv root_domain root_domain_old
mv usr_domain usr_domain_old
mv clone_root root_domain
mv clone_usr usr_domain
echo cleaning up ...
cd /mnt
umount dr
umount du
umount dv
rmdir dr du dv
echo all done
exit 0
r-a-c.de
search results for "tru64 freeware" were quite a wasteland so here comes the first bunch
these are built on 5.1b1 with the native compilers, optimized as hell, reasonable feature selection ... the usual
i don't know about backwards compatibility so just try in doubt. unless noted otherwise all packages are provided as native "kits".
http://www.nekochan.net/downloads/index ... z%2Ftru64/
MIRROR
append the package name since there's no listing
these are built on 5.1b1 with the native compilers, optimized as hell, reasonable feature selection ... the usual
i don't know about backwards compatibility so just try in doubt. unless noted otherwise all packages are provided as native "kits".
- gettext_0146.tar.bz2
- glib_1210.tar.bz2
- gtk_1210.tar.bz2
- jpeg_6b.tar.bz2
- libiconv_112.tar.bz2
- libpng_1616.tar.bz2
- openssl_098ze.tar.bz2
- sylpheed_106.tar.bz2
- zlib_128.tar.bz2
http://www.nekochan.net/downloads/index ... z%2Ftru64/
MIRROR
Code: Select all
r-a-c.de/dist/tru64/
append the package name since there's no listing
r-a-c.de
yeah i remember the pkgsrc discussions in the past although i never saw a need to involve a 3rd party "system" for something that's included by default.
as for the compilers, they're pretty good. both c and c++, i hardly had to step in at all. would be intersting to see how fortran does ...
as for the compilers, they're pretty good. both c and c++, i hardly had to step in at all. would be intersting to see how fortran does ...
r-a-c.de
5.3 doesn't come with flexlm at all. wordperfect would have to supply it on its own.
r-a-c.de
- apr_0920.tar.bz2
- aprutil_0920.tar.bz2
- curl_7400.tar.bz2
- diff_281.tar.bz2
- essentials.tar.bz2
- gimp_125.tar.bz2
- libpng_1212.tar.bz2
- libxml2_277.tar.bz2
- ncftp_325.tar.bz2
- patch.bz2
- screen.bz2
- svn_155.tar.bz2
- tiff_403.tar.bz2
- xchat_1811p1.tar.bz2
essentials is just a tar with a few binaries that have no deps and need no specific location which i consider a good start. things like tcsh, wget, bzip2, nedit and such. just throw them wherever you like.
the older libpng was needed for this gimp. the newer one i uploaded before didn't work. don't use both at the same time because they'd overwrite some files of each other.
was surprised to find out that neither the diff nor the patch program that comes with 5.1b is decent given that other base stuff is quite up to date for its time hence wrapped up an alternative diff and patch. the latter as well screen are just the naked binaries because there's not more to it
r-a-c.de
of course, never trash the old gems. some are irreplaceable
r-a-c.de
- pcre_836.tar.bz2
- mariadb_5214.tar.bz2
- apache_2412.tar.bz2
- m4_1412.tar.bz2
- autoconf_266.tar.bz2
- libevent_2021.tar.bz2
- php_5329.tar.bz2
- memcached_1422.bz2
maria, php and apache are no kits but just a single dir that should go to /usr/local. as usual i like to have certain programs in one place instead of scattered all across the filesystem.
apache comes with the prefork mpm simply because that works best with php by far.
php is quite pure because i tried to keep as much stuff as module as possible so that it's as flexible as possible. only the included modules are in this pack. externals such as apc or memcache will follow soon as a separate package. some of them have special deps of course depending on what they do.
r-a-c.de
media related foss stuff is usually a no-go on unix. most of it is exclusively targeted at linux/x86 and requires a crapload of other things that come with most linux distris. and even if it gets through somehow it won't run well.
xmms1 was an honorable exception but the rest is just not reasonable and also not what the professional unix devirates were made for.
if you have a look at the old mplayer threads here you'll see how much work it required and even then didn't play much and wasn't fast enough either.
the usual rule of thumb applies here as in commercial unix was made for professional work. if you wanna play get windows
xmms1 was an honorable exception but the rest is just not reasonable and also not what the professional unix devirates were made for.
if you have a look at the old mplayer threads here you'll see how much work it required and even then didn't play much and wasn't fast enough either.
the usual rule of thumb applies here as in commercial unix was made for professional work. if you wanna play get windows
r-a-c.de
vishnu wrote: Welcome to The Cloister and congratulations on your ordination to The Order!
what he said
r-a-c.de
i'd get an external scsi plextor or one of the few others that were accepted by most risc machines. it's really handy having that around not only for sgi but any other boxes you might have
r-a-c.de
- squid_27s9.tar.bz2
- postgresql_7430.tar.bz2
- postgresql_8422.tar.bz2
- postgresql_941.tar.bz2
- libssh2_143.tar.bz2
- gftp_2017.tar.bz2
- climm.bz2
- gtar.bz2
- openssh_66p1.tar.bz2
squid and the postgres are normal tars. matching postgre php modules coming next ...
r-a-c.de
cosmo code was free at some point. sgi themselves provided an unlimited license on their site so no legal issues in this case
r-a-c.de
ivelegacy wrote: let me understand: does this software require a flex license in order to work ?
if so, how to run it on an IRIX 6.5.22 machine ( my little O2+) ?
well to run it you'd need the mentioned license
r-a-c.de
- libiconv_112.tar.bz2
- php_5329-cgi.tar.bz2
- nginx_179.tar.bz2
- lighttpd_1435.tar.bz2
- php_modules.tar.bz2
- zsh_404.tar.bz2
yes, libiconv replaced. the last one didn't have static libs.
php modules include modules for all 3 postgre versions, xcache, memcache and suhosin.
the php cgi version is independent and can be used as an fcgi runner for use with the other webservers. it contains all modules.
r-a-c.de
vishnu wrote: Against IBM! And we all saw how that worked out for them...
at the time they offered mips systems ibm didn't play much of a role in the hpc sector and putting their money on that was logical because that's where the big bucks came from. gfx and entertainment was nice but not the main source of the annual revenue. military and scientific projects washed in the dollars and that for many years.
the reasons why they ended up the way they did were quite different but i think we talked about that more than enough here already
r-a-c.de
absolutely preserve it. and maybe even share it if you feel like it
r-a-c.de