The collected works of canavan - Page 7

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of VRMLview at June 29, 2014, 9:56 a.m...

Apparently, I have it, but my webserver is currently down. PM me your email address.


-rw-r--r-- 1 canavan user 3355413 Sep 17  2005 /usr/people/canavan/install/vrmlview_irix.tar.gz

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of Pretty - a few new desktop backgrounds for IRIX at June 30, 2014, 1:22 p.m...

Nice backgrounds, but apparently the subsystem with the exitop script is optional and the exitop for the images fails when it's not installed:

Code: Select all


Pre-installation check completed


Installing new versions of selected sp_pretty.art subsystems


Assigning id 60000 to unknown user/group name avi


sp_pretty.art.icons: ( /usr/sillypages/bin/root_icon_install )


/bin/sh: /usr/sillypages/bin/root_icon_install:  not found


ERROR: An error occurred while:


Installing new versions of selected sp_pretty.art subsystems


Command "( /usr/sillypages/bin/root_icon_install )" failed (return status 127, subsystem sp_pretty.art.icons 20147)


Removing orphaned directories

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of Pretty - a few new desktop backgrounds for IRIX at July 27, 2014, 3:08 p.m...

You need a prereq section that defines the dependencies. For example, in the neko_imagemagick package, the sw.eoe component doesn't work without the sw.lib bits:

Code: Select all


image sw


id "software"


version 17


order 9999


subsys eoe default


id "execution only env"


replaces self


prereq (


neko_imagemagick.sw.lib 17 maxint


)


exp neko_imagemagick.sw.eoe


endsubsys

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of usb audio at July 27, 2014, 3:31 p.m...

I don't think any specific chip is required. IRIX should support "any" USB audio class device, i.e. "generic" audio devices that doesn't require specific drivers to run.

http://www.usb.org/developers/devclass_docs

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of Porting libtorrent/rtorrent at Aug. 3, 2014, 10:46 a.m...

I haven't tried building anything current. How about trying yourself?

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of Netsurf 3.1 at Aug. 10, 2014, 1:07 p.m...

There's still the problem that you can't mix C++ code generated by g++ and CC, since they use different name mangling algorithms.

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of upgrading ram on sgi indy, indigo2 at Aug. 10, 2014, 1:17 p.m...

Trippynet wrote: @jan-jaap - I have had some HDDs fail to work in an Indigo2. I have two different models of 300GB SCA drives. One model works fine in my Indigo2, the other doesn't. Both work fine in my O2 however, so there is a compatibility issue there somewhere!

Some of the newer LVDS drives just don't support single ended SCSI anymore. A more common problem is that some drives just don't work if the unused 8 bit of the bus is not terminated, thus you need an SCA <> 50pin adapter that has a terminator for the upper 8 bit and that has to be enabled.

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of neko_librsvg-2.18.2 missing dependancy? at Aug. 10, 2014, 2:01 p.m...

The prerequisites of neko_librsvg-2.18.2.tardist are

Code: Select all


prereq (


neko_gtk.sw.lib 7 maxint


neko_libgsf.sw.lib 1 maxint


neko_libcroco.sw.lib 2 maxint


neko_libart.sw.lib 2 maxint


neko_popt.sw.lib 2 maxint


)

and the directly linked libraries are

Code: Select all


$ /usr/nekoware/gcc-4.7/bin/readelf -Wa /usr/nekoware/lib/librsvg-2.so | grep NEED


0x00000001 (NEEDED)                     Shared library: [libgnomevfs-2.so.1]


0x00000001 (NEEDED)                     Shared library: [libbonobo-2.so.1]


0x00000001 (NEEDED)                     Shared library: [libgconf-2.so.5]


0x00000001 (NEEDED)                     Shared library: [libbonobo-activation.so.5]


0x00000001 (NEEDED)                     Shared library: [libORBit-2.so.1]


0x00000001 (NEEDED)                     Shared library: [libgthread-2.0.so.1]


0x00000001 (NEEDED)                     Shared library: [libpthread.so]


0x00000001 (NEEDED)                     Shared library: [libgsf-1.so.2]


0x00000001 (NEEDED)                     Shared library: [libgdk_pixbuf-2.0.so.1]


0x00000001 (NEEDED)                     Shared library: [libxml2.so.3]


0x00000001 (NEEDED)                     Shared library: [libpangoft2-1.0.so.1]


0x00000001 (NEEDED)                     Shared library: [libpangocairo-1.0.so.1]


0x00000001 (NEEDED)                     Shared library: [libpango-1.0.so.1]


0x00000001 (NEEDED)                     Shared library: [libgobject-2.0.so.1]


0x00000001 (NEEDED)                     Shared library: [libgmodule-2.0.so.1]


0x00000001 (NEEDED)                     Shared library: [libglib-2.0.so.1]


0x00000001 (NEEDED)                     Shared library: [libintl.so.4]


0x00000001 (NEEDED)                     Shared library: [libcairo.so.3]


0x00000001 (NEEDED)                     Shared library: [libglitz.so.2]


0x00000001 (NEEDED)                     Shared library: [libpng12.so.0]


0x00000001 (NEEDED)                     Shared library: [libm.so]


0x00000001 (NEEDED)                     Shared library: [libXrender.so.1]


0x00000001 (NEEDED)                     Shared library: [libX11.so.1]


0x00000001 (NEEDED)                     Shared library: [libfontconfig.so.2]


0x00000001 (NEEDED)                     Shared library: [libfreetype.so.7]


0x00000001 (NEEDED)                     Shared library: [libz.so]


0x00000001 (NEEDED)                     Shared library: [libc.so.1]

So I'd say there's a whole lot more missing than just gnomevfs.

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of upgrading ram on sgi indy, indigo2 at Aug. 16, 2014, 12:56 p.m...

The internal SCSI bus in an Indigo2 is of course an 8 bit bus, and therefore has termination for those 8 data lines and the 3 address lines that go along with it. At least some of the drives you linked to are 80 pin SCA, that implies a 16 bit wide data bus, with 4 adress lines. Those extra 8 data lines and the 1 address line are not terminated by the resistors already installed in the Indigo, they just plain aren't connected. Back when I had an Indigo2, the brand new IBM drive I bought refused to work, because it wanted all those floating lines terminated as well. The SCA adapters I have here have an optional little extra board in one corner that's called "9 lines terminator" with a bunch of resistors, a transistor and a few capacitors.

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of upgrading ram on sgi indy, indigo2 at Aug. 17, 2014, 3:41 p.m...

robespierre wrote: I don't mean to quibble, but what do you mean "address lines"? that is not part of the parallel scsi interface.
edit: on a SCA connector there are 4 ID pins, but these are not bussed, they are part of enclosure services.

You're of course correct, there's no such thing, I just misremembered that part.

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of 80 pin SCSI drive in Indigo2 - how? at Aug. 17, 2014, 3:47 p.m...

smj wrote: I hadn't thought the connectors were compatible with proper SCA - I thought they were inverted or something at minimum - but live and learn. Even so loose drives rattling around will work, but it's nice to use the sleds. You're going to be stuck with 10MBps SCSI either way. But if you secure the drive to the sled, in 2 months to 2 years from now when you move the machine you won't have to hear the sickening "thunk" to remember "Oh right, the drive is loose in there!"

One could still cut off the back plate of the original drive sled and make a pair of U shaped brackets to attach the drive further up and to the back (and optionally provide mounting holes for the front cover). Maybe that drive sled without the electrical bits I have in a drawer somewhere will be useful again some day if I get my hands on an Indigo 2.

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of DM3 / DM5 LVDS cable pinout at Sept. 6, 2014, 12:40 p.m...

viewtopic.php?f=4&t=16727186&start=15

The claim about the possible alternative cable in the Wiki is suspicious. Someone needs to verify my measurements.

http://www.nekochan.net/wiki/DMediaPro

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of Netsurf 3.1 at Sept. 7, 2014, 7:47 a.m...

i remember that it's been mentioned that some libs required by netsurf were compiled with g++?
if so that's a likely reason

If that were the case, it would already fail when linking, and not produce an executable at all.

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of sgi_led: Control your front panel LED at Sept. 7, 2014, 9:14 a.m...

Hmm. Might have to see if this can be modified for Fuel.

I haven's tried it on my fuel, but it works on my tezro, although you have to be root - no error message is printed if the access is denied.

canavan
Who joined July 21, 2003, 6:19 a.m.

Wrote on the subject of beta: autoconf 2.69 at Sept. 29, 2014, 12:45 p.m...

I'm uploading neko_autoconf-2.69_6.tardist to /incoming. It does include m4sugar/foreach.m4 (no Idea why that was missing, I seem to have neglected to rebuild the .tardist before uploading, my local .idb was already fixed), drops Struct.pm, which shouldn't have been in there in the first place, properly installs the info pages and tags the binaries as 0x620 (is that "correct"?), and bumps the version to 6.