The collected works of canavan - Page 6

And freetype 2.4.12 has actually been released in the meantime. I'll have to check how that new renderer is enable (runtime or compile time)...

There's also the question, which fonts will be rendered by that CFF renderer at all, apparently it's not the TrueType ones. I this is intended for Postscript Type 1 fonts.
It should not be necessary to set the *Background explicitly, it should just be taken automatically from the scheme.
I think you need the Development Foundation 1.3 CD to install MipsPro 7.4. There's nothing on there that matches any of the missing dependencies you mentioned, but I think that CD was the one I was missing all the time I that I was unable to install MipsPro 7.4.
Another vote for /current for neko_pango-1.28.4.
hamei wrote:
Doesn't much matter in this case tho, looks like the commuuuuuunity got its greasy paws into Postgresql 9.2.4 ->
checking thread safety of required library functions... no
configure: error: thread test program failed
This platform is not thread-safe.  Check the file 'config.log' or compile
and run src/test/thread/thread_test for the exact reason.
Use --disable-thread-safety to disable thread safety.

Oh right, I forgot that Irix isn't thread-safe.

They probably just failed at implementing the test. Maybe there's just a -lpthread missing or at the wrong place in the command line. It's almost as if they knew that would happen, seeing that they appear to have spent more time with that verbose error message than the test itself..

Since we're talking about C++ libraries: Has anyone considered writing a simple program that re-links a gcc c++ library with the symbols just re-mangled MIPSPro-style? Maybe it's even possible to have the symbols available in both mangling styles in the same library.
I'm trying to compile scummvm 1.6.0 with the mipspro compilers. Right now, I'm stuck with their singleton hack. The error I'm getting is
cc-3234 CC: ERROR File = graphics/fonts/ttf.cpp, Line = 479
Explicit specialization of member
"Common::Singleton<Graphics::TTFLibrary>::_singleton" must precede
its first use.

template<> Graphics::TTFLibrary *Singleton<Graphics::TTFLibrary>::_singleton = 0;

Although the constuct is supposed to prevent this exact error:
* Note that you need to use this macro from the Common namespace.
* This is because C++ requires initial explicit specialization
* to be placed in the same namespace as the template.
template<> T *Singleton<T>::_singleton = 0

} // End of namespace Common
Does anyone have an Idea how to work around this for MIPSPro C++?
The namespace is actually correct:
namespace Common {
} // End of namespace Common

It turns out that MIPSPro is way more picky about this, and just shufflihg the DECLARE_SINGLETON just after the declaration of the :TTFLibrary makes this file compileable.
I'd suspect there's either something wrong with the implementation of some Class invloved, or it's just the geometry or constraints of the Widgets. Try editres to read or change some height, bottomAttachment, bottomWidget or whatever suspect resources you can find.
Has anyone encountered this one:
ld32: FATAL 2 : Internal: at ../../ld/pass1.c Total size of all GPREL sections exceeds 64K
I'll check where I got neko_libidn 4 from, and/or downgrade to 3.
hamei wrote:
smj wrote:
Uncompressing the gzip output with pigz, using just 1 thread took 24.9 seconds (single run, but typical of others). But the same operation with 8 threads took 32.5 seconds
Is that correct or a misprint ? Uncompressing with 8 threads took 7.5 seconds longer than a single thread ?

That's no typo, it took several seconds longer trying to use multiple threads -- in this specific example. Maybe a much larger compressed file would yield different results.[/quote]

If I'm not mistaken, the documentation for pigz explicitly states that it does not provide any benefit for archives generated with 'normal' gzip. I'm surprised to see that it's so much worse.
In that screenshot, the first thing I'd suspect is the XmScrolledWindow, the parent of the XmScrollBars and the XmForm.
It appears that neko_libidn 4 is something that has failed to escape into the wild, and is currently only found on a few of my own SGIs. I've downgraded my libidn and re-build wget. The fixed package is in /incoming, so it may appear in /beta soon.
Obviously, you shouldn't include stuff from internal/ directly, so
$ grep -r limits_core.h /usr/include/
/usr/include/CC/climits:#include <internal/limits_core.h>
/usr/include/limits.h:#include <internal/limits_core.h>

While you're there, check the man page for limits
hamei wrote: May as well hang for a second ... there's a brandy-newiest libidn 1.27 that just came out yesterday ....
What the heck, the updated translations might be worthwhile ...

Well, 1.27 is in /incoming. You can check if the translations are any good as soon as the package hits /beta.
I can't install it, because I'm missing neko_bzip2.sw.lib 6. The last version I can find is the version 5 (that's bzip2 1.0.6) that I built about 1 year ago..
Just so that future generations can find this more easily: the NIC discussed here is the "number in a can" eprom(?) which is used in e.g. Octane and O2 to store the serial number, not the network interface...
I think the various bzip2 versions 1.0x should all be compatible, so reducing the required version should not hurt.
I don't think sticking a /usr/nekoware/etc/fonts/local.conf into the firefox package is a good idea. You could update the neko_fontconfig package, make a patch package that "updates" the local.conf or provide a separate package with only the local.conf.

On the other hand, instead of replacing local.conf you could just place a file in conf.avail and make a symlink in conf.d in a postinstall script.
Helvetica seems to work for me in any size in left, right or justified text.

If you want ted to be more SGIified, you'll have to add Mnemonics to the Menu and reduce the height of the menu bar.

The properties dialog doesn't close when pressing Alt+F4 or using the system menu, even the cancel button ceases to work. The cancel button starts to work again, if one re-opens the (already open) dialog via the file menu

Redrawing is inefficient, ted doesn't combine multiple redraw requests, instead it redraws hundreds of little slices if one moves another window across ted.

Any Idea how I can set the default page format to A4?

How do I zoom the page display?

It sometimes spills a few warnings / messages:
Class: XmScrollBar
The specified scrollbar value is greater than the maximum
scrollbar value minus the scrollbar slider size.

tedMouseInput.c(766) docX=197 docY=212

Aisde from those small problems (most of which aren's the package's fault), neko_ted seems to work fine for me.
If you want to turn this into a nekoware package, you should not compile against freeware, but instead use the (probably much more current) packages from nekoware. Since the gnuchess in freeware is almost a decade old, you may want to update that first and build a nekoware for that.
miro didn't make PCs, they started with PC graphics cards, and in cooperation with SGI, made a PC card with an R3000 on it: .
The jumpers for some of the IBM drives I've had are at back of the PCB, towards the motor, not at the connector side. You may have to take the drive out of the caddy, but I'd recommend reading the documentation first because the jumpers may or may not be labeled properly.
I don't seem to be able to convince the Xsgi that my monitor is indeed DPMS capable. I've tried just setting an "Unknown" entry in moninfo, an entry with the name given by gfxinfo, and setting a monitor type via chantab, I always get "Display is not capable of DPMS". Is there anything I've overlooked? There's been some discussion about this topic in viewtopic.php?f=7&t=16726867 and claims that power management is set in /usr/lib/X11/xdm/Xlogin, probably referring to the /usr/bin/X11/xset s 600 1200 in there.
DPMS (Energy Star):
Display is not capable of DPMS
#  cat  /var/X11/Xvc/chantab
# always EIZO-S2243 connected
0.0.1: EIZO-S2243
# tail -n 40  /var/X11/Xvc/moninfo

LineRate=31000-94000 FrameRate=55-76
VerticalBlanking=920000-20000000 VerticalSync=64000-333330
HorizontalBlanking=3000 HorizontalSync=670-9000
HeightMM=296 WidthMM=474
QueryResponseSerial="ENC 8512"

LineRate=31000-94000 FrameRate=55-76
VerticalBlanking=920000-20000000 VerticalSync=64000-333330
HorizontalBlanking=3000 HorizontalSync=670-9000
HeightMM=296 WidthMM=474
QueryResponseSerial="ENC 8512"

LineRate=31000-94000 FrameRate=55-76
# /usr/gfx/gfxinfo
Graphics board 0 is "ODYSSEY" graphics.
Managed (":0.0") 1920x1200
BUZZ version B.2
PB&J version 1
128MB memory
Banks: 4, CAS latency: 3
Monitor 0 type: Unknown
Dual Channel Display option
Monitor 1 type: ENC 8512        Monitor 2 type: Unknown
Channel 0:
Origin = (0,0)
Video Output: 1920 pixels, 1200 lines, 60.00Hz (1920x1200_60p)
You can actually read the ISOs on an SGI, if you have a spare harddrive connected and just dd the .iso over the entire drive. IT may be neccessary to run scsiha or fx or something similar to re-read the partition table on that disk. That should work both for proper, bootable EFS images as well for normal iso 9660 images.
xset dpms doesn't help. It doesn't complain, but after running xset s 5 10 and xset dpms 15, the screen turns black after 5s, but the backlight remains on even after 1 minute. I think DPMS works with an O2, a 7of9 and and the appropriate adapter, an Octane V6 and a 20D11, and if I'm not mistaken even an Indy.
I think prboom (plus) is the doom version to use. The package may need some improvements regarding the GUI integration, but the game itself works. The non-plus Version has red/blue reversed for me, but the plus fork is fine.

4 downloads, and nobody complained that the opengl graphics mode doesn't work? The lack of aligment of the GLSprites (and the 64bit Flags therein) cause bus errors immediately after starting a game (or demo). In the fixed version below, the sky boxes are still rendered as cubes (I think) and have very visible borders, but I'll leave that to an OpenGL experts. I think a replacement for the GLEXT_glMultiTexCoord2fvARB would be required to fix this.
If you want to read an ISO, you can skip the formatting and partitioning. Just dd the entire iso to /dev/rdsk/dks?d?vol, not into a partition thereon.
I haven't tried that with a firewire device, but that's basically the method I used to find out how to make bootable installation CDs, or produce UDF images to burn to DVDs. The same principle should apply to ISOs.
What ajw99uk means is the difference between the computer sending the signal to power down / go to sleep to the monitor (usually by turning off one or both of the sync signals), or just sends a black "image," while still sending H+V sync signals. I don't think the power management cares if the sync signals are separate, combined or on green.
A proper tardist for c-ray could easily include an optimized version for any (of the popular) IPxx / CPU combinations....
jpstewart wrote:
blowout6 wrote: Also, I'd imagine there should be driving / racing simulation software.

I wonder how feasible it would be to get TORCS running on IRIX. I suspect it would be better on an Onyx than anything in my humble collection, if it could run on IRIX at all..
If you want to run TORCS, you may have to build an older version, since the current Version requires OpenGL 1.3 for "driving" ... e&artid=30

Minimum: 550MHz CPU, 128MB RAM, OpenGL 1.3 compatible graphics card with 32 MB RAM.

The oldest version of that document on may indicate that TORCS 1.3.0 was the last Version that worked with OpenGL 1.2: ... e&artid=30
emilen72 wrote: hi recondas, thanks for the welcome here...
in the meantime hi fix the trouble.... I discovered that the default installation of pango and fontconfig don't set all the sub option... I reinstalled the two packages with all the subs set on... and now e-uae starts...

That would be a bug in the Package dependencies. Someone should fix that...
I'm not very satisfied from performance of the emulator.

I suspect the JIT doesn't support MIPS, so Intel CPUs should have a much more efficient emulation.
neko_xteddy ist still in /beta, just in case you should need it.
The VT510 I have here seems to be single session, the setup menu only lets me assign S1 and print functions to the various serial and parallel ports.
There's a srterror_r in gnulib: This function is missing on some platforms: NetBSD 3.0, Minix 3.1.8, HP-UX 11.23, IRIX 6.5, Solaris 9, mingw, MSVC 9. Sadly, using gnulib either implies that you use autoconf/automake, or do a lot of manual work to resolve all the macros. ... 005fr.html
Wasn't the "original" supported USB audio device the Griffin iMic? That thing should have had at least one input...
hamei wrote:
I've still got my old HP ScanJet 4p hooked up to my Indigo2 through SCSI. Works great with Impressario. :)

That would be the nicest, but old stuff is rare as hen's teeth here. And waaaay overpriced. Oh well :(

Don't look for old scanners in stores, but on the trash. That's where I got my scanjet 6100c from. I have no idea if it works with impressario, but it does work with xsane.
I just remembered that I also had a USB audio device, a SoundBlaster something (sb0270). When both analog and digital ports are on, IRIX complains that uaudio_init_channels() Could not get MIN/MAX/CUR/RES value for control 0x02 on channel 1 - usberr 0x625f0016, and apanel complains that there's "no audio driver at all", but it's still listed in hinv. With just the digital ports enabled, there are no complaints when booting or plugging the thing in, and apanel just shows a single output device.
You can't really put IMD application icons anywhere but /usr/lib/images, so those few packages that I have produced so far that include icons put them there as well. If you want to update the icons for specific nekoware packages, I'd prefer if you updated that specific package; it's probably outdated anyway.

/usr/include/X11/bitmaps/sgidesktop/ already contains files from various packages and subsystems, namely sysadmdesktop.sw.sysadm desktop_eoe.sw.envm and desktop_eoe.sw.control_panels.

For the backgrounds, It would be cool if you could include an exitop / removeop that automatically updates /usr/lib/X11/system.backgrounds when installing and removing the package.