SGI: Development

beta test: neko_pango-1.14.10 - Page 1

diegel
Who joined Nov. 17, 2009, 2:08 a.m.
and authored 158 notes

Wrote the following at March 14, 2013, 8:12 a.m...

neko_pango-1.14.10.tardist is available in beta now. This version of pango make firefox3 run a lot more stable.

_________________
:Tezro:

hamei
Who joined Feb. 24, 2004, 4:10 p.m.
and authored 8702 notes

Wrote the following at March 14, 2013, 5:41 p.m...

diegel wrote:

neko_pango-1.14.10.tardist is available in beta now. This version of pango make firefox3 run a lot more stable.

I've only had it installed for five minutes now but the FF is noticeably quicker. And the lousy scrolling issues seem to be either gone or much better.

I'd recommend trying this to anyone who is actively using Firefox.

Out of curiosity- what happens at a newer pango ? It's up to 32 or something, so there must have been a reason you chose this version ?

_________________
waiting for flight 1203 ...

diegel
Who joined Nov. 17, 2009, 2:08 a.m.
and authored 158 notes

Wrote the following at March 15, 2013, 5:55 a.m...

hamei wrote:

Out of curiosity- what happens at a newer pango ? It's up to 32 or something, so there must have been a reason you chose this version ?

I didn't try any newer version - pango 1.14 is the minimum requirement for firefox 3.5 so it's probably a good choice for firefox.

_________________
:Tezro:

hamei
Who joined Feb. 24, 2004, 4:10 p.m.
and authored 8702 notes

Wrote the following at March 15, 2013, 6:10 a.m...

diegel wrote:

pango 1.14 is the minimum requirement for firefox 3.5 so it's probably a good choice for firefox.

Ah. Makes sense.

Anyway, it's definitely an improvement. I didn't get to drive the FF hard today but hasn't crashed once and renders faster. Thanks

I am ready to give it a "move to /current" vote. It's vastly better.

_________________
waiting for flight 1203 ...

recondas
Moderator
Who joined June 6, 2004, 4:55 p.m.
and authored 4966 notes

Wrote the following at March 15, 2013, 6:23 a.m...

Code:

% versions neko_pango

I = Installed, R = Removed

Name                 Date        Description

I  neko_pango           03/15/2013  pango-1.14.10 GTK2 Core and Text Font Handling

I  neko_pango.man       03/15/2013  man pages

I  neko_pango.man.html  03/15/2013  html documentation

I  neko_pango.man.manpages  03/15/2013  man pages

I  neko_pango.opt       03/15/2013  optional software

I  neko_pango.opt.relnotes  03/15/2013  release notes

I  neko_pango.sw        03/15/2013  software

I  neko_pango.sw.eoe    03/15/2013  execution only env

I  neko_pango.sw.hdr    03/15/2013  header

I  neko_pango.sw.lib    03/15/2013  shared libraries

Worked well here too. I'll second the vote to move neko_pango-1.14.10 to /current.

_________________
***********************************************************************
Welcome to ARMLand - 0/0x0d00
running...(sherwood-root 0607201829)
* InfiniteReality/Reality Software, IRIX 6.5 Release *
***********************************************************************

jirka
Who joined May 4, 2004, 6:55 a.m.
and authored 120 notes

Wrote the following at March 15, 2013, 1:15 p.m...

diegel wrote:

This version of pango make firefox3 run a lot more stable.

I can confirm that, a site which was problematic previously (flickr.com) is now OK (IRIX 6.5.27, SGI O2 R10k). Thanks much!

_________________
:O2:

hamei
Who joined Feb. 24, 2004, 4:10 p.m.
and authored 8702 notes

Wrote the following at March 17, 2013, 2:21 a.m...

Is this Chicago ? Can I vote again ?

Seriously, after a couple days use, not one crash. This should maybe be a Firefox dependency in the future, it's that much better.

Thanks !

_________________
waiting for flight 1203 ...

canavan
Who joined July 21, 2003, 5:19 a.m.
and authored 687 notes

Wrote the following at March 17, 2013, 12:45 p.m...

I just tried building a current ImageMagick with neko_pango-1.14.10, but it doesn't build any pango support because the minimum version is 1.28.1. There's no Pango support inGraphicsMagick, so that's not a problem there.

hamei
Who joined Feb. 24, 2004, 4:10 p.m.
and authored 8702 notes

Wrote the following at March 17, 2013, 5:39 p.m...

canavan wrote:

Another reason to switch to Graphics Magick

It's painless, they even have a set of wrapper scripts to emulate ImageMagick so the change doesn't break anything.

I gave a newer pango a shot : it doesn't compile easily for a nooby. Now if someone skillful were to step in there ....

_________________
waiting for flight 1203 ...

diegel
Who joined Nov. 17, 2009, 2:08 a.m.
and authored 158 notes

Wrote the following at March 17, 2013, 11:27 p.m...

canavan wrote:

Another reason why I choose 1.14 is the fact that we have 1.12 in current and a lot more newer versions up to 1.28 in obsolete. The only explanation for me is that there is a problem with never versions. If someone can give some more details of this problems I may have a look at a current version of pango.

_________________
:Tezro:

hamei
Who joined Feb. 24, 2004, 4:10 p.m.
and authored 8702 notes

Wrote the following at March 18, 2013, 3:10 a.m...

diegel wrote:

I am not confident that you can get much information from the versions in obsolete. There were some tardists made for a while that were ... ummm ... "lacking" in the quality control department. I just threw 1.18 in the blender and got back a number of errors. Probably none of them are difficult but it was enough to stop me in my tracks.

If you want real feedback, you'll probbaly have to get it from your own experience

_________________
waiting for flight 1203 ...

diegel
Who joined Nov. 17, 2009, 2:08 a.m.
and authored 158 notes

Wrote the following at March 18, 2013, 9:16 a.m...

Just tried pango 1.28.4. It compiles and works for me. In opposite to 1.14 it contains c++ code and with this my gcc build will not be nekoware compliant. The problems of the previous pango packages are most probably the usual mips pro optimization failures. If someone wants to fix it feel free, I am out here.

_________________
:Tezro:

canavan
Who joined July 21, 2003, 5:19 a.m.
and authored 687 notes

Wrote the following at March 18, 2013, 9:25 a.m...

I have tried building pango 1.33.7, but that requires newer versions of cairo, harfbuzz, poppler and so on in a circular dependency hell.

hamei
Who joined Feb. 24, 2004, 4:10 p.m.
and authored 8702 notes

Wrote the following at March 18, 2013, 5:48 p.m...

diegel wrote:

In opposite to 1.14 it contains c++ code and with this my gcc build will not be nekoware compliant.

The scary part about gcc builds is not the religious aspect, but the things you find when doing a MIPSPro build of community software. The community apparently pretty much lacks q/c.

I am a turkey who knows nothing but on several occasions I have found places where gcc should have refused to compile the code. Places that were flat-out wrong. The code accidentally works under gcc but that's not a good thing ; a hidden cancer is not better than an obvious one ... and the people in charge at many community software projects are quite dismissive of real flaws. "It works fine here, just use gcc."

Yeah, well, that doesn't solve the underlying problem, does it ?

This is one reason I really like xpdf and Graphics Magick. Both those projects actually listen to problems and try to find solutions. Real solutions, not bandaids.

_________________
waiting for flight 1203 ...

foetz
Who joined April 14, 2003, 3:34 a.m.
and authored 4732 notes

Wrote the following at March 18, 2013, 7:23 p.m...

hamei wrote:

I am a turkey who knows nothing but on several occasions I have found places where gcc should have refused to compile the code. Places that were flat-out wrong. The code accidentally works under gcc but that's not a good thing ; a hidden cancer is not better than an obvious one ... and the people in charge at many community software projects are quite dismissive of real flaws. "It works fine here, just use gcc."

Yeah, well, that doesn't solve the underlying problem, does it ?

AMEN to that!
that goes for so many softwares today ... quality seems to have died together with most of the proprietary unix machines and the few ones left try to adopt to the crap instead of what'd be right :cry:

but hey, we got ipads now with hd video - hooray :twisted:

_________________
r-a-c.de

diegel
Who joined Nov. 17, 2009, 2:08 a.m.
and authored 153 notes

Wrote the following at March 18, 2013, 11:11 p.m...

hamei wrote:

I didn't talk about refused code, I am tired to solve problems that don't exist at O1 and show up at O2 or O3 optimization level. The optimization of gcc is much more reliable than mips pro. Of course I can build an O1 optimized pango 1.28, but it will be slow, even slower than 1.12.

_________________
:Tezro:

hamei
Who joined Feb. 24, 2004, 4:10 p.m.
and authored 8449 notes

Wrote the following at March 19, 2013, 5:37 a.m...

diegel wrote:

I didn't talk about refused code, I am tired to solve problems that don't exist at O1 and show up at O2 or O3 optimization level.

Code:

urchin 32% pwd

/usr/people/dev/pango-1.18.4

setenv CC cc

setenv CXX CC

setenv F77 f77

setenv CFLAGS '-mips4 -O1 -c99  -I/usr/nekoware/include -I/usr/include'

setenv CPPFLAGS '-mips4 -O1 -c99  -I/usr/nekoware/include -I/usr/include'

setenv CXXFLAGS '-mips4 -O1 -c99 -I/usr/nekoware/include -I/usr/include'

setenv LDFLAGS '-mips4 -L/usr/nekoware/lib -L/usr/lib32 -Wl,-rpath -Wl,/usr/nekoware/lib'

setenv LD_LIBRARY_PATH '/usr/nekoware/lib /usr/lib32'

setenv LD_LIBRARYN32_PATH '/usr/nekoware/lib /usr/lib32'

setenv MAKE gmake

./configure --prefix=/usr/nekoware

(stuff)

configure: creating ./config.status

config.status: creating Makefile

config.status: creating pango/Makefile

config.status: creating pango/mini-fribidi/Makefile

config.status: creating pango/opentype/Makefile

config.status: creating pango/pango.rc

config.status: creating pango/pangoft2.rc

config.status: creating pango/pangowin32.rc

config.status: creating pango-view/Makefile

config.status: creating modules/Makefile

config.status: creating modules/arabic/Makefile

config.status: creating modules/basic/Makefile

config.status: creating modules/hangul/Makefile

config.status: creating modules/hebrew/Makefile

config.status: creating modules/indic/Makefile

config.status: creating modules/khmer/Makefile

config.status: creating modules/syriac/Makefile

config.status: creating modules/thai/Makefile

config.status: creating modules/tibetan/Makefile

config.status: creating examples/Makefile

config.status: creating docs/Makefile

config.status: creating docs/version.xml

config.status: creating tools/Makefile

config.status: creating tests/Makefile

config.status: creating pango.pc

config.status: creating pangox.pc

config.status: creating pangowin32.pc

config.status: creating pangoft2.pc

config.status: creating pangoxft.pc

config.status: creating pangocairo.pc

config.status: creating pango-uninstalled.pc

config.status: creating pangox-uninstalled.pc

config.status: creating pangowin32-uninstalled.pc

config.status: creating pangoft2-uninstalled.pc

config.status: creating pangoxft-uninstalled.pc

config.status: creating pangocairo-uninstalled.pc

config.status: creating pango-zip.sh

config.status: creating tests/runtests.sh

config.status: creating config.h

config.status: config.h is unchanged

config.status: executing depfiles commands

config.status: executing pango/module-defs.h commands

config.status: executing pango/module-defs-x.c commands

config.status: executing pango/module-defs-fc.c commands

config.status: executing pango/module-defs-win32.c commands

config.status: executing pango/module-defs-atsui.c commands

config.status: executing pango/module-defs-lang.c commands

config.status: executing pango/pango-features.h commands

config.status: creating pango/pango-features.h

config.status: pango/pango-features.h is unchanged

configuration:

backends: FreeType X Xft Cairo

urchin 34% gmake

gmake  all-recursive

gmake[1]: Entering directory `/usr/people/dev/pango-1.18.4'

Making all in pango

gmake[2]: Entering directory `/usr/people/dev/pango-1.18.4/pango'

( cd . && glib-mkenums \

--fhead "#ifndef __PANGO_ENUM_TYPES_H__\n#define __PANGO_ENUM_TYPES_H__\n\n#include <glib-object.h>\n\nG_BEGIN_DECLS\n" \

--fprod "/* enumerations from \"@filename@\" */\n" \

--vhead "GType @enum_name@_get_type (void);\n#define PANGO_TYPE_@ENUMSHORT@ (@enum_name@_get_type())\n"         \

--ftail "G_END_DECLS\n\n#endif /* __PANGO_ENUM_TYPES_H__ */" \

pango.h pango-attributes.h pango-break.h pango-context.h pango-coverage.h pango-engine.h pango-font.h pango-fontmap.h pango-fontset.h pango-glyph.h pango-glyph-item.h pango-gravity.h pango-item.h pango-language.h pango-layout.h pango-matrix.h pango-modules.h pango-renderer.h pango-script.h pango-tabs.h pango-types.h pango-utils.h ) > tmp-pango-enum-types.h \

&& (cmp -s tmp-pango-enum-types.h pango-enum-types.h || cp tmp-pango-enum-types.h pango-enum-types.h ) \

&& rm -f tmp-pango-enum-types.h \

&& echo timestamp > s-enum-types-h

/bin/sh: glib-mkenums:  not found

gmake[2]: *** [s-enum-types-h] Error 127

gmake[2]: Leaving directory `/usr/people/dev/pango-1.18.4/pango'

gmake[1]: *** [all-recursive] Error 1

gmake[1]: Leaving directory `/usr/people/dev/pango-1.18.4'

gmake: *** [all] Error 2

I'm not going to say that I know what I'm doing but it configures with no problem, then crashes at the very first instruction. Optimization is set to -O1.

It's kind of frustrating. This sort of thing is not that uncommon. I have not had any problems that I know of with -O3 but in future, I'll try that when there is an unexplained problem. Thank you for the pointer.

_________________
lemon tree very pretty and the flower very sweet ...

diegel
Who joined Nov. 17, 2009, 2:08 a.m.
and authored 153 notes

Wrote the following at March 19, 2013, 6:34 a.m...

hamei wrote:

It's kind of frustrating. This sort of thing is not that uncommon. I have not had any problems that I know of with -O3 but in future, I'll try that when there is an unexplained problem. Thank you for the pointer.

Compiles without errors on the first try here. Here is my standard build environment:

Code:

export CC=c99

export CXX=CC

export CFLAGS="-O1 -OPT:Olimit=0:roundoff=3:IEEE_arithmetic=3:alias=TYPED -mips4 -n32 -INLINE -woff 1174"

export CXXFLAGS="-O1 -OPT:Olimit=0:roundoff=3:IEEE_arithmetic=3:alias=TYPED -mips4 -n32 -Zf,_245 -INLINE -woff 1110,1171,1201,1355,3201"

export LDFLAGS="-L/usr/nekoware/lib"

export PERL="/usr/nekoware/bin/perl"

export CPPFLAGS="-I/usr/nekoware/include -I/usr/nekoware/include/glib-2.0"

_________________
:Tezro:

jpstewart
Who joined Sept. 21, 2010, 2:31 p.m.
and authored 165 notes

Wrote the following at March 19, 2013, 7:02 a.m...

hamei wrote:

Code:

/bin/sh: glib-mkenums: not found

...
I'm not going to say that I know what I'm doing but it configures with no problem, then crashes at the very first instruction.

It's telling you that glib-mkenums is not installed. You'll need the neko_glib package installed (including the development stuff which might not be installed by default) to build pango. It's a shame that the configure step doesn't actually check for the presence of glib-mkenums!

Also (from my own past experience), glib-mkenums is a Perl script which might not run so well with IRIX perl. You might need to change its first line to use Nekoware perl instead of /usr/bin/perl5 (if you get more strange errors from glib-mkenums). That same advice applied to tools/gen-color-table.pl from the Pango source when I tried to compile pango-1.29.5. I had to change it to use a newer Perl, according to my notes from a past IRIX porting project. (Although my notes don't say why I felt I had to do that.) IRIX perl is just plain ancient.

_________________

Sun SPARCstation 20, Blade 2500
HP C8000

hamei
Who joined Feb. 24, 2004, 4:10 p.m.
and authored 8449 notes

Wrote the following at March 19, 2013, 8:35 p.m...

This is pretty much exactly what I was talking about concerning the problems with gcc .. or, more accurately, some gcc users :

diegel wrote:

... environment settings with /usr/nekoware/include/glib-2.0 added ...

Thought this was a good idea for a dumb problem but it didn't work. No luck, still

Code:

/bin/sh: glib-mkenums: not found

Someone please tell me what's the point of a configure script if it doesn't work ? How is it that every other application in the universe can find the glib includes except for this moron-written piece of shit ? How is it that a configuration script can happily create the Makefiles without even mentioning that it wasn't able to find a major component ? But oh man, it did create

Code:

config.status: creating modules/arabic/Makefile

config.status: creating modules/basic/Makefile

config.status: creating modules/hangul/Makefile

config.status: creating modules/hebrew/Makefile

config.status: creating modules/indic/Makefile

config.status: creating modules/khmer/Makefile

config.status: creating modules/syriac/Makefile

config.status: creating modules/thai/Makefile

config.status: creating modules/tibetan/Makefile

I'm so happy ! look, mom ! configure crated Makefiles for hangul ! and thai ! and khmer ! and tibetan ! We are so international ! and tolerant of other cultures ! Diversity ! we be so k00l and l33t !

Too bad the fucking useless piece of shit couldn't find glib, which was the actual point of the operation.

It makes me sad to think that those no-talent little twats are using up food and air that would be better served by feeeding the world's population of dung beetles

jpstewart wrote:

It's telling you that glib-mkenums is not installed.

And I'm telling it that their useless configure script is too retarded to find it with a map glued to their nose.

Quote:

You'll need the neko_glib package installed (including the development stuff which might not be installed by default) to build pango.

Code:

urchin 6% pwd

/usr/nekoware/include/glib-2.0

urchin 7% ls -l

total 24

drwxr-xr-x    2 root     sys          4096 Jan 30  2011 gio

drwxr-xr-x    2 root     sys          4096 Jan 30  2011 glib

-rw-r--r--    1 root     sys          1511 Jan 30  2011 glib-object.h

-rw-r--r--    1 root     sys          2910 Jan 30  2011 glib.h

-rw-r--r--    1 root     sys          3613 Jan 30  2011 gmodule.h

drwxr-xr-x    2 root     sys          4096 Jan 30  2011 gobject

Quote:

It's a shame that the configure step doesn't actually check for the presence of glib-mkenums!

Yes. It's a shame that the people involved in this software are so stupid. I suppose it's better than having them run a non-profit to help the disadvantaged or deal with the poor self-image of battered orangutans ... but not much.

Quote:

You might need to change its first line to use Nekoware perl instead of /usr/bin/perl5

Code:

urchin 15% which perl

/usr/nekoware/bin/perl

I suppose their Down's baby script is too advanced and modern to use traditional Unix concepts ? Such as the path statement ? Or does the Slimy Slug release of Droolaroo Loonnix not support the deprecated 'legacy' non- mooodern concept of a path ?

jp, I will accept that I don't know what I am doing. But this crap is beyond retarded. These people have the mental ability of a dead skunk in the middle of the road. I bet they flunked kindergarten. Useless, drooling morons. You name it, it's wrong with this project.

This is exactly why, until diegel got FF3 to run, I was unwilling to get within 100 yards of gtk2. Those people should be placed in a weighted burlap sack and dropped off a bridge into the river. The amount of air and food on earth is limited. We shouldn't waste it on this no-talent scum.

_________________
lemon tree very pretty and the flower very sweet ...