Mac G5 X11 performance

[Mark Millard]

On 2019-May-23, at 17:10, Jason Bacon <bacon4000 at gmail.com> wrote:


> Netsurf works fine, but unfortunately depends on a legacy libutf8proc, which conflicts with the more modern utf8proc required by many other ports (including subversion).
> 
> All firefox dependencies build ("make depends" runs clean), so we're very close.
> 
> BTW, there are ~27,000 binary packages for ppc64, ~31,000 for amd64.  Not bad, and that number will jump when a few key ports like qt5-core are fixed.

I'm able to build devel/qt5-core based on:

# svnlite diff /usr/ports/Mk/Uses/
Index: /usr/ports/Mk/Uses/qt-dist.mk
===================================================================
--- /usr/ports/Mk/Uses/qt-dist.mk	(revision 501994)
+++ /usr/ports/Mk/Uses/qt-dist.mk	(working copy)
@@ -251,7 +251,7 @@
 .if ${_QT_VER:M5}
 post-patch: gcc-post-patch
 gcc-post-patch:
-	${REINPLACE_CMD} 's|%%LOCALBASE%%|${LOCALBASE}|' \
+	${REINPLACE_CMD} 's|%%LOCALBASE%%|${LOCALBASE}|g' \
 		${WRKSRC}/mkspecs/common/gcc-base.conf \
 		${WRKSRC}/mkspecs/freebsd-g++/qmake.conf
 	${REINPLACE_CMD} 's|%%GCC_DEFAULT%%|${GCC_DEFAULT}|g' \

I noticed that my build logs had %%LOCALBASE%% text that had
not been replaced, in particular the 2nd one on a particular
line. That lead to noticing the missing "g".

(I'm not a ports committer or maintainer.)

I experiment with a system-clang-8 and base/binutils based
powerpc64 environment, instead of gcc 4.2.1 based. So direct
inference of others being able to do the same would not
apply. But it seems likely to me.


Side note:

qt-dist.mk sometimes uses -e with ${REINPLACE_CMD} even when
only one command is being fed to sed, a context in which the -e
is optional.

> On 2019-05-23 14:02, Francis Little wrote:
>> I find WindowMaker fairly responsive.
>> 
>> I've not had a lot of luck with a web browser yet, Midori builds and comes up but cannot load webpages for some reason.
>> 
>> 
>> 
>> 
>> On Thu, 23 May 2019, 19:25 Jason Bacon, <bacon4000 at gmail.com <mailto:bacon4000 at gmail.com>> wrote:
>> 
>> 
>>    This is not a very important issue, but thought I'd bring in up in
>>    case
>>    there's a simple solution.
>> 
>>    I have an iMac G5, used mainly for port testing, but I've found that
>>    it's very close to being a usable desktop system (running SLiM login
>>    manager and LXDE as other mainstream desktop environments don't
>>    currently build).  All it would need is a few more key ports fixed
>>    (e.g.
>>    firefox or chromium, qt5-*) and improved X11 performance.
>> 
>>    X11 performance is slow to the point where a dragged window won't
>>    even
>>    keep up with the mouse.
>> 
>>    Chipset and Xorg info is below.
>> 
>>    Loading radeonkms causes a complete system freeze.
>> 
>>    I haven't found anything on the freebsd-ppc pages or elsewhere to
>>    help.
>> 
>>    If anyone knows of any ways to improve the situation, I'd appreciate
>>    hearing them.  If they're not too hackish, I would add support to
>>    sysutils/desktop-installer so others with PowerPC Macs can configure
>>    them with minimal effort.
>> 
>>    vgapci0 at pci0:0:16:0:    class=0x030000 card=0x41501002
>>    chip=0x41501002
>>    rev=0x00 hdr=0x00
>>         vendor     = 'Advanced Micro Devices, Inc. [AMD/ATI]'
>>         device     = 'RV350 [Radeon 9550/9600/X1050 Series]'
>>         class      = display
>>         subclass   = VGA
>> 
>>    [    66.665]
>>    X.Org X Server 1.18.4
>>  . . .
>> 
> 


===
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)