View previous topic :: View next topic |
Author |
Message |
steveL Watchman
Joined: 13 Sep 2006 Posts: 5153 Location: The Peanut Gallery
|
Posted: Tue Sep 16, 2008 3:36 am Post subject: |
|
|
steveb wrote: | Wrong. If you ever have build Gentoo for an embedded system, then you will see that bash is not the number one choice for an shell. |
True, but all ebuilds are in bash, and eclasses in the tree are supposed to be able to rely on bash being available; forcing that to sh compatibility isn't the right way to get embedded builds imo, and only slows down the rest. A serious embedded developer doesn't build on the target ime. The only other use-case I've heard of that could possibly use it, is building for a PDA or a phone, and again I'd much rather see better desktop/host tools for that, than slowing ebuild/eclass devs down. If you really want portable sh, you don't even use [ a.b. wrote: | Honestly, I haven't even thought about the possibility that a bashrc might be sourced by anything other than bash so I didn't think about that while writing.
And in at least one place, I could replace use bash's '=~' operator to replace an ugly "cat | grep" construct which I find much worse. |
That's called a useless use of cat; grep 'foo' file instead of cat file | grep 'foo'. (I am not looking at the original script, but cat | grep is always one of those ime, and you sound like a beginning shell-scripter; the url will be useful to you, if so.)
Quote: | Code: | while read atom data; do
if matchatom "${atom}"; then
if [ ${append} -eq 0 ]; then
ret="${data%%#*}"
else
ret="${ret} ${data%%#*}"
fi
fi
done < ${1} |
Does that kind of expression have a name I can google for? |
It's called Parameter Expansion. This FAQ shows what you can do with it (the array stuff will not work in sh.)
BTW wrapping vars in unnecessary braces is a waste of time. You need to quote "$1" there however as it's a filename. You don't need to quote ret=${data%%#*} as word-splitting does not take place in assignments. You do with the second one as there is a space there. Note you can also use ret+=$foo in general (again the space there would necessitate quoting.)
You can also use if ((append)); then which would swap the meaning of the condition or if ((!append)); if you wanted to keep it as-is. (Assuming append is an integer string.)
http://wooledge.org/mywiki/BashGuide is the place to start if you want to know #bash.
edit: fixed [
Last edited by steveL on Fri Sep 19, 2008 7:31 am; edited 1 time in total |
|
Back to top |
|
|
steveb Advocate
Joined: 18 Sep 2002 Posts: 4564
|
Posted: Tue Sep 16, 2008 4:40 am Post subject: |
|
|
steveL wrote: | True, but all ebuilds are in bash, and eclasses in the tree are supposed to be able to rely on bash being available; forcing that to sh compatibility isn't the right way to get embedded builds imo, and only slows down the rest. A serious embedded developer doesn't build on the target ime. The only other use-case I've heard of that could possibly use it, is building for a PDA or a phone, and again I'd much rather see better desktop/host tools for that, than slowing ebuild/eclass devs down. If you really want portable sh, you don't even use | Gentoo on BSD has some of the restrictions as well. So if you want a script to run on almost all of the Gentoo supported platforms, then go as portable as possible. I personally don't see a reason why not trying to support the others as well. And using the external commands in bashrc in Portage gives you errors/warnings. So why ignoring them?
// SteveB |
|
Back to top |
|
|
steveL Watchman
Joined: 13 Sep 2006 Posts: 5153 Location: The Peanut Gallery
|
Posted: Tue Sep 16, 2008 12:10 pm Post subject: |
|
|
steveb wrote: | steveL wrote: | True, but all ebuilds are in bash, and eclasses in the tree are supposed to be able to rely on bash being available; forcing that to sh compatibility isn't the right way to get embedded builds imo, and only slows down the rest. A serious embedded developer doesn't build on the target ime. The only other use-case I've heard of that could possibly use it, is building for a PDA or a phone, and again I'd much rather see better desktop/host tools for that, than slowing ebuild/eclass devs down. If you really want portable sh, you don't even use [ |
Gentoo on BSD has some of the restrictions as well. So if you want a script to run on almost all of the Gentoo supported platforms, then go as portable as possible. |
If you can define the restrictions clearly and one of them isn't "no bash." Bash is portable in the same way GNU make is. It might not be that fast, but it's definitely fast enough (that script needs major optimisation, which we'll do for the next version, but it works fine as-is on my old desktop and screams on my laptop) and it might be bloated, but it's not that bloated. When you look at the range of tools that get pulled in to build all these packages from source, bash is really not a problem, imo. And binpkgs can simply be untarred into the root of the target.
Quote: | I personally don't see a reason why not trying to support the others as well. And using the external commands in bashrc in Portage gives you errors/warnings. So why ignoring them? |
Well to be totally honest, I've never really used bashrc, so I don't have any experience of that situation; I was just answering wrt the "no bash" debate. It's much easier to maintain and write clean bash[1], especially for things like libs, than it is to maintain and write truly portable sh. Plus you can do a lot more without having to resort to eval etc. Yes you can argue that it should all be wrapped in an EAPI, so what difference does it make if hardly any ebuild devs can work with it? I'd argue that just makes it even harder for people to get involved, and doesn't give any long-term benefit in terms of use-cases, while also making the scripts a pig to maintain, when we all know devs are overstretched as it is.
I do agree with you however wrt non-GNU userland and the options that are used there. Personally I think there's a strong case for sticking to POSIX options wrt command line utilities, and allowing aliases or variables for say sed -r vs sed -E I think it is on BSD. (ie where the ebuild dev explicitly asks for an ERE.)
[1] Note that I don't actually consider Gentoo in-house style to be clean bash. ${var} is simply yuck in the general case, and often leads beginners into thinking they don't need quotes, and gives zero benefit but needs more typing (with awkward shift'ing which is really annoying when you know it doesn't do anything. Well it is for me, anyhow.) As for == in a shell script outside (( or [[ -n "$foo" ]] .. ;) |
|
Back to top |
|
|
steveb Advocate
Joined: 18 Sep 2002 Posts: 4564
|
Posted: Tue Sep 16, 2008 6:01 pm Post subject: |
|
|
SteveL! I did not wanted to attack you. In no way. I just said, that not everything is pure Bash in Gentoo. Look for example at Getnoo Bugzilla Bug 230690 and Getnoo Bugzilla Bug 235152. See that info about "let i++"? And it's such small things to which I am referring. Avoiding them does not cost much time and helps others as well. You are right about portability of pure Bash scripts. But that's not the whole game. I contributed some time ago a Bash script for a package and then some one complained that the Bash script does not work on his embedded Gentoo because he does not have Bash there. He could install Bash and things would be okay. But he has his reasons (which are from my viewpoint 100% valid for an embedded installation) and so I took the time to make the very small changes in order to get the script run under pure sh. It's not that much work. Especially if you are the creator of the original script, then changing it to be runnable under pure sh is as easy as 1-2-3.
But hey! Who am I? Nobody! No one needs to follow my advice. I just said that if a.b does not want Portage to bark about external used commands, then he should try to reduce the need for external commands. That's it. Nothing fancy.
// SteveB[/bug] |
|
Back to top |
|
|
steveL Watchman
Joined: 13 Sep 2006 Posts: 5153 Location: The Peanut Gallery
|
Posted: Wed Sep 17, 2008 12:38 pm Post subject: |
|
|
SteveB it's cool I don't feel under attack from you at all :-) Sorry for sounding defensive.
i=$(expr $i + 1) is just awful; if you want to replace let i++ in POSIX sh I'd do i=$(($i+1)) since it doesn't require a subshell, nor a call to an external.
And yeah, scripts for a package itself, as opposed to for building the package, should be sh-compatible unless they are, for instance, to support a GUI app. initscripts must not require BASH sh, for example. == outside Arithmetic context works in sh when it's linked to BASH, but not in standard POSIX, and is usually an indication that the author doesn't really grok shell scripting ime.
steveb wrote: | I just said that if a.b does not want Portage to bark about external used commands, then he should try to reduce the need for external commands. That's it. Nothing fancy. |
Yeah, I think that's why I became a bit defensive; I'd missed that bit of the conversation, and jumped in at the end about BASH vs SH. Sorry about that. (And I agree, of course.)
Ah found the link which best explains just how nasty truly portable sh gets. Reading that is very useful, I do recommend it, but I really do not want to be writing scripts like that in general, and if the package manager already needs python, or C++ and boost, you're way out of embedded space already (even C++ alone is not usually considered for embedded systems) and thus doing the standard "build on a host for a target" thing. Focussing on making that work properly is a much better use of developer time, imo.
(TBH even the thought of a shellscript interpreter for an embedded system seems wrong to me, but then I am old ;)
Another type of ebuild, eg a pbuild for python, could easily restrict the sh used since much more of the logic would be in python, and sh would only be called minimally, as in a Makefile, if at all. In that case, sticking to POSIX sh would be useful and would not impact on the maintainability or flexibility of the resultant builds or libs. |
|
Back to top |
|
|
Thaidog Veteran
Joined: 19 May 2004 Posts: 1053
|
Posted: Mon Nov 10, 2008 3:05 pm Post subject: |
|
|
I do not have a "/etc/portage/bashrc" file in my system. Should I make one or is that something regarding an older system type? _________________ Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
Back to top |
|
|
mattst88 Developer
Joined: 28 Oct 2004 Posts: 422
|
Posted: Mon Nov 10, 2008 4:22 pm Post subject: |
|
|
Thaidog wrote: | I do not have a "/etc/portage/bashrc" file in my system. Should I make one or is that something regarding an older system type? |
Yes, you should create it. _________________ My Wiki page |
|
Back to top |
|
|
Thaidog Veteran
Joined: 19 May 2004 Posts: 1053
|
Posted: Mon Nov 10, 2008 5:06 pm Post subject: |
|
|
|mattst88| wrote: | Thaidog wrote: | I do not have a "/etc/portage/bashrc" file in my system. Should I make one or is that something regarding an older system type? |
Yes, you should create it. |
Thanks! It's working now. Do you know if the kernel will automatically compile correctly with Icc or do I need to apply a patch to it? _________________ Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds
Last edited by Thaidog on Mon Nov 10, 2008 8:20 pm; edited 1 time in total |
|
Back to top |
|
|
Thaidog Veteran
Joined: 19 May 2004 Posts: 1053
|
Posted: Mon Nov 10, 2008 5:35 pm Post subject: |
|
|
Well now everything I try to compile is failing with the exception of python. I am trying the "safe" packages and most errors say "emake failed". firefox 3 died with no error message.
Edit: the flags are: ICCCFLAGS="-O2 -xW -tpp7 -gcc"
I added -gcc and now some "safe" packages are comping but I screwed up nss for firefox and the 3.12 version is not making it:
Generating DSA Key Pair..../var/tmp/portage/dev-libs/nss-3.12/work/nss-3.12/mozilla/security/nss/cmd/shlibsign/./sign.sh: line 2: 20899 Segmentation fault ${2}/shlibsign -v -i ${5}
make[2]: *** [../../../dist/Linux2.6_x86_icc_glibc_PTH_OPT.OBJ/lib/libsoftokn3.chk] Error 139
make[2]: Leaving directory `/var/tmp/portage/dev-libs/nss-3.12/work/nss-3.12/mozilla/security/nss/cmd/shlibsign'
make[1]: *** [libs] Error 2
make[1]: Leaving directory `/var/tmp/portage/dev-libs/nss-3.12/work/nss-3.12/mozilla/security/nss/cmd'
make: *** [libs] Error 2
"ERROR: dev-libs/nss-3.12 failed.
* Call stack:
* ebuild.sh, line 49: Called src_compile
* environment, line 2400: Called die
* The specific snippet of code:
* emake -j1 BUILD_OPT=1 XCFLAGS="${CFLAGS}" CC="$(tc-getCC)" || die "nss make failed"
* The die message:
* nss make failed" _________________ Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
Back to top |
|
|
ZeLegolas Tux's lil' helper
Joined: 26 Apr 2005 Posts: 128
|
Posted: Thu Nov 13, 2008 9:17 pm Post subject: Re: [HOWTO] Intel C++ Compiler and per-package CFLAGS |
|
|
|mattst88| wrote: | I've rewritten Gentoo-Wiki's HOWTO on using the Intel C++ Compiler.
http://gentoo-wiki.com/HOWTO_Intel_C_Compiler
The old guide was terribly written and horribly messy.
It also made Portage compile with icc unless you explicitly told it to use gcc. This is bad. Far too many packages don't work with icc to use this behavior. My guide makes Portage compile only specified packages with icc and all others with gcc which will save a few headaches in the long run. It also will give your system the ability to use per-package CFLAGS for gcc and icc!
Need feedback on the guide. Suggestions are welcome. |
http://gentoo-wiki.com/HOWTO_Intel_C_Compiler is gone |
|
Back to top |
|
|
mattst88 Developer
Joined: 28 Oct 2004 Posts: 422
|
Posted: Thu Nov 13, 2008 9:22 pm Post subject: Re: [HOWTO] Intel C++ Compiler and per-package CFLAGS |
|
|
ZeLegolas wrote: | |mattst88| wrote: | I've rewritten Gentoo-Wiki's HOWTO on using the Intel C++ Compiler.
http://gentoo-wiki.com/HOWTO_Intel_C_Compiler
The old guide was terribly written and horribly messy.
It also made Portage compile with icc unless you explicitly told it to use gcc. This is bad. Far too many packages don't work with icc to use this behavior. My guide makes Portage compile only specified packages with icc and all others with gcc which will save a few headaches in the long run. It also will give your system the ability to use per-package CFLAGS for gcc and icc!
Need feedback on the guide. Suggestions are welcome. |
http://gentoo-wiki.com/HOWTO_Intel_C_Compiler is gone |
Yep. gentoo-wiki lost all its content.
I'm not planning to rewrite it any time soon.
This definitely shows the importance of off site backups. Mike Valstar: what the hell man? I know it wasn't your fault, but no backups?! _________________ My Wiki page |
|
Back to top |
|
|
Thaidog Veteran
Joined: 19 May 2004 Posts: 1053
|
|
Back to top |
|
|
mattst88 Developer
Joined: 28 Oct 2004 Posts: 422
|
Posted: Thu Nov 13, 2008 9:55 pm Post subject: |
|
|
If someone is interested in getting the cached version back into Gentoo Wiki, then be my guest. _________________ My Wiki page |
|
Back to top |
|
|
ZeLegolas Tux's lil' helper
Joined: 26 Apr 2005 Posts: 128
|
Posted: Fri Nov 14, 2008 4:17 am Post subject: Re: [HOWTO] Intel C++ Compiler and per-package CFLAGS |
|
|
Thanks
When I execute emerge for icc i have this:
Code: | $ emerge -pv icc
These are the packages that would be merged, in order:
Calculating dependencies... done!
[ebuild NS ] sys-devel/gcc-3.3.6-r1 [4.3.2] USE="doc fortran (multilib) nls objc (-altivec) -bootstrap -boundschecking -build-gcj -gtk (-hardened) -ip28 -ip32r10k -libffi -multislot (-n32) (-n64) -nocxx -nopie -nossp -test -vanilla" 23,534 kB
[ebuild N ] virtual/libstdc++-3.3 0 kB
[ebuild N ] dev-lang/icc-10.1.018 39,769 kB
Total: 3 packages (2 new, 1 in new slot), Size of downloads: 63,303 kB |
Do I have to install gcc 3 to emerge icc???? |
|
Back to top |
|
|
Thaidog Veteran
Joined: 19 May 2004 Posts: 1053
|
Posted: Fri Nov 14, 2008 11:42 am Post subject: |
|
|
|mattst88| wrote: | If someone is interested in getting the cached version back into Gentoo Wiki, then be my guest. |
I don't mind adding that one back... just need to be pointed in the right direction on how to do it. Is the server back up yet? _________________ Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
Back to top |
|
|
juantxorena Apprentice
Joined: 19 Mar 2006 Posts: 201 Location: The Shire
|
Posted: Sun Nov 16, 2008 6:00 pm Post subject: Re: [HOWTO] Intel C++ Compiler and per-package CFLAGS |
|
|
ZeLegolas wrote: | Do I have to install gcc 3 to emerge icc???? |
Nope, icc depends on virtual/libstdc++, which depends on sys-libs/libstdc++-v3, sys-libs/libstdc++-v3-bin OR sys-devel/gcc-3.3*. You should install sys-libs/libstdc++-v3 (with --oneshot flag), and then install icc as usual, it won't pull gcc anymore.
I had this problem some days ago and I don't understand why virtual/libstdc++ try to install gcc first. Somebody can enlighten me? _________________ I cannot write English very well. Please, correct any mistake so that I can improve. |
|
Back to top |
|
|
thedarave n00b
Joined: 21 Jul 2004 Posts: 19
|
Posted: Fri Dec 05, 2008 10:06 pm Post subject: |
|
|
Any timeline on when ICC 11.0 will hit the portage tree or should I just submit a bug with an ebuild for it?
As for the virtual/libstdc++-v3, that's probably due to when that ebuild was made. It probably should get revised on x86/amd64 to invert the order. There's still platforms that use gcc-3.3 as the default compilier (I think) which would be why the gcc-3.3 install would be the default. A little rewriting of the ebuild should fix that. _________________ -Steve- |
|
Back to top |
|
|
thedarave n00b
Joined: 21 Jul 2004 Posts: 19
|
Posted: Fri Dec 05, 2008 10:11 pm Post subject: |
|
|
Here's a toy for those that can understand what this is trying to do. If you don't understand what it's doing/does or how to work it, its probably not meant for you.
Yes, its a silly little item, but it fills a slight void I've encountered.
Code: |
#!/bin/bash
clear
export OPTLEVEL=2
export IPOLEVEL=2
export FPLEVEL=1
export PARALEVEL=1
export GCCVLEVEL=2
if [[ `cat /proc/cpuinfo | grep -c " sse4"` -ne 0 ]]
then export INSDFLAG=" -xS"
elif [[ `cat /proc/cpuinfo | grep -c " ssse3"` -ne 0 ]]
then export INSDFLAG=" -xT"
elif [[ `cat /proc/cpuinfo | grep -c " sse3"` -ne 0 ]]
then export INSDFLAG=" -xP"
elif [[ `cat /proc/cpuinfo | grep -c " sse2"` -ne 0 ]]
then export INSDFLAG=" -xW"
elif [[ `cat /proc/cpuinfo | grep -c " sse "` -ne 0 ]]
then export INSDFLAG=" -xK"
else export INSDFLAG=" -axS"
fi
export INSLEVEL=1
grep -v "$1 " /etc/portage/package.icc-cflags > /etc/portage/package.icc-cflags.temp
while [[ 0 -eq 0 ]]
do
if [[ ${OPTLEVEL} -eq 2 ]]
then
export OPTFLAG="-O3"
elif [[ ${OPTLEVEL} -eq 1 ]]
then
export OPTFLAG="-O2"
else
export OPTFLAG="-O1"
fi
if [[ ${IPOLEVEL} -eq 2 ]]
then
export IPOFLAG=" -ipo"
elif [[ ${IPOLEVEL} -eq 1 ]]
then
export IPOFLAG=" -ip"
else
export IPOFLAG=""
fi
if [[ ${FPLEVEL} -eq 1 ]]
then
export FPFLAG=" -fomit-frame-pointer"
else
export FPFLAG=""
fi
if [[ $PARALEVEL -eq 1 ]]
then
export PARAFLAG=" -parallel"
else
export PARAFLAG=""
fi
if [[ $INSLEVEL -eq 1 ]]
then
export INSFLAG="${INSDFLAG}"
else
export INSFLAG=""
fi
if [[ $GCCVLEVEL -eq 2 ]]
then
export GCCVFLAG=""
elif [[ $GCCVLEVEL -eq 1 ]]
then
export GCCVFLAG=" -gcc-version=420"
else
export GCCVFLAG=" -gcc-version=330"
fi
echo $1 ${OPTFLAG}${IPOFLAG}${FPFLAG}${INSFLAG}${PARAFLAG} -gcc${GCCVFLAG} > /etc/portage/package.icc-cflags
emerge -1 $1
if [[ $? -eq 0 ]]
then
cat /etc/portage/package.icc-cflags /etc/portage/package.icc-cflags.temp | sort -u > /etc/portage/package.icc-cflags.new
rm -f /etc/protage/package.icc-cflags /etc/portage/package.icc-flags.temp
mv /etc/portage/package.icc-cflags.new /etc/portage/package.icc-cflags
exit 0
else
if [[ ${IPOLEVEL} -gt 0 ]]
then
export IPOLEVEL=$(( ${IPOLEVEL} - 1 ))
else
export IPOLEVEL=2
if [[ ${PARALEVEL} -gt 0 ]]
then
export PARALEVEL=$(( ${PARALEVEL} - 1 ))
else
export PARALEVEL=1
if [[ ${FPLEVEL} -gt 0 ]]
then
export FPLEVEL=$(( ${FPLEVEL} - 1 ))
else
export FPLEVEL=1
if [[ ${GCCVLEVEL} -gt 0 ]]
then
export GCCVLEVEL=$(( ${GCCVLEVEL} - 1 ))
else
export GCCVLEVEL=2
if [[ ${OPTLEVEL} -gt 0 ]]
then
export OPTLEVEL=$(( ${OPTLEVEL} - 1 ))
else
grep -vx $1 /etc/portage/package.icc | sort -u > /etc/portage/package.icc.temp
mv -f /etc/portage/package.icc.temp /etc/portage/package.icc
mv -f /etc/portage/package.icc-cflags.temp /etc/portage/package.icc-cflags
rm -rf /var/tmp/portage/$1*
exit 1
fi
fi
fi
fi
fi
fi
done
|
Take it or leave it, but I'm not answering questions about it! _________________ -Steve- |
|
Back to top |
|
|
tranquilcool Veteran
Joined: 25 Mar 2005 Posts: 1179
|
Posted: Mon Feb 16, 2009 2:33 am Post subject: |
|
|
icc is trying to install outside sandbox.
anybody knows what's happening?
packages can only be installed with FEATURES="-sandbox" without
ACCESS DENIED errors. _________________ this is a strange strange world. |
|
Back to top |
|
|
mattst88 Developer
Joined: 28 Oct 2004 Posts: 422
|
Posted: Tue Feb 17, 2009 8:06 pm Post subject: |
|
|
tranquilcool wrote: | icc is trying to install outside sandbox.
anybody knows what's happening?
packages can only be installed with FEATURES="-sandbox" without
ACCESS DENIED errors. |
Yes, this is noted in bug 246516.
I think the best thing to do at this point is stick with icc 10.1.xxx.
At some point, I'll also get around to transplanting the guide onto my website so we don't lose it again. _________________ My Wiki page |
|
Back to top |
|
|
Thaidog Veteran
Joined: 19 May 2004 Posts: 1053
|
|
Back to top |
|
|
rufnut Apprentice
Joined: 16 May 2005 Posts: 247
|
Posted: Thu Feb 26, 2009 8:13 am Post subject: |
|
|
Thaidog wrote: | If anyone wants to compile their kernel with ICC please check out my project: |
Thanks
I thought you had given up for a while.
This ought to be interesting and may eventually be a good little kick for all those Atom based machines out there.
|
|
Back to top |
|
|
Thaidog Veteran
Joined: 19 May 2004 Posts: 1053
|
Posted: Thu Feb 26, 2009 8:51 pm Post subject: |
|
|
rufnut wrote: | Thaidog wrote: | If anyone wants to compile their kernel with ICC please check out my project: |
Thanks
I thought you had given up for a while.
This ought to be interesting and may eventually be a good little kick for all those Atom based machines out there.
|
Down but never out _________________ Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
Back to top |
|
|
tranquilcool Veteran
Joined: 25 Mar 2005 Posts: 1179
|
Posted: Fri Feb 27, 2009 9:02 am Post subject: |
|
|
Thaidog wrote: | rufnut wrote: | Thaidog wrote: | If anyone wants to compile their kernel with ICC please check out my project: |
Thanks
I thought you had given up for a while.
This ought to be interesting and may eventually be a good little kick for all those Atom based machines out there.
|
Down but never out |
i have tried to compile the kernel in many ways but it fails
with the pid.o errors.
Mod edit: Answered here in the support thread (see next post). --timeBandit _________________ this is a strange strange world. |
|
Back to top |
|
|
timeBandit Bodhisattva
Joined: 31 Dec 2004 Posts: 2719 Location: here, there or in transit
|
Posted: Tue Apr 28, 2009 4:07 am Post subject: |
|
|
This has somehow slipped under the radar, for about a hundred replies. To wit: Guidelines for Tips & Tricks wrote: | - This is *not* a support forum. Do not ask questions here. The moderators and myself would like to stress this point, if you do ask a question, your post will be deleted. Use the other forums to ask questions, if you are asking a question regarding a tip or trick, you are encouraged to include a link to the tip/trick you are referring to make it easier for other to help you.
| I've split off recent discussion (to provide some context) to a new thread for support on this HOWTO. If you have read this far and still have questions, please post them there. _________________ Plants are pithy, brooks tend to babble--I'm content to lie between them.
Super-short f.g.o checklist: Search first, strip comments, mark solved, help others. |
|
Back to top |
|
|
|