<WadeBerrier[m]>
any recommendations about distributing layers between vendors and customers for private source? ie: do you provide the layers and all the tarballs for each recipe? Shared source control access? Other ideas about "best practices" for this scenario?
troth has quit [Ping timeout: 268 seconds]
troth has joined #yocto
xmn has joined #yocto
dagmcr has quit [Ping timeout: 245 seconds]
dagmcr has joined #yocto
rmmr has quit [Ping timeout: 245 seconds]
goliath has quit [Quit: SIGSEGV]
rmmr has joined #yocto
vd has quit [Quit: Client closed]
vd has joined #yocto
coldspark29 has quit [Ping timeout: 240 seconds]
coldspark29 has joined #yocto
akiCA has quit [Ping timeout: 268 seconds]
ahs3 has joined #yocto
vladest has quit [Remote host closed the connection]
vladest has joined #yocto
paulg_ has quit [Ping timeout: 256 seconds]
paulg has quit [Ping timeout: 268 seconds]
camus1 has joined #yocto
camus has quit [Ping timeout: 268 seconds]
camus1 is now known as camus
akiCA has joined #yocto
jmiehe1 has joined #yocto
jmiehe has quit [Ping timeout: 268 seconds]
jmiehe1 is now known as jmiehe
codavi has joined #yocto
akiCA has quit [Ping timeout: 240 seconds]
codavi has quit [Ping timeout: 240 seconds]
jpuhlman_ has joined #yocto
jpuhlman is now known as Guest3124
Guest3124 has quit [Killed (molybdenum.libera.chat (Nickname regained by services))]
WenwuZhang has quit [Remote host closed the connection]
WenwuZhang has joined #yocto
WenwuZhang has quit [Remote host closed the connection]
zpfvo has joined #yocto
rob_w has joined #yocto
zpfvo has quit [Ping timeout: 252 seconds]
zpfvo has joined #yocto
zpfvo has quit [Ping timeout: 268 seconds]
zpfvo has joined #yocto
zpfvo has quit [Ping timeout: 252 seconds]
zpfvo has joined #yocto
zpfvo has quit [Ping timeout: 268 seconds]
zpfvo has joined #yocto
luc4 has joined #yocto
luc4 has quit [Client Quit]
sstiller has joined #yocto
rfuentess has joined #yocto
mckoan|away is now known as mckoan
<JosefHolzmayrThe>
Wade Berrier: completely depends. you could also do a push-strategy.
Flumpy33 has joined #yocto
Flumpy33 is now known as Sion
Sion has quit [Client Quit]
JeffreySimons has joined #yocto
<JosefHolzmayrThe>
the first thing to think about is, are you providing a layer and the software stack it contains for consumption? Or are you delivering a kind of build setup, e.g. a specific machine/distro setup.
camus has quit [Ping timeout: 268 seconds]
camus has joined #yocto
<coldspark29>
Is there a way you can turn off all the information about users leaving or entering the room?
coldspark29 has quit [Quit: Konversation terminated!]
coldspark29 has joined #yocto
pregn has joined #yocto
<pregn>
good morning
<pregn>
is there a simple way to add a german keyboard to a sato based image and switch to it ?
<barometz>
coldspark29: most IRC clients have some way to do that, yeah, but that's client-specific so you'll have to look there for how
JaMa has quit [Quit: reboot]
camus1 has joined #yocto
camus has quit [Read error: Connection reset by peer]
camus1 is now known as camus
zpfvo has quit [Ping timeout: 252 seconds]
zpfvo has joined #yocto
<wyre>
Saur, I'm having problems with the method you described in the wiki
leon-anavi has joined #yocto
<wyre>
despite I've hashed the password the system cannot verify it
<wyre>
but I'm able to verify the hash is right with 'openssh passwd -salt <random gen salt> -5 <my pass>'
<wyre>
that will produce the very same hash I've hardcoded, but ... when I try to log in the system is not able to authenticate me
vd33 has joined #yocto
<coldspark29[m]>
<barometz> "coldspark29: most IRC clients..." <- Ok thanks
vd has quit [Ping timeout: 256 seconds]
michalkotyla_ has quit [Ping timeout: 240 seconds]
florian has joined #yocto
rr12zer has quit [Ping timeout: 240 seconds]
michalkotyla has joined #yocto
rr12zer has joined #yocto
camus1 has joined #yocto
camus has quit [Remote host closed the connection]
camus1 is now known as camus
vd33 has quit [Ping timeout: 256 seconds]
<wyre>
definitely ... the system is not able to authenticate me with your yesterday modifications in the wiki, Saur 😞
michalkotyla has quit [Ping timeout: 256 seconds]
michalkotyla has joined #yocto
<hmw[m]>
Hi my weston instance is crashing 1minit and 30 seconds after it is loaded after boot. if i restart it manualy befor the 1:30 or after that time it keeps running. i also get: Startup finished in 917ms (kernel) + 1min 56.140s (userspace) = 1min 57.058s. after weston crashes
camus has quit [Read error: Connection reset by peer]
camus has joined #yocto
<dvorkindmitry>
how to get {layername} from BBFILE_PATTERN_{layername} in the recipe?
<dvorkindmitry>
is there are any global variable with "layername" available?
kayterina has joined #yocto
<kernelspace>
hi, need a suggestion. Trying to compiel a 4.4 kernel in dunfell, but gcc9 fails with some errors. Is it better to downgrade gcc ? Is it possible ?
<qschulz>
kernelspace: just disable the warning to errors flags that are problematic
<qschulz>
by appending to KERNEL_CC IIRC
<kernelspace>
qschulz: will try, thanks
lucaceresoli has joined #yocto
<qschulz>
kernelspace: or apparently KERNEL_EXTRA_ARGS could also be used to pass arguments to make of the kernel
camus1 has joined #yocto
camus has quit [Ping timeout: 240 seconds]
camus1 is now known as camus
<kernelspace>
qschulz: thanks
JaMa has joined #yocto
davidinux has quit [Quit: WeeChat 2.8]
davidinux has joined #yocto
michalkotyla has quit [Ping timeout: 240 seconds]
michalkotyla has joined #yocto
<user_>
Hi, We are facing issue while integrating cups in yocto. The yocto name is dunfell, cups filter version is 1.27.2 and cups version is 2.2.12. The error we are facing is ParseError at /home/user/yocto/imx-yocto-bsp/sources/poky/meta/recipes-extended/cups/cups-filters.inc:10: unparsedline: 'DEPENDS:class-native = "poppler-native glib-2.0-native dbus-native pkgconfig-native gettext-native libpng-native"'
<qschulz>
user_: which dunfell version of poky?
<qschulz>
I assume your dunfell version is too old and does not support the new override syntax
<qschulz>
so 1) update to latest poky dunfell (highly recommended)
<JosefHolzmayrThe>
2) have a drink
<qschulz>
wait.. cups-filters comes from poky... did you backport patches directly into your dunfell poky from a newer release of poky by any chance?
michalkotyla has quit [Read error: Connection reset by peer]
michalkotyla has joined #yocto
kriive has joined #yocto
<kriive>
What's a production-proven way to secure ssh access to a Yocto IoT device?
<JosefHolzmayrThe>
kriive: massively depends on your definition of production proven, and the processes you have in place
<JosefHolzmayrThe>
kriive: but the actually best way is not shipping ssh.
michalkotyla_ has joined #yocto
michalkotyla has quit [Ping timeout: 256 seconds]
michalkotyla has joined #yocto
michalkotyla_ has quit [Ping timeout: 256 seconds]
dtometzki has quit [Ping timeout: 240 seconds]
zpfvo has quit [Ping timeout: 245 seconds]
zpfvo has joined #yocto
zpfvo has quit [Ping timeout: 260 seconds]
zpfvo has joined #yocto
dkl has quit [Quit: %quit%]
dkl has joined #yocto
coldspark29__ has joined #yocto
t_unix[m] is now known as T_UNIX[m]
coldspark29__ has quit [Ping timeout: 245 seconds]
kayterina has quit [Ping timeout: 240 seconds]
troth has quit [Ping timeout: 256 seconds]
camus has quit [Ping timeout: 240 seconds]
camus has joined #yocto
kayterina has joined #yocto
coldspark29__ has joined #yocto
goliath has joined #yocto
troth has joined #yocto
coldspark29__ has quit [Ping timeout: 268 seconds]
xmn has joined #yocto
<smurray>
JosefHolzmayrThe: heh, I believe Mender have a remote access solution ;)
zpfvo has quit [Ping timeout: 252 seconds]
zpfvo has joined #yocto
<JosefHolzmayrThe>
smurray: they do, and it is quite neat - however recommending it is not really worthwhile if it doesn'T fit the rest of the ecosystem.
<JosefHolzmayrThe>
ossystems has something too, which is essentially only ssh, so thats another option.
camus has quit [Quit: camus]
zpfvo has quit [Ping timeout: 240 seconds]
zpfvo has joined #yocto
<smurray>
JosefHolzmayrThe: yes, but I was amused since they're your new employer
sakoman has joined #yocto
<kriive>
well, we're already using mender so that's super neat
<kriive>
i didn't know about remote terminal! thanks
mvlad has joined #yocto
<JosefHolzmayrThe>
kriive: heh thats what i meant by "it depends". yeah, the remote terminal is very neat, however it requires an additional license AFAIK. if you need further clarification, then get in touch with your contact person, or me in private, however you prefer.
<JosefHolzmayrThe>
kriive: (note: i work for mender)
<kriive>
As far as I can understand from your website, remote terminal is included in the open source license we're using atm for our tests
<kriive>
I am trying it out right now haha
<kriive>
If I have any question I will gladly your offer for clarifications, thanks :)
<kriive>
gladly accept*
Ad0 has joined #yocto
zpfvo has quit [Ping timeout: 260 seconds]
zpfvo has joined #yocto
Minvera has joined #yocto
zpfvo has quit [Ping timeout: 252 seconds]
zpfvo has joined #yocto
<jonmason>
I personally blame zeddii for poky-tiny not working and generally being a PITA
prabhakarlad has joined #yocto
<zeddii>
I'm ok with that!
<jonmason>
[ 0.000000] Linux version 5.14.21-yocto-tiny (oe-user@oe-host) (arm-poky-linux-musleabi-gcc (GCC) 11.2.0, GNU ld (GNU Binutils) 2.37.20210721) #1 PREEMPT Mon Nov 29 01:18:09 UTC 2021
<jonmason>
zeddii: if it is interesting to you, I can a cron job that uses lei and b4 to auto-apply patches off my mailing list to my git tree and run CI
<jonmason>
its super cool to wake up to CI that ran on a new patch while I was asleep
<zeddii>
nice. I was messing around with b4 just last night.
<zeddii>
but since I don't use lei, getting the message IDs is a PITA.
<jonmason>
you can get it from mailx and grep
camus has joined #yocto
<jonmason>
lei is better, but I had an early version that just used mail to dump everything outstanding and grep for message-id
<jonmason>
but with lei, I can use a unique mailbox that only contains mail from the past hour, and run it every hour
paulg_ has joined #yocto
paulg has joined #yocto
jpuhlman_ has joined #yocto
jpuhlman is now known as Guest2683
Guest2683 has quit [Killed (strontium.libera.chat (Nickname regained by services))]
jpuhlman_ is now known as jpuhlman
rob_w has quit [Quit: Leaving]
zpfvo has quit [Ping timeout: 252 seconds]
zpfvo has joined #yocto
zpfvo has quit [Ping timeout: 260 seconds]
zpfvo has joined #yocto
zpfvo has quit [Ping timeout: 260 seconds]
zpfvo has joined #yocto
zpfvo has quit [Ping timeout: 245 seconds]
zpfvo has joined #yocto
sakoman has quit [Ping timeout: 240 seconds]
<ziga>
One question... I tried enabling systemd and it works if I put configuration at the end of "conf/local.conf". I also tried putting the same configuration to my layers "conf/layer.conf" but it does not work... It is true that I did not put it at the end of the file, but how come? Afteral ".conf" files are global right?
<qschulz>
ziga: don't put things in conf/layer.conf except what bitbake-layers create-layer puts there
<qschulz>
to share your work with a systemd enabled image, just create a new distro, that's how you're supposed to do it
<ziga>
So I create a new layer that provides a new distro annd I do that inside the distro's conf file?
<ziga>
I guess using conf/local.conf is not the proper way even if it works right?
kiran_ has joined #yocto
<qschulz>
yes to both questions
Sunny has joined #yocto
<Sunny>
Hello All. I am trying to pass variables from do_compile_append() to a custom python function (${@custom_python_function(d)}) inside both inside an .inc recipe file. It seems the python function cannot dereference the value of the variable and is empty. Any ideas on how to pass variables from do_compile_append() into a python function inside a
<Sunny>
recipe?
<moto-timo>
don't abuse the distro conf either. Some folks put IMAGE_INSTALL there... instead create a proper image (you can just 'require' the base image you want to extend). And probably packagegroup(s).
<moto-timo>
and sometimes you want to create a "bsp" layer, even if you aren't creating a MACHINE... you can put kernel changes there and add MACHINEOVERRIDES
sakoman has joined #yocto
<Sunny>
my custom python function looks like:
<Sunny>
def custom_python_function(d):
<Sunny>
print(d.getVar('do_compile_var'))
JavierPeces[m] has quit [Quit: You have been kicked for being idle]
<qschulz>
moto-timo: systemd is a distro feature AFAIR
<moto-timo>
qschulz: yes, I forgot to mention that is a good example of what SHOULD be in distro conf
<qschulz>
basically, anything that is hw specific, in machine conf, everything that cna be put in an image should be there, and what's left, in a distro.conf. I think that's an ok summary?
<qschulz>
Sunny: are you setting a variable in one task and trying to get it from another?
<Sunny>
qschulz: I am setting the variable in one task and passing it to a custom python function just outside of the task.
<qschulz>
yeah that's not possible
<qschulz>
except if your python function is called from the task
<qschulz>
in which case, d.getVar('var') will work just fine
<Sunny>
Yes, I am calling the python function from within my do_compile_append() task i.e. ${@custom_python_function(d)} and passing d as a parameter
<qschulz>
ah yes
<qschulz>
that's not possibel IIRC
<Sunny>
But inside the python function, I do a print but don't see anything printed
<qschulz>
because ${@} will be resolved at parsing time and your variuable won't be set yet
<qschulz>
at least that's my guess
<Sunny>
Ah ok. That makes sense. No wonder i wasn't seeing anything
<Sunny>
Do you know how would be able to pass a variable from a task to a python function?
<ziga>
qschulz & moto-timo: Thank you for the philosophy explanation. I am re-watching Josef's video on the "holy trinity" i.e. distro, image, machine. There it is well explained!
<qschulz>
Sunny: I don't think postfuncs allows to call a python function otherwise that would have been my suggestion
<qschulz>
it'd be do_compile[postfuncs] += "custom_python_function" for example
JeffreySimons has quit [Quit: Konversation terminated!]
<qschulz>
I think one can call shell functions from a python script but I don't know if the opposite is possible (I mean it is, just it does not share the datastore because it's resolved/parsed too early)
<qschulz>
time to summon the veterans :)
<qschulz>
RP: rburton: ^ sharing a variable set in a task with a python function called from the same task, is that possible? if so how
vd33 has joined #yocto
<moto-timo>
why would a recipe that is RDEPENDS in a packagegroup not build and no sign of it in the image? Hmmm.
kayterina has quit [Read error: Connection reset by peer]
<moto-timo>
also, are packagegroups one of the places we still need to set PR to make sure changes trigger a rebuild?
* moto-timo
has Friday foggy brain
<qschulz>
moto-timo: RDEPENDS syntax typo?
<moto-timo>
qschulz: good point... I better check that
<moto-timo>
I already caught a INSANE_SKIP_${PM} (note the M) typo yesterday
<moto-timo>
my fingers are flying on the keyboard, but maybe not where they should be
<qschulz>
have we not migrated INSANE_SKIP to use the new override syntax?
<moto-timo>
this is dunfell... sorry brain is split
<qschulz>
moto-timo: ok then both override syntaxes are fine
<moto-timo>
with new enough bitbake, right
<Sunny>
qschulz: Ah I see.
<qschulz>
moto-timo: indeed
zpfvo has quit [Ping timeout: 240 seconds]
<qschulz>
Sunny: so, sorry but I don't have the answer :/
<qschulz>
Sunny: if you ever find out, please let us know
<qschulz>
this feels like something we should document
<sgw>
rburton: well, maybe I did not do the d.setVar() correctly, setting a variable did not help the multiple warns from anon python(), I think it sets it in the recipe scope still, not in the global scope.
<qschulz>
Sunny: yup, that's what you do from inline python functions
<qschulz>
Sunny: or to call a python function from a python task
<qschulz>
d is the datastore
<qschulz>
Sunny: also, except if you're using the very outdated 2.3 poky release, please use docs.yoctoproject.org
<qschulz>
basically, anything that is currently supported is on docs.yoctoproject.org
<qschulz>
alright folks, time for me to put the computer on the side for two weeks. Happy holidays and "see" you next year :) /me waves
angolini has joined #yocto
<vd33>
qschulz: I'm adding a custom distro which you can base your distro on, the best would be to centralize all these in my base distro then. Can you generate the documentation somehow for only your distro variables?
xmn has quit [Ping timeout: 256 seconds]
<Sunny>
qschulz: Thank you.
<Sunny>
moto-timo: Thank you.
<Sunny>
Happy holidays :)
florian has quit [Quit: Ex-Chat]
rfuentess has quit [Quit: CHELAS! MUCHAS CHELAS! CHELAS EVERYWHERE!]
mckoan is now known as mckoan|away
coldspark29__ has joined #yocto
<Tartarus>
Well, hunh.
<Tartarus>
So, can you really safely share, concurrently, SSTATE_DIR / DL_DIR between multiple builds now?
florian has joined #yocto
coldspark29__ has quit [Ping timeout: 260 seconds]
florian has quit [Ping timeout: 240 seconds]
camus has quit [Remote host closed the connection]
leon-anavi has quit [Quit: Leaving]
Guest80 has joined #yocto
<Guest80>
Hi, I have problems with a dynamic library. If I compile it in yocto directly, it's about 400kb. If I compile the same code using the SDK, it becomes 57MB. Any hints how to debug this or what the problem could be?
mariusz has joined #yocto
Guest80 has quit [Quit: Ping timeout (120 seconds)]
<Sunny>
Can I convert bitbake tasks into bitbake python functions?
<Sunny>
When I do this, I get a strange error "unable to parse recipe. Error during parse shell code"
<Tartarus>
You want "${@python ....}"
<Tartarus>
Grep around on "${@ for examples
Herrie has joined #yocto
Guest80 has joined #yocto
<tlwoerner>
Tartarus: that's what RP's been saying for years (i mostly keep them separate, probably more out of habit)
<tlwoerner>
i believe the AB is setup that way, which does a lot of building of lots of stuff
<Tartarus>
Yeah, I've long long recommended separate too, but just had a customer point to a thread they had with RP
<Tartarus>
I updated my answer to be that I'm not gonna argue against RP, but I still keep them separate out of caution as space is cheaper than engineering time
<Tartarus>
Doubly so since they had some "sstate corruption" problem
<tlwoerner>
sounds good
<Tartarus>
And I'm heartened I'm not the only one that still keeps this separate :)
Guest80 has quit [Quit: Ping timeout (120 seconds)]
<Herrie>
I'm trying to get PinePhonePro dev unit (RK3399S) to work with fitImage, DTB and initramfs, but I must be missing something. I get everything to build correctly, however I'm stuck at the BL31 in loop with the fitImage kernel. When I use separate kernel I get a bit further. My WIP tree (don't mind the mess) is at: https://github.com/webOS-ports/meta-pine64-luneos/commit/dee5b2b3a7d27955c8a310f
<Herrie>
fc9f1404096ec6db5
<Herrie>
I tried to find some more documentation, but it seems very limited
<Herrie>
I'm probably missing something basic somewhere, just not sure what
<Herrie>
I've cherry-picked and matched a few bits from meta-arm and meta-rockchip to not pull the whole layers in for now, but might end up pulling them in if easier/needed going forward
HerrieTP_ has joined #yocto
<tlwoerner>
Herrie: that build is quite a mish-mash. your kernel is external, and you're building the rest with hardknott
<tlwoerner>
Herrie: could you possibly try building the entire image by using meta-rockchip+meta-arm and adding your machine.conf in there (on master)?
<tlwoerner>
the rock pi 4b machine conf would be a good one to look at (to copy)
<HerrieTP_>
Yes I know... Kernel is external because that's a known working mainline kernel for other distros
<tlwoerner>
HerrieTP_: oh... are there lots of patches? not mainline?
<HerrieTP_>
Well it's mainline with integrated patches
<HerrieTP_>
Probably similar to what's in linux-yocto-dev for rk3399
HerrieTP has joined #yocto
HerrieTP_ has quit [Read error: Connection reset by peer]
HerrieTP has quit [Read error: Connection reset by peer]
HerrieTP_ has joined #yocto
<Herrie>
I doubt the issue is the kernel as such, but more some config or defconfig issue somewhere...
<Herrie>
I could try a build with those layers of course
<tlwoerner>
with an external kernel i don't know it's config, and there's a chance it's the config that's the issue
<tlwoerner>
[meh, the first "its" shouldn't have the apostrophe, grr]
<Herrie>
OK, but you have had a working fitImage + DTB + initramfs before with RK3399 right with the meta-rockchip?
<Herrie>
I could switch to honister if that's easier
<Herrie>
We have the layers ready for that as well so can just as well do that
<tlwoerner>
Herrie: not quite. we build up the wic image by placing the various parts (tf-a, u-boot spl, u-boot, rootfs) in their correct locations
florian has joined #yocto
<tlwoerner>
we don't use an initramfs
<Herrie>
tlwoerner: Well that's what others seems to do as well with PKGBUILD but in a lot less elegant way imho with boot.txt, boot.scr
<tlwoerner>
and although we build the fitimage (probably a default part of the u-boot build) we don't use it
<Herrie>
I could also just go with master and take it from there to be on bleeding edge
<Herrie>
tlwoerner: ^
<vd33>
FOO ?= "A" FOO += "B", ${FOO} will equal "B", right?
<vd33>
hum maybe not
kiran_ has joined #yocto
kiran_ has quit [Ping timeout: 240 seconds]
<ziga>
I think it should equal to "B"!
<ziga>
?= and += are evaluated imediatelly
<rfs613>
it will depend if FOO had a value previously...
<vd33>
so if they are evaluated immediately and FOO ?= "A" is before FOO += "B", then FOO must equal "A B"
<vd33>
rfs613 true, I'm assuming FOO had no value before the ?= is encountered.
jadudm has joined #yocto
<ziga>
rfs613: True
<jadudm>
Hi all; where is a good place to ask questions about a yocto deployment and managing it? I have a project where we've piloted some RPi-based sensors (n=~20), and we'd like to improve the infrastructure in terms of repeatability, stability, and maintainability. I've built a test Yocto image, but questions about how best to provision and maintain a
<jadudm>
second-round pilot fleet (n=~50) is not obvious to me.
kiran_ has joined #yocto
ziga has quit [Ping timeout: 240 seconds]
kiran_ has quit [Quit: Konversation terminated!]
jmiehe has joined #yocto
Minvera has quit [Quit: Leaving]
jadudm has quit [Quit: Client closed]
xmn has joined #yocto
Tokamak has joined #yocto
lucaceresoli has quit [Ping timeout: 252 seconds]
zenstoic has joined #yocto
<dvorkindmitry>
mc:myarch:glib-2.0-1_2.68.4-r0 do_package_qa: QA Issue: glib-2.0-ptest rdepends on locale-base-fr-fr, but it isn't a build dependency? [build-deps]
<dvorkindmitry>
I have a lot of warnings like this. how can I get rid of it?
Tokamak has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<RP>
Tartarus: the autobuilder stresses that *massively*. It is engineered to work
<RP>
qschulz: in theory it would be set into the datastore and the function should see the same datastore and be able to retrieve it. A lot 'depends' on what you're doing though
florian has quit [Ping timeout: 240 seconds]
Tokamak has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Tartarus>
RP: Yeah, and I can't argue on behalf of the customer that reports "sstate corruption" and doesn't have a reproducer. Still feels odd to share the actual SSTATE_DIR / DL_DIR among concurrent users but I guess I need to update my personal guidance
<Tartarus>
We are tracking down a different wtf with sharing TMPDIR between two similar MACHINES and somehow the wrong file being used out of SSTATE_DIR, but that's different, the sstate itself is correct
<perdmann_>
Hi, i just stumbled upon DRM / FB . I thought i could just use an existing tool (fbi) to print a picture to Framebuffer. Now it turns out that I can't, its all DRM now. Does anyone know a tool or a Framework which i could use? I know about libdrm but its not well documented...
<WadeBerrier[m]>
<JosefHolzmayrThe> "the first thing to think about..." <- mostly a couple of layers for consumption. What would you suggest for these two different scenarios? (Also, what did you mean by a "push strategy" ?)