Hi!
I will be doing extensive hardware maintenance + upgrades today in a
data center, which means that gnumonks.org and some other related
systems will be offline for some amount of time today, particularly
during the morning and early afternoon (MESZ time zone).
The only Osmocom system/service affected is people.osmocom.org.
It's just likely that you will receive notifications about e-mail delays
to my personal e-mail address, but those will be automatically
re-transmitted after the systems are back online.
Regards,
Harald
--
- Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)
Hey guys,
I'm a bit disappointed that I see more merges to osmo-bsc master that tweak
MGCP and codec preferences, moving things around files and so on. Please please
do not merge this kind of stuff now, I really cannot use any of that while I'm
turning insane to finally get the osmo-bsc refactoring out of the door. It's
really mean to load merge conflicts on top of that.
If fixes really need quick merging, please do everything in your might to AVOID
moving functions around and changing file names. That's ... yeah, you get the
picture.
What is also more than welcome is to work on the branch tip of
neels/inter_bsc_ho and apply cosmetic moves to that. Just please not between
current master and the refactoring.
(And to test that branch would also be welcome, if time permits.)
Another alternative is to *talk* to me.
Many thanks.
~N
Hi Philipp,
just now I'm trying to use pySim to read a SIM card, but I notice:
ImportError: No module named pytlv.TLV
I can find no pytlv in the debian packaging; and I notice this dependency was
added only recently. When going back to
859e0fd1ce08699930064c12fb9f7908ef972b50 I can use pySim fine without that.
a) I think the readme should hint at dependencies that pySim requires, which
should include pytlv and where to find it.
b) In python, it is easy to import pytlv only in case it is actually used,
which would simplify dependencies for users that don't need to "get file/record
length from FCP (USIM)" -- would that make sense?
What are your thoughts?
~N
TLDR: Tried to find another way of managing changes to a large patch on gerrit
by submitting a merge-commit, but turns out not so useful. Just sharing findings.
When I fix and change the current large osmo-bsc refactoring patch, on the one
hand I'd like to squash new changes into it to commit the correct result right
away. On the other hand, I want to keep the new changes in separate reviewable
bits, so that reviewers don't need to read the entire patch again, and so that
the new changes can be reviewed one by one, separately.
So far my strategy was to add a new patch set for each small change. Now I
thought it might make sense to have a non-fast-forward merge combining commits
and review that. I tried it in the sandbox repos.
Turns out that gerrit will have one patch for the merge-commit including the
smaller commits, showing all combined changes. However, there will also be
separate patches in the list for each single part of that merged commit. So
there will be a series of small commits onto current master that need to be
reviewed and merged one-by-one. And in the end that is followed by one
basically useless merge-commit; meaning, if we got as far as accepting the
smaller parts of that merge, we don't need to also merge that empty
merge-commit, master already has gotten the smaller bits as separate commits.
So the conclusion is that indeed pushing new small patch-sets onto the same
large commit makes the most sense, and where changes stand on their own,
separate follow-up patches make sense.
~N
Hi Philipp,
I notice that 5 of the LCLS tests in ttcn3_bsc_tests fail saying
"unexpected number of MGW-MDCX transactions"
pass->FAIL BSC_Tests_LCLS.TC_lcls_gcr_bway_connect
pass->FAIL BSC_Tests_LCLS.TC_lcls_gcr_bway_connect_hr
pass->FAIL BSC_Tests_LCLS.TC_lcls_gcr_bway_codec_mismatch
...
pass->FAIL BSC_Tests_LCLS.TC_lcls_connect_break
pass->FAIL BSC_Tests_LCLS.TC_lcls_connect_clear
It appears that your osmo-ttcn3-hacks change
"BSC_Tests: count MGCP operations in as_media"
lacks the proper MGCP operation counts for the LCLS tests?
(The tests have been broken for four days)
https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-bsc-test/test_resu…
Could you fix that plz?
Thanks!
~N
I have build following components from source.
1. SoapySDR v0.6.12. Lime Suite v18.06.03. UHD v3.12.0.0-rc14. SoapyUHD v0.3.45. osmo-trx v0.4.0 configure --with-lms6. osmo-nitb7. osmo-bts-trx
however i have only following options are available with osmo-trx
root@miniBTS:/home/pi/osmocom/osmo-trx# osmo-trx-lms -h
Options: -h This text -C Filename The config file to use -V Print the version of OsmoTRX
I have run each instance in separate windows but I cannot show network, log file of following screens along with config files attached. would you please help me whats going wrong, I am newbie
terminal 1 : osmo-trx-lms -C osmo-trx-limesdr.configterminal 2: osmo-nitb -c openbsc.configterminal 3: osmo-bts-trx -c osmobts.config
attached config files have grabbed from https://github.com/myriadrf/cellular-network-configs/tree/master/osmocom
Regards
Babar Ali
On Saturday, 14 July 2018, 4:31:43 AM GMT+5, Pau Espin Pedrol <pespin(a)sysmocom.de> wrote:
Hi Babar,
If you want to use a LimeSDR-mini, you should build latest master, and
use osmo-trx-lms binary. To generate that binary, you first need to
build and install latest LimeSuite master [1], and then build osmo-trx
with --with-lms configure flag. As you are building for raspberry pi3,
you may want to enable instruction set optimizations as explained in the
user manual section "12 OsmoTRX hardware architecture support" [2]. Have
a look at the wiki page too for further information [3].
Also a reminder: we have osmo-trx-lms already available as a debian
package for debian9 [4], but I think we don't enable NEON on that build.
[1] https://github.com/myriadrf/LimeSuite
[2] http://ftp.osmocom.org/docs/latest/osmotrx-usermanual.pdf
[3] https://osmocom.org/projects/osmotrx/wiki/OsmoTRX
[4]
https://download.opensuse.org/repositories/network:/osmocom:/nightly/Debian…
On 12/07/18 01:36, Babar Ali wrote:
> Hi Harald
>
> Thank very much for your support please, i’ll test & share the results.
>
> Further kindly let me clear one more thing please, can I build osmotrx
> on RaspbianOS, I have already tried to build osmotrx v0.3 on RasbianOS
> it was build successfully but when I run osmo-trx —help, it shows only
> one option that is -C which I can use with osmo-trx, no other options
> are available, is it cannot be build completely on raspberry pi 3? or
> what’s gone wrong? any suggestions/comments in this regard please.
>
> Thank you for your support & time please.
>
> Regards
> Babar
>
> On Thursday, July 12, 2018, 1:26 AM, Harald Welte <laforge(a)gnumonks.org>
> wrote:
>
> On Wed, Jul 11, 2018 at 04:53:54PM +0000, Babar Ali wrote:
>
> > Anyone has an experience of Setuping 2G Network using NITB &
> LimeSDR Mini, which version of osmotrx is working fine with LimeSDR
> Mini.
>
>
> We have seen very good results using the new osmo-trx-lms with
> LimeSDR-mini.
>
> --
> - Harald Welte <laforge(a)gnumonks.org <mailto:laforge@gnumonks.org>>
> http://laforge.gnumonks.org/
> ============================================================================
> "Privacy in residential applications is a desirable marketing option."
> (ETSI EN 300
> 175-7 <tel:300%20175-7> Ch. A6)
>
--
- Pau Espin Pedrol <pespin(a)sysmocom.de> http://www.sysmocom.de/
=======================================================================
* sysmocom - systems for mobile communications GmbH
* Alt-Moabit 93
* 10559 Berlin, Germany
* Sitz / Registered office: Berlin, HRB 134158 B
* Geschaeftsfuehrer / Managing Director: Harald Welte
Anyone has an experience of Setuping 2G Network using NITB & LimeSDR Mini, which version of osmotrx is working fine with LimeSDR Mini.
Thank You
Regards
Babar
Hello,
I am trying some setup to have multiple channels working all together: I
have an UmTRX and a LimeSDR boards working fine.
The problem is that I cannot launch osmo-trx-uhd and osmo-trx-lms at the
same time as the VTY TCP port are defined in libosmocore.
Is it a way of disabling the VTY or change the TCP port it listen to?
Have a nice day,
Antony
Hi there,
I'm really happy to make my LimeSDR working with the master branch, good
job guys!
I'm now trying to enable the multi-arfcn multiplexed over one physical
channel but I do not succeed to have it working (No response from
transceiver for phy0.1 (CMD RXTUNE 894000)).
Do I need to configure or pass some parameters (see further for what I'm
doing)?
I am starting osmo-bts-trx with "-t 2" option, and here is the extract of
part of my configuration file:
----- osmo-trx-lms.cfg
trx
multi-arfcn enable
swap-channels disable
egprs enable
tx-sps 4
rx-sps 4
chan 0
tx-path BAND1
rx-path LNAW
----- osmo-bts-trx.cfg
phy 0
instance 0
osmotrx rx-gain 10
phy 0
instance 1
osmotrx rx-gain 10
bts 0
band 900
trx 0
phy 0 instance 0
trx 1
phy 0 instance 1
----- osmo-bsc.cfg
bts 0
type sysmobts
band GSM900
trx 0
arfcn 10
trx 1
arfcn 20
Thanks in advance,
Antony
Hi!
We've not used the old build1.osmocom.org machine since late April, so I
am finally pulling the plug and cancelling the contract as of tomorrow.
I manually looked around and the only parts that looked worth backing
up was ~osmocom-build/neels which I've now copied over to
build2.osmocom.org:~root/neels/from_build1/
Regards,
Harald
--
- Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)
