Hi all
This continues the discussion on author lists, versions, DOIs etc…
Ben raises the point if the title of the DOI for the VM should include the version or not (see email below). In my opinion, a version-specific DOI should lead to a web page with the version name clearly stated. My understanding of the zenodo “generic” DOI thing is that it always points to the most recent version (i.e. it does not point to a “generic page” that points to several versions).
When I look at the links for the SIRF releases, I see that
- The v1.1.1 link https://zenodo.org/record/2707949#.XOgJkkgo-Uk has the version number in the title.
- If you click on the “newer version”, you get to 2.0.0 at present (same if you click on version 2.0.0 in the box on the right). That page doesn’t have 2.0.0 in the title. You have to check in the “versions box” to see which version it is.
I assume that this is how the “github release to zenodo” thing works. I also assume that if we release 2.1, then the DOI for v2.0.0 will lead to a page with v2.0.0 clearly in the title. I don’t think it’s ideal, but it’s fine for me.
If this is how it works for the software, we should do the same for the VM. How? I have no idea…
Author lists and versions then. I see that all SIRF versions have the same author list. I’m afraid I think that makes no sense to me. I’m assuming that this is because Casper manually edited the list according to my recommendation for v2.0.0? Sadly, there doesn’t seem to be a github way to get “contributors for a certain release”. Does this mean we have to agree on a list of authors for previous releases? It’s going to be fun.
Kris
From: Thomas, Ben <[log in to unmask]>
Sent: 24 May 2019 14:29
To: Pasca, Edoardo (STFC,RAL,SC) <[log in to unmask]>
Subject: Re: VM 2.0.0
I meant “v.2.0.0”. Shouldn’t the title be “SIRF Virtual Machine”, with each new upload getting a new version number and DOI?
Ben
From: Edoardo Pasca - UKRI STFC <[log in to unmask]>
Date: Friday, 24 May 2019 at 14:26
To: "Thomas, Ben" <[log in to unmask]>
Subject: RE: VM 2.0.0
What “v”?
;)
Thanks
Edo
From: Thomas, Ben <[log in to unmask]>
Sent: 24 May 2019 14:24
To: Pasca, Edoardo (STFC,RAL,SC) <[log in to unmask]>
Subject: Re: VM 2.0.0
Did you want to write v2.0.0 in the title?
Ben
From: CCP-PETMR Developers list <[log in to unmask]> on behalf of Edoardo Pasca - UKRI STFC <[log in to unmask]>
Reply-To: "Pasca, Edoardo (STFC,RAL,SC)" <[log in to unmask]>
Date: Friday, 24 May 2019 at 14:09
To: "[log in to unmask]" <[log in to unmask]>
Subject: Re: VM 2.0.0
https://doi.org/10.5281/zenodo.3228199
I hope it’s all right.
Edo
From: Thielemans, Kris <[log in to unmask]>
Sent: 24 May 2019 14:04
To: [log in to unmask]; Pasca, Edoardo (STFC,RAL,SC) <[log in to unmask]>
Subject: Re: VM 2.0.0
Change
providing the scientific community of a open source software
To
providing the scientific community open source software
Kris
From: Edoardo Pasca - UKRI STFC
Sent: Friday 24 May, 13:56
Subject: RE: VM 2.0.0
To: Thielemans, Kris, [log in to unmask]
If you want to proof-read this is the description I added:
The Synergistic Image Reconstruction Framework (SIRF) is a software suite aiming at providing the scientific community of a open source software for reconstruction of data from real PET/MR scanners, whilst being simple enough to be used in a teaching context.
SIRF provides user-friendly Python and MATLAB interfaces built on top of C++ libraries. SIRF uses advanced PET and MR reconstruction software packages (currently STIR, Software for Tomographic Image Reconstruction, for PET and Gadgetron for MR) and image registration tools (currently NiftyReg).
The current download is a VirtualBox appliance with SIRF and prerequisites pre-installed in a Ubuntu 18.04 machine.
Edo
PS In a rare case of recurrent application I was forced to upload to zenodo from inside the VM. Wonderful.
From: Thielemans, Kris <[log in to unmask]>
Sent: 24 May 2019 12:01
To: Pasca, Edoardo (STFC,RAL,SC) <[log in to unmask]>; [log in to unmask]
Subject: RE: VM 2.0.0
Difficult, but either GPL3 or “Other”
From: Edoardo Pasca - UKRI STFC <[log in to unmask]>
Sent: 24 May 2019 11:46
To: Thielemans, Kris <[log in to unmask]>; [log in to unmask]
Subject: RE: VM 2.0.0
What license will I attach to this zenodo upload, do you reckon?
Edo
From: Thielemans, Kris <[log in to unmask]>
Sent: 24 May 2019 11:14
To: Pasca, Edoardo (STFC,RAL,SC) <[log in to unmask]>; [log in to unmask]
Subject: RE: VM 2.0.0
Great!
A reboot (almost) always helps!
Don’t forget that you cannot export the VM where you ran everything (unless you do judicious cleaning). (So, either export first, or create a snapshot after the zerofill)
Kris
From: CCP-PETMR Developers list [mailto:[log in to unmask]] On Behalf Of Edoardo Pasca - UKRI STFC
Sent: 24 May 2019 11:04
Subject: Re: VM 2.0.0
OK,
All the notebooks work. Also the ones of which one must download the real data with the download_MR_data.sh and PET scripts.
I’ll quickly package up the VM and put it on Zenodo with the author as in the previous email
Edo
From: Brown, Richard <[log in to unmask]>
Sent: 24 May 2019 10:15
To: Pasca, Edoardo (STFC,RAL,SC) <[log in to unmask]>
Cc: <[log in to unmask]> <[log in to unmask]>
Subject: Re: VM 2.0.0
Hi Edo,
This error comes from not being able to have the file open from multiple sources.
To make sure that no other jupyter notebooks are using the file, make sure you shut their kernels down (click the "running” tab and shut them down from there).
Let me know if the problem persists.
Rich
On 24 May 2019, at 10:06, Edoardo Pasca - UKRI STFC <[log in to unmask]> wrote:
I just run the grappa_basic.py in the examples via Python directly
python grappa_basic.py
It seems a Filesystem problem of file problem.
---
reading in file /home/sirfuser/devel/buildVM/sources/SIRF/data/examples/MR/simulated_MR_2D_cartesian_Grappa2.h5...
ERROR: HDF5 Error in ../../../src/H5FDsec2.c, line 939: unable to flock file, errno = 11, error message = 'Resource temporarily unavailable'
ERROR: HDF5 Error in ../../../src/H5FD.c, line 1821: driver lock request failed
ERROR: HDF5 Error in ../../../src/H5Fint.c, line 1168: unable to lock the file or initialize file structure
ERROR: HDF5 Error in ../../../src/H5F.c, line 579: unable to open file
ERROR: File Error in /home/sirfuser/devel/buildVM/sources/ISMRMRD/libsrc/dataset.c, line 969: Failed to open file.
File: /home/sirfuser/devel/install/python/sirf/Gadgetron.py
Line: 747
check_status found the following message sent from the engine:
??? "ISMRMRD File Error in ismrmrd_open_dataset (/home/sirfuser/devel/buildVM/sources/ISMRMRD/libsrc/dataset.c:969: Failed to open file.\nISMRMRD HDF5 Error in H5Fopen (../../../src/H5F.c:579: unable to open file\nISMRMRD HDF5 Error in H5F_open (../../../src/H5Fint.c:1168: unable to lock the file or initialize file structure\nISMRMRD HDF5 Error in H5FD_lock (../../../src/H5FD.c:1821: driver lock request failed\nISMRMRD HDF5 Error in H5FD_sec2_lock (../../../src/H5FDsec2.c:939: unable to flock file, errno = 11, error message = 'Resource temporarily unavailable'" exception caught at line 490 of /home/sirfuser/devel/buildVM/sources/SIRF/src/xGadgetron/cGadgetron/cgadgetron.cpp; the reconstruction engine output may provide more information
The other thing I noticed is that it runs ok once and then I get this error.
Edo
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Edoardo Pasca - UKRI STFC
Sent: 24 May 2019 08:42
Subject: Re: VM 2.0.0
Well, I don’t think this is the problem, as it is working sometimes and sometimes not.
I’ve rebooted and I’ll see if that solves the issue
edo
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Thielemans, Kris
Sent: 23 May 2019 17:00
Subject: Re: VM 2.0.0
Good spot! Yes, that’s it of course. Note however that we don’t want to distribute the VM with the data on-board, as then it’s going to be huge…
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Brown, Richard
Sent: 23 May 2019 16:52
Subject: Re: VM 2.0.0
Could the third error from having not downloaded / correctly symlinked the extra SIRF_data?
On 23 May 2019, at 16:24, Thielemans, Kris <[log in to unmask]> wrote:
Hi
No clue. Sorry. What updating your previous one then?
Kris
PS: what’s the full log of the error in point 3. At present you give only the SIRF error, but not the STIR one. Could be that it’s redirected in one of the .txt files (depends on the script if it does it or not)
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Edoardo Pasca - UKRI STFC
Sent: 23 May 2019 16:01
Subject: Re: VM 2.0.0
Guys,
I need some help. I fixed the small things below and recreated a VM from scratch.
Build and ctest is fine.
I now have a number of problems that I didn’t have before:
Jupyter doesn’t work if accessed from the host machine. Well, I get to enter the password, but then it’s stuck on a HTTP 302 (redirect).I installed firefox and jupyter works from inside the guest machineLots of troubles with reading files from the example data error: ??? "'Error opening file 20170809_NEMA_60min_UCL.l.hdr\\n' exception caught at line 267 of /home/sirfuser/devel/buildVM/sources/SIRF/src/xSTIR/cSTIR/cstir.cpp; the reconstruction engine output may provide more information"
Edo
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Edoardo Pasca - UKRI STFC
Sent: 23 May 2019 09:52
Subject: Re: VM 2.0.0
I’m working on it.
Edo
From: Thielemans, Kris <[log in to unmask]>
Sent: 23 May 2019 08:40
To: Pasca, Edoardo (STFC,RAL,SC) <[log in to unmask]>; [log in to unmask]
Subject: RE: VM 2.0.0
Hi Edo
Sadly, the “Gadgetron notebook” PR related to https://github.com/CCPPETMR/SIRF-Exercises/issues/29 was not merged. I have done this now, after 1 more fix.
As you noted, there’s one more notebook failure. I’ve created an issue with diagnostics and suggested fix.
Edo, can you have a look at this?
Indeed, the checklist says to run the exercises before the release :-;
As it now appears we need a new VM anyway, I suggest to address
Edo, can you take care of this?
My suggested course of action:
Fix https://github.com/CCPPETMR/SIRF-Exercises/issues/31Import a new VM from your current ovaUpdate SIRF-ExercisesCreate a spyder config file by handZerofillExportUpload to zenodo as releaseContinue with the checklist
This might be the fastest way. But feel free to use whatever you prefer of course.
Kris
PS: can you send me a draft of the announcement email? Need to include that updating the 1.x VM will fail unfortunately.
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Edoardo Pasca - UKRI STFC
Sent: 22 May 2019 11:18
Subject: Re: VM 2.0.0
Please find the latest version on http://edosil.net/stfc/sirf/SIRF_2.0.0.ova
Please test and give me the go code
Edo
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Edoardo Pasca - UKRI STFC
Sent: 21 May 2019 16:25
Subject: Re: VM 2.0.0
I am currently rebuilding the VM. Tomorrow I will be uploading it for your testing.
Edo
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Thielemans, Kris
Sent: 21 May 2019 15:24
Subject: Re: VM 2.0.0
Hi
So… we should never point to personal forks. If we ever need to point to a fork, it’ll need to be on CCPPETMR.
This is
(just merged).
We have a question now if that PR is an important enough change to tag the SB with 2.0.1. I wouldn’t harm the VM, but then we have the fun if desyncing the SB and SIRF. My feeling therefore is to leave the tagging of the SB for a bit longer. Especially as presumably this is dragging in other NiftyReg changes which we haven’t tested properly either.
Kris
PS: By the way, I had suggested to NOT tag the CPPPETMR_VM either, but too much noise on the list I guess :-;
From: CCP-PETMR Developers list [mailto:[log in to unmask]] On Behalf Of Brown, Richard
Sent: 21 May 2019 15:14
Subject: Re: VM 2.0.0
Apologies, it was pointing to my fork and since my PR got accepted, I deleted the branch.
I’ve restored the branch, so could you give it another go? Hopefully that should have sorted it.
Since my NiftyReg PR has been accepted, from current master onwards the KCL fork will be used (I’ve updated version_config.cmake), but I’ll make sure not to delete the branch on my fork that was used for SIRF 2.0.0.
Rich
On 21 May 2019, at 14:49, Edoardo Pasca - UKRI STFC <[log in to unmask]> wrote:
Hi guys,
I tested the PR 137 in the VM. I added a commit. Merged to master and tagged.
Update_VM fails ...
[ 1%] Creating directories for 'NIFTYREG'
[ 2%] Performing download step (git clone) for 'NIFTYREG'
Cloning into 'NIFTYREG'...
warning: remote HEAD refers to nonexistent ref, unable to checkout.
fatal: reference is not a tree: 22f24db1113973507689aeabbc3f17ddeca1d7f6
CMake Error at /home/sirfuser/devel/buildVM/builds/NIFTYREG/tmp/NIFTYREG-gitclone.cmake:75 (message):
Failed to checkout tag: '22f24db1113973507689aeabbc3f17ddeca1d7f6'
Checking the SuperBuild tag is v2.0.0 and SIRF_TAG is v2.0.0
Apparently NIFTYREG_TAG is wrong?
Edo
-----Original Message-----
From: CCP-PETMR Developers list <[log in to unmask]> On Behalf Of Edoardo Pasca - UKRI STFC
Sent: 17 May 2019 22:02
Subject: Re: VM 2.0.0
sure.
The only thing I don't have access to is Zenodo.
Edo
________________________________
From: CCP-PETMR Developers list <[log in to unmask]> on behalf of Thielemans, Kris <[log in to unmask]>
Sent: Friday, May 17, 2019 4:47:42 PM
Subject: Re: VM 2.0.0
Hi
I think the best thing to do is now that Edo run this on his VM, re-zeroes and re-exports, and then upload later. Amazingly, zenodo is down, so we cannot upload there anyway...
So, release on Monday. Edo, can you take care of this then? (website, email with suitably enthusing message to all lists, rest of the checklist anyway). I can help if need be.
Kris
-----Original Message-----
From: CCP-PETMR Developers list <[log in to unmask]> On
Behalf Of Brown, Richard
Sent: 17 May 2019 16:41
Subject: Re: VM 2.0.0
we fixed Evgueni's error with "sudo pip uninstall pyder-kernels".
We're not overly sure what the problem was about, nor why this fixed
it...
I'll update the VM script.
Rich
On 17 May 2019, at 16:17, Evgueni Ovtchinnikov - UKRI STFC
<[log in to unmask]> wrote:
sorry it appears i replied to edoardo only...
________________________________
From: Ovtchinnikov, Evgueni (STFC,RAL,SC)
Sent: Friday, May 17, 2019 3:36 PM
To: Pasca, Edoardo (STFC,RAL,SC)
Subject: Re: VM 2.0.0
Everything seems ok save for this error message when starting Spyder
(which
seems to work fine regardless):
File
"/usr/local/lib/python2.7/dist-packages/spyder/otherplugins.py",
line 79,
in _import_plugin
module = _import_module_from_path(module_name, plugin_path) File
"/usr/local/lib/python2.7/dist-packages/spyder/otherplugins.py",
line 101, in
_import_module_from_path
info = imp.find_module(module_name, [plugin_path])
ImportError: No module named spyder_kernels-0.4.2.dist-info
By the way, when I open a file in Spyder, 'Open file' window looks
different
from previous VM (which I got from your OneDrive link) - have you
installed different version of Spyder lately?
________________________________
From: CCP-PETMR Developers list <[log in to unmask]> on
behalf of Edoardo Pasca - UKRI STFC <[log in to unmask]>
Sent: Friday, May 17, 2019 12:20:29 PM
Subject: VM 2.0.0
Hi there,
If someone wants to try and comment on the VM with SIRF 2.0.0 it can
be downloaded from
si
l.net%2Fstfc%2Fsirf%2FSIRF_2.0.0.ova&data=02%7C01%7Ck.thielemans%4
0UCL.AC.UK%7C46b37bcde41c4d71a35c08d6dade2f7d%7C1faf88fea9984c5b93
c921
0a11d9a5c2%7C0%7C0%7C636937045151742495&sdata=pze2eLd6%2BX
ML4qqjN8
oosDCC0MYS8T5GDtmlDFYRdN0%3D&reserved=0
Edo
--
Edoardo Pasca, PhD
Senior Visualisation Scientist
Scientific Computing Department, STFC Rutherford Appleton Laboratory
Fermi Avenue Chilton Didcot OX11 0QX
0044 (0)1235 44 5660, R89 F16
:wq
________________________________
To unsubscribe from the CCP-PETMR-DEVEL list, click the following link:
jiscmail.ac.uk%2Fcgi-bin%2Fwebadmin%3FSUBED1%3DCCP-PETMR-
DEVEL%26A%3D1
&data=02%7C01%7Ck.thielemans%40UCL.AC.UK%7C46b37bcde41c4d71a3
5c08d
6dade2f7d%7C1faf88fea9984c5b93c9210a11d9a5c2%7C0%7C0%7C636937045
151742
495&sdata=Jd%2BLGhPg5oUkXcpCDJVS3Onc1h4ptS6ClaX62Lpp6jQ%3D&a
mp;res
erved=0
################################################################
######
##
To unsubscribe from the CCP-PETMR-DEVEL list, click the following link:
jiscmail.ac.uk%2Fcgi-bin%2Fwebadmin%3FSUBED1%3DCCP-PETMR-
DEVEL%26A%3D1
<p class="
To unsubscribe from the CCP-PETMR-DEVEL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP-PETMR-DEVEL&A=1