update readme build/install instructions

[gidden]

I've done this exercise for cyclopts, cyclus, and cycamore. This readme is a bit more involved, so I'll let someone else take a crack at it.

[gidden]

@scopatz, we should probably tackle this too (making sure the readme is up to date)

[gidden]

also, cycstub is currently still reported as forked form katyhuff/cycstub (which it was, but its the only repo in the cyclus group that is forked)

[scopatz]

Yes, @katyhuff would have to delete her cycstub repo or we would have to delete the cyclus org one to 'fix' this. I am happy either way or letting it be. I'll let @katyhuff decide.

[scopatz]

Also I think the readme here should just be replaced with the hello world example once I finish it. In this case I am of the opinion that no docs are better that old doc. So if it comes down to it, we'll kill this file.

[katyhuff]

Just a note, related to this issue, for when it gets tackled: Usage instructions for the rename script may need to be more specific.

khuff@hypatia:~/repos/cyclusv1/cycstub(develop) $ python rename.py
usage: rename.py [-h] name
rename.py: error: too few arguments
khuff@hypatia:~/repos/cyclusv1/cycstub/src(develop) $ ls
CMakeLists.txt           stub_facility.h          stub_inst.cc             stub_inst_tests.cc       stub_region.h            stub_unit_test_driver.cc
stub_facility.cc         stub_facility_tests.cc   stub_inst.h              stub_region.cc           stub_region_tests.cc
khuff@hypatia:~/repos/cyclusv1/cycstub/src(develop) $ cd ..
khuff@hypatia:~/repos/cyclusv1/cycstub(develop) $ python rename.py fco_transition_region
Traceback (most recent call last):
  File "rename.py", line 34, in <module>
    main()
  File "rename.py", line 30, in main
    with open(f.replace('stub', low), 'w') as out:
IOError: [Errno 2] No such file or directory: '/Users/khuff/repos/cyclusv1/cycfco_transition_region/CMakeLists.txt'

Re: forking, I tend to prefer not erasing history. I'd prefer to let it be as-is unless it's causing some actual difficulty. I believe it's functionally identical either way.

[scopatz]

I'll copy over the hello world example now.

I am not sure what state you got cycstub in, but it certainly seems wrong! From a clean clone, everything seems to work:

scopatz@ares ~ $ git clone cycstub/
fatal: destination path 'cycstub' already exists and is not an empty directory.
scopatz@ares ~ $ git clone https://github.com/cyclus/cycstub.git fco
Cloning into 'fco'...
remote: Counting objects: 1648, done.
remote: Compressing objects: 100% (795/795), done.
remote: Total 1648 (delta 806), reused 1648 (delta 806)
Receiving objects: 100% (1648/1648), 1.43 MiB | 560.00 KiB/s, done.
Resolving deltas: 100% (806/806), done.
Checking connectivity... done.
scopatz@ares ~ $ cd fco/
scopatz@ares ~/fco develop $ ./rename.py fco_transition_region
scopatz@ares ~/fco develop $ ls
AUTHORS.rst  CMakeLists.txt  README.rst  cmake  input  install.py  license.rst  rename.py  src
scopatz@ares ~/fco develop $ ls src
CMakeLists.txt                     fco_transition_region_facility_tests.cc  fco_transition_region_inst_tests.cc  fco_transition_region_region_tests.cc
fco_transition_region_facility.cc  fco_transition_region_inst.cc            fco_transition_region_region.cc      fco_transition_region_unit_test_driver.cc
fco_transition_region_facility.h   fco_transition_region_inst.h             fco_transition_region_region.h

Re: forking, I tend to prefer not erasing history. I'd prefer to let it be as-is unless it's causing some actual difficulty. I believe it's functionally identical either way.

Well, this is more about ownership and less about (git) history since no history would be destroyed. Currently, it seems weird that the root owner for cyclus and cycamore is the cyclus organization, but for cycstub it is in your userspace. I don't think that this is a big deal but would be nice to change.

[katyhuff]

I have the right cycstub. I just called it with "python rename.py" rather than "./rename.py".

[katyhuff]

Actually, nope, still fails.
In the initial fail, I was using revision fadd00b, which was a clean clone 2 hours ago.

[katyhuff]

and, it still fails with 037cee9:

khuff@hypatia:~/repos/cyclusv1/fco_comparision/src(develop) $ git pull cyclus develop
remote: Counting objects: 5, done.
remote: Compressing objects: 100% (5/5), done.
remote: Total 5 (delta 1), reused 0 (delta 0)
Unpacking objects: 100% (5/5), done.
From github.com:cyclus/cycstub
 * branch            develop    -> FETCH_HEAD
   fadd00b..037cee9  develop    -> cyclus/develop
Updating fadd00b..037cee9
Fast-forward
 README.rst | 455 ++++++++++++++++++++++++++++++++++++------------------------------------------------------------------------------------------------------------------------------
 1 file changed, 99 insertions(+), 356 deletions(-)
khuff@hypatia:~/repos/cyclusv1/fco_comparision/src(develop) $ ./rename.py fco_transition_region
-bash: ./rename.py: No such file or directory
khuff@hypatia:~/repos/cyclusv1/fco_comparision/src(develop) $ cd ..
khuff@hypatia:~/repos/cyclusv1/fco_comparision(develop) $ ./rename.py fco_transition_region
Traceback (most recent call last):
  File "./rename.py", line 34, in <module>
    main()
  File "./rename.py", line 23, in main
    with open(f, 'r') as inp:
IOError: [Errno 2] No such file or directory: '/Users/khuff/repos/cyclusv1/fco_comparision/CMakeLists.txt'

[rwcarlsen]

It looks like our cycstub 1.0.0 tag is not on the most recent cycstub
commit.

On Sat, May 31, 2014 at 4:53 PM, Katy Huff [email protected] wrote:

Actually, nope, still fails.
In the initial fail, I was using revision fadd00b
fadd00b, which was a clean
clone 2 hours ago.

[image: image]
https://cloud.githubusercontent.com/assets/393899/3140207/923c0cee-e90d-11e3-835f-d980db4e7723.png

—
Reply to this email directly or view it on GitHub
#6 (comment).

[rwcarlsen]

What hash should 1.0.0 be on? - I can update it.

[katyhuff]

Ah, so, the first failed run of rename.py deleted my CMakeLists.txt. So, there you go. Better now.

[scopatz]

Ah, so, the first failed run of rename.py deleted my CMakeLists.txt. So, there you go. Better now.

Great!

What hash should 1.0.0 be on? - I can update it.

The current develop head is probably best if for no other reason than the readme is updated. Thanks @rwcarlsen!