I have looked over the download instructions at:
and, for example,
Together these are a very well done and complete reference for people who already know more or less how it all works.
I think that the pages would benefit by having a cheat sheet that is targeted at less experienced users. An example of such a cheat sheet is given below. I am not sure where the cheat should live in the wiki organization. There is a danger that this may create multiple points of maintenance - so it would be good to learn how to minimize that.
- Talk with your experiment's Offline coordinator to learn the qualified version of art that you need.
- An example might be: v1_12_02 -qs5:e6:prof
- You can learn about the meanings of the qualifier mnemonic names at https://cdcvs.fnal.gov/redmine/projects/cet-is-public/wiki/AboutQualifiers
- Learn the architecture of the machine on which you plan to run art.
- The mnemonic names for these architectures are things like slf5, slf6, d12 or d13.
- To learn what these names mean see (fixme: url to the right place).
- Decide on the absolute path to the directory that will hold the downloaded ups products
- Make sure that there is at least xx GB of free space to hold the installed products
- cd to a scratch directory to hold some temporary files that are created by the process
- Make sure that there is at least xx GB of free space to hold these files
- Go to: https://scisoft.fnal.gov
- If you see your experiment's name, click on that link.
- If you are not sure which to click, consult someone on your experiment
- If you just want to test drive art, or if you want to run the art workbook, choose artdev.
- On the next page, click on the link for the version number of art that you learned in step 1.
- On the next page you will see a directory listing showing an html file and some other files. Click on the link to view the html file.
- Read the instructions in the section "Download Prebuilt Binaries", but don't follow the instructions yet.
- Consider the example of the qualified version of art: v1_12_02 -qs5:e6:prof
- The instructions tell you to download and run a script named pullProducts
- Hopefully it's clear where the version number ( v1_12_02) goes in the argument list.
- The qualifier gets split into two arguments:
- The prof part of s5:e6:prof is the final argument
- The s5:e6 needs to be written as s5-e6 ; note that the colon was changed to a dash.
- If the ups qualifiers that you are given contain plus signs, ignore them; they are not needed in the arguments to pullProducts.
- Follow the instructions to run pullProducts
- If you want to install additional qualified versions of art, simply rerun the pullProducts command giving the appropriately revised arguments.
- That is, if a particular fully qualified product was downloaded by the first run of pullProducts and is requested by the second run, pullProducts is smart enough to know that it does not need to redo the work on the second run.
- Be sure that there is enough disk space to hold both the scratch files and the installed files
- Repeat until you have all of the fully qualified versions that you need
A future improvement to pull products might be to accept a qualifier list in native ups format so that the script, not the human, can do the parsing. Ideally it should accept anything between s5:e6:prof and -q+s5:+e6:+prof
I think that it would also be a good idea on https://scisoft.fnal.gov to list the names of the experiments that use the nu distribution and the larsoft distribution. Many beginners will not know this.
#4 Updated by Lynn Garren about 4 years ago
- Status changed from Assigned to Resolved
- % Done changed from 0 to 100