add examples to intro episode

[carpentries-incubator]

add examples to intro episode

[jstaf]

I think either the instructor notes (https://hpc-carpentry.github.io/hpc-intro/guide/) or as callout boxes in "why use a cluster" (https://hpc-carpentry.github.io/hpc-intro/00-hpc-intro/) are the places to put this depending on whether we want instructors or students to see this. We already have a short version in the "why use a cluster" session, so maybe there?

[psteinb]

At CarpentryCon, people came up with these 3 profiles:

X is a molecular biologist that soon will receive a large DNA sequence dataset. 
Her adviser has told her she will need to use the university's HPC cluster to analyze them. 
A colleague that is supposed to teach her how has left her a set of written instructions, but is on parental leave and thus cannot help her at the moment.

An email from the sequencing centre arrives with a set of download links for X to obtain the datasets. 
X's adviser asks her to quickly investigate the quality of the data using an appropriate software available on the cluster, 
so decision can be made to pay for the cost of data generation.

The workshop will teach X how to transfer the data to an appropriate place on the cluster and access and schedule and run jobs.
 Afterwards, X will understand enough of the written instructions to be able to start the correct analyses for her data.
(By Lex Nederbragt)

------

Y is an environmental biologist that uses DNA signatures obtained from soils to study species diversity in the environment. 
She need to compare DNA sequences to large databases. So far, she has been able to use web-based tools for her limited datasets.

Recently, Y has started working with much larger datasets, and discovered that the online tool he uses has a limit of 50 entries on the online server. 
He has heard it should be possible to run the same tool through the commandline, and managed to install it on his local Laptop. 
Now, however, it takes several days before each of the analyses are finished.

The workshop will teach Y to move his data to and from the university's computer cluster, and submit jobs using pre-installed software on the cluster. 
Afterwards, Y will be able to analyze her own data and pre-installed command-line based versions of the tool
to spread the analysis over several dozen cores so it finishes in a few hours.
(By Lex Nederbragt)
 
-------

Z is a skilled bioinformatician running analysis pipelines on the computer cluster. 
A new tool has been published and Z has decided it she wants to see whether it could be replace an existing tool. 
She submits a request to the HPC IT support staff to have the new tool installed in the cluster's environment module system. 
The response is that support is swamped with requests and for now has to prioritize issues with already installed tools. 
Support suggests that Z installs the tool herself.

Z has never installed software before and quickly realizes the instructions given are way above her head. 
Meanwhile, more and more researchers report on Twitter how good the new tool is working relative to the one she uses now.

This workshop will enable Z to make a local installation of the new tool and test it out in her pipeline. 
(By Lex Nederbragt)

--------

New PhD student is given a task to select parameters for their simulation.  They need to run a set of calculations on several thousand
  combinations of parameters.  One calculation takes several minutes.  They set up the problem on their laptop but quickly realise 
  that it would take more than one month to complete the task. 
  They are told to use local HPC but they are not sure how this would help them.

Should I make them individual issues so we can discuss them and bring them into the material?

[psteinb]

My feelings are that learners should be exposed to this without larger obstacles. My hope would be that this helps them identify themselves with the main character, i.e. the material, better.

[reid-a]

The difference between what's currently (Jan. 2021) in the lesson and what's suggested here is mainly the stress on the fact that the users are HPC cluster novices. This helps learners "see themselves", but isn't really the answer to "Why use a cluster?", it's closer "What will someone like me get out of this lesson?". Retitle the lesson? Or the subsection with the profiles?

[tkphd]

As @reid-a points out, the current lesson includes three user profiles based on @psteinb's proposed blurbs. If more refinement would be useful, a more specific issue should be filed.