Research Computing is a tough field that combines the challenges of many disciplines.
Like IT or commercial software development, we need to deliver real usable tools to support our users; but like research, our projects are often extremely open-ended, with complexity coming not from unvalidated requirements but the uncertaintainty of the new, requiring experimentation and discovery. Like academia, we often work with team members who are trainees, not employees; like nonprofits we are called on to enact real changes with ongoing programmes or products while funded only by budgets dependent on multiple short-term grants.
But while there are many websites and podcasts, newsletters and tutorials, on the bytes and flops and Mbit/s of research computing, there is very little out there on the genuinely hard day-to-day work of designing, building and managing R&D computing teams, projects, and software. Commercial or open-source software development, research, nonprofit, IT, business - our field is too different for advice from those fields to routinely directly carry over, even if there are lessons we can learn from them.
This is a newsletter that started in mid-Feb 2020, covering “Everything about research computing except the computers”. You can view previous posts here.
The intended audience
- Trained in or experienced with research
- Technically capable in R&D computing
- Responsible, or intending to be responsible, for managing or designing R&D computing teams or efforts
- Looking for information on and a community around managing R&D computing teams and efforts
- Weekly link roundups with occasional longer posts
- Occasional (no more than weekly) short emails with links
- Reply to any message (it’ll go directly to me and only to me) with thoughts, comments, and feedback
Feedback from subscribers will greatly shape the topics we cover. Likely early topics include:
- Hiring from outside research, writing job ads, onboarding, the responsibility ladder
- One-on-Ones, giving feedback, coaching and training, distributed teams
- “Managing upwards”
- Building and sustaining support for projects, shepharding multi-institution collaborations, communications
- Career paths, for our team members and ourselves
- R&D computing and academic credit
- Automation and reproducibility
- R&D Software planning and development throughout the lifecycle
- Grants and Sustainability
- Working with reserach communities.
Issues that would not be on topic include GPU vs FPGA, POSIX vs Object stores, Python vs C++, etc; on the other hand, a story about why and how an R&D computing team planned and succeeded at a migration from GPU/POSIX/Python to FPGA/Object Store/C++ might well be on topic.
(Jonathan’s twitter rants about things like MPI and exascale will not be on topic for this newsletter.)
Q: So you are an expert on this stuff, right?
A: Absolutely not. I do have a lot of experience in this area from a lot of different perspectives, but more importantly I continue trying to learn and improve, and anyone interested is welcome to join as I write my way through that process..
Q: Will the posts be archived anywhere?
A: That wasn’t originally the intention, but people seem to prefer it; the archives can currently be found here.
Q: Will my email address get sold/used for anything else?
A: No. I’ll use it only for sending you my thoughts on R&D computing teams, and when you no longer care to continue hearing those thoughts, unsubscribe at any time, and I’ll remove the unsubscribed member’s information routinely. No one other than me and whatever tool I’m using at the time to send out the newsletter will ever have access to your email address.
Q: How much tracking is being done in the emails?
A: None; I’ve turned off tracking and analytics for the emails.
Below are resources on intended topics that I’ve found to be useful in the past. I’ll keep this list updated with sources that come up in the link roundups.
- General managment topics
- Software development managment
- Nonprofit management
- Data-science teams