You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Farmyard manure developments by Theresa Boas seem to also incorporate winter wheat and fruit tree changes, all of which are huge features. Based from our past discussion I'm strongly against merging such a monster PR. This time I'm willing to compromise as it takes a huge effort to disentangle farmyard manure+winterwheat+fruit tree fixes; more importantly these 3 features have always been tested together by @tboas, @AGonzalezNicolas, and @jjokella, so it can argued that a passing BGC test case is sufficient to merge all these developments. This upcoming PR should be treated as a special case, and I hope this is the first and last PR where we have to do this difficult merge. I propose the following acceptance criteria:
Differences should be as minimal as possible to reduce review burden. Whitespace changes like this should not show up in the PR.
A new single-point test in BGC mode that makes use of the new features must be added to the GitHub CI.
Backwards compatibility breakages are tolerated, as long as they are minimal and documented as GitHub issues.
Test results showcasing the new BGC features should be documented in the PR, just like how it was done for the new snow thermal conductivity feature.
At least both I and @s-poll should review and approve the PR. Having Stefan's review is crucial since such a big change could potentially affect his TSMP2 runs.
I merged the eCLM into this development and removed some whitespace changes, but for now only for testing. For example, the original branch does not contain PDAF-related code, but this merged version compiles for eCLM-PDAF. I yet have to go deeper into comparing results, though.
I deliberately did not open a PR yet, because I wanted to discuss it first.
However, I still think opening this issue can be a good step forward, so thanks for the initiative @kvrigor .
Also, with the advent of AI text inflation tools these days, a short and to-the-point PR description would ease the review burden. For me it's distracting to review PRs that read like a mini-technical report; paradoxically such text is more difficult to understand than code, and I'd rather spend my time understanding the prime source of truth which is code 😉. When writing PR descriptions, I usually:
provide a short high-level summary of the changes,
add references to relevant external docs/repos/issues or published journals (if any), and
highlight significant gotchas (e.g. new namelist switches)
Most of the value in a well-documented PR is in the discussion itself, which usually happens around the context of a code change. No need to spend effort in unnecessarily inflating text 😜
Activity
jjokella commentedon Sep 30, 2026
A few additions for clarity.
I have generated this branch https://github.com/HPSCTerrSys/eCLM/tree/eCLM_FYM_fixes_merged_with_master that you label Farmyard manure developments.
It is coming from original work by @tboas towards a PR (and backward compatability) here: https://github.com/tboas/eCLM/tree/eCLM_FYM_fixes
I merged the eCLM into this development and removed some whitespace changes, but for now only for testing. For example, the original branch does not contain PDAF-related code, but this merged version compiles for eCLM-PDAF. I yet have to go deeper into comparing results, though.
I deliberately did not open a PR yet, because I wanted to discuss it first.
However, I still think opening this issue can be a good step forward, so thanks for the initiative @kvrigor .
kvrigor commentedon Sep 30, 2026
Also, with the advent of AI text inflation tools these days, a short and to-the-point PR description would ease the review burden. For me it's distracting to review PRs that read like a mini-technical report; paradoxically such text is more difficult to understand than code, and I'd rather spend my time understanding the prime source of truth which is code 😉. When writing PR descriptions, I usually:
Most of the value in a well-documented PR is in the discussion itself, which usually happens around the context of a code change. No need to spend effort in unnecessarily inflating text 😜