good problems: the PR backlog !

26 views
Skip to first unread message

Simon Michael

unread,
May 16, 2021, 12:28:46 AMMay 16
to Stephen Morgan, hledger
Hi Stephen.. thanks for your recent productivity. The PRs have been piling up in a way we have not seen before in the project. I have been having trouble keeping up - a combination of code review burnout, client work, a move etc..

I'd like to get more efficient / find other ways to increase our throughput, without of course jeopardising the project. My usual fear is that we could easily merge a couple of PRs with unforeseen bad consequences that we discover only slowly and after some time has passed, which would be quite costly. I could be wrong but I always I feel hledger is more vulnerable to this than many projects, in spite of all our tests.

Ie, worse-is-better and move-fast-and-break things doesn't seem the best approach for us.

Actually, it probably would win in the long run, and I *would* like to find ways to do that more. But ideally without causing regressions for users and unaffordable fixup/busywork for myself and all of us contributors.

Ideas on any of this welcome, all.



Simon Michael

unread,
May 16, 2021, 12:30:45 AMMay 16
to hledger, Stephen Morgan
PS, specifically what I wanted to ask you is could you help me prioritise your PRs for review - low hanging fruit, or important/old ones best completed first. I seem to have about of your 20 PRs/threads flagged to follow up on.
> --
> You received this message because you are subscribed to the Google Groups "hledger" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to hledger+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/hledger/A9DC3EFA-F565-4F91-9F56-1D02279B205A%40joyful.com.

morgan.s...@gmail.com

unread,
May 16, 2021, 12:47:21 AMMay 16
to hledger
I don't have any particular insights on the general problem, but I can definitely offer some help on prioritising my PRs. I've categorised them as follows: each category is in roughly increasing complexity both top to bottom and left to right.

Small and simple PRs: #1555, #1561
Simple to understand, but with a few dozen lines of code changed: #1551, #1543, #1537, #1542, #1544, #1560(!)
Simple to understand, but with a some feedback needed before merging: #1545(!)
Major changes to how hledger works, requires major discussion and testing: #1554
Simple changes, but depends on dropping support for GHC<8.6: #1557

I've flagged #1560 because it significantly simplifies the valuation/costing code, which would be helpful for both #1554 and @chvp's #1432.
I've flagged #1545 because it causes a significant change in the naming of ReportSpec's fields which is likely to conflict with other PRs, and also because the introduction of lenses makes it much easier to write new code: writing a few of these PRs I've definitely wished that this one were merged already.

Best,
Stephen

Simon Michael

unread,
May 16, 2021, 1:48:55 AMMay 16
to hle...@googlegroups.com
Thanks!

On May 15, 2021, at 18:47, morgan.s...@gmail.com <morgan.s...@gmail.com> wrote:



morgan.s...@gmail.com

unread,
Jun 7, 2021, 8:50:04 PMJun 7
to hledger
Just a quick update on the PRs, since things have changed a bit since I last responded.

Trivial changes: #1557
Reviewed and ready to merge: #1544
Simple to understand, but with possible implications throughout the codebase: #1560(!), #1564
Simple to understand, but with a some feedback needed before merging: #1545(!)
Need to decide whether to move ahead with this PR, or instead implement proposals in #1559: #1542
Major changes to how hledger works, requires discussion and testing: #1554

I've flagged #1560 because it significantly simplifies the valuation/costing code, which would be helpful for both #1554 and @chvp's #1432.
I've flagged #1545 because it causes a significant change in the naming of ReportSpec's fields which is likely to conflict with other PRs, and also because the introduction of lenses makes it much easier to write new code: writing a few of these PRs I've definitely wished that this one were merged already.

Simon Michael

unread,
Jun 7, 2021, 11:25:58 PMJun 7
to hledger
Thanks Stephen! Noted.


Reply all
Reply to author
Forward
0 new messages