Mailing List Archive
tlug.jp Mailing List tlug archive tlug Mailing List Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]Re: [tlug] Chrome ate my profile
- Date: Fri, 6 Apr 2018 13:53:14 +0900
- From: "Stephen J. Turnbull" <turnbull.stephen.fw@example.com>
- Subject: Re: [tlug] Chrome ate my profile
- References: <23205.16405.527594.15212@turnbull.sk.tsukuba.ac.jp> <20180311165823.GA1669@wismut> <ba01bbb1-32d0-7b3c-b3b3-b4d0ed4c6d9a@dcook.org> <23210.6885.521848.717987@turnbull.sk.tsukuba.ac.jp> <20180315104007.26t7l46c6hqfrdxm@iambic.cynic.net> <23211.61661.137765.239583@turnbull.sk.tsukuba.ac.jp> <20180318014444.i5wg6pbq6b4fqwsm@idyllic.cynic.net> <23215.28961.204351.418662@turnbull.sk.tsukuba.ac.jp> <20180319093203.njtzgav5kc2h3hdr@iambic.cynic.net>
Coming back to this after spending a week solving a knapsack problem.[1] Curt Sampson writes: > On 2018-03-19 17:13 +0900 (Mon), Stephen J. Turnbull wrote: > > > I don't understand how it can be easier than save, fix, save (look ma, > > no explicit git here!) > > ... > > Sounds exactly like what I do have except that I never need to invoke > > git explicitly for this functionality, and it's never on a branch that > > I might push. > > I don't see how overall that's easier because you still need to turn > these "saved" things (wherever you're saving them) into commits They *are* commits. They're just on a branch that is used only by this function, which function is invoked by XEmacs automatically every time I save. This is the same workflow that you describe, except that I never invoke git explicitly until "git rebase --interactive": I just save. The after-save-hook makes sure that I'm on that branch, then commits. Then when I'm done with the work that requires a certain amount of concentration (I can fix a typo but not much more than that or I might lose the thread) I switch into "coalesce related commits" mode. So not only does saving mean I don't lose work if XEmacs crashes (this doesn't happen as often as Mac OS locking up, so it's not really a thing any more :-), it provides "parentheses" around small tasks that distract from my main workflow. In the rebase phase, all the typo fixes go on a "for immediate release" branch, any small bug fixes or tiny improvements that still need review go on another branch, and the "real work" goes on a long-lived feature branch. Then I reset the "auto" branch head to the end of the feature branch, and start the process again. Footnotes: [1] You have a bunch of objects of various sizes, and a vessel of fixed size you wish to fill exactly. Choose a subset of the objects. This is in general NP-hard[2], but it turns out to be quite tractable when the objects are purchases (mostly of Kindle books, but there were a couple of industrial strength laser printers and the like in there too), and the vessel is a 100,000 yen budget that needs to be spent to the last yen. :-) [2] No, I didn't solve the NP-hardness part. ;-)
- Follow-Ups:
- Re: [tlug] Chrome ate my profile
- From: Curt Sampson
Home | Main Index | Thread Index
- Prev by Date: [tlug] fcitx/mozc IME: black on black
- Next by Date: Re: [tlug] Chrome ate my profile
- Previous by thread: [tlug] fcitx/mozc IME: black on black
- Next by thread: Re: [tlug] Chrome ate my profile
- Index(es):
Home Page Mailing List Linux and Japan TLUG Members Links