#taskwarrior

/

      • rmk0 joined the channel
      • rmk0 has quit
      • rmk0 joined the channel
      • calmar has quit
      • pbeckingham
        coddingtonbear: Bad in that it doesn't do a good job of validating certs. The question is, what version of GnuTLS is it built with? That's what determines restarts and not responding.
      • rmk0 has quit
      • riclima joined the channel
      • riclima joined the channel
      • coddingtonbear
        ahh -- 2.12!
      • im'ma upgrade right now
      • riclima has quit
      • bqf
        coddingtonbear: did you see my note about "task import" for updating tasks being ready for testing? still no rush, i just want to check "ping coddingtonbear about possibly using import for taskw" off my mental TODO list.
      • coddingtonbear
        ahh, yeah, I did! thanks for the heads-up
      • it'll be a little while before I have much time to look at it, though, I'm afraid
      • incidentally -- what's the minimum version of taskwarrior for using `task import`?
      • we'll have to build in a little logic for using the old ways for earlier versions
      • bqf
        using "task import" for updating - 2.4.5, which has not been released yet. "task import" for importing JSON - no idea, just tried scanning the changelog only to be none the wiser. :)
      • and as i said - no rush. it should "just work"(tm), and the 2.4.5 branch is currently stable enough to play around with. i'd prefer someone apart from me play with it before it gets released as part of 2.4.5, but if that isn't in the cards then so be it. someone will find my screwups sooner or later.
      • i couldn't let it go. cdda791c84c659e0e0855ea780c0ceef041e94d6, Sun Aug 7 02:10:50 2011 -0400, "Implemented the import command, which only suports JSON. (...)". so Taskwarrior 2.0.1+, with probability of bugs being pretty high.
      • Haudegen has quit
      • Haudegen joined the channel
      • Kalendium joined the channel
      • rmk0 joined the channel
      • fling
        How to merge tasks together?
      • I have a dupe ;>
      • haircode joined the channel
      • haircode has left the channel
      • haircode joined the channel
      • haircode has quit
      • haircode joined the channel
      • haircode has quit
      • Kalendium has quit
      • haircode joined the channel
      • haircode has quit
      • bqf
        fling: there's no process for this, because it's impossible to know which parts you want from each. your best bet is modifying one of them to include the attributes you need from the other, then delete the redundant one.
      • fling
        bqf: already done this way, thanks…
      • Just upgraded to 2.4.4 and it works fine.
      • bqf
        fling: i'd hope so. a stable release not working would be pretty bad. ;)
      • fling
        There were issues with 2.4.1~ version afair…
      • bqf
        fling: which is exactly why there is now a release >2.4.1 - because the previous ones were bad in some way.
      • ... where "bad" can mean as little as "it's missing feature $X", or "hooks are broken, we need to fix that". the latter is probably what you are remembering about 2.4.1 :)
      • fling
        haha ok
      • haircode joined the channel
      • bqf
        djp: just googled "taskwarrior imask". the taskextras mirror of the old task wiki is all over the results, and there is no indication whatsoever that the contents are outdated. that's... i'm not going to say "bad", but probably not "good" either. it's definitely not not bad.
      • pbeckingham: recur.cpp:updateRecurrenceMask() - "mask" scope looks... funky. line 393 needs to go. i haven't looked at recurrence enough to know WTH is supposed to happen there, so i'm only pinging you about it. i'd say that whole else block is weird, but what do i know. :)
      • lydgate joined the channel
      • Brock joined the channel
      • latestrevision joined the channel
      • oddshocks has quit
      • coddingtonbear has quit
      • awwaiid has quit
      • tbabej has quit
      • oddshocks joined the channel
      • tbabej joined the channel
      • haircode has quit
      • haircode joined the channel
      • haircode has quit
      • RenRenJuan joined the channel
      • JuanDaugherty has quit
      • haircode joined the channel
      • haircode has quit
      • haircode joined the channel
      • haircode has quit
      • haircode joined the channel
      • antroy joined the channel
      • antroy
        Hi all. I have been using taskwarrior for a few weeks, and so far am loving it.
      • However I am having problems with hyphens.
      • I am using JIRA task IDs as project names in taskwarrior: task add Fix the bug project:cx-123
      • However I can't then search on the project id
      • e.g. task project:cx-123 next
      • gives the error:
      • Cannot subtract strings
      • haircode has quit
      • I have tried single quoting the entire command, single and double quoting the project name, with and without escape backslashes.
      • Any help on this appreciated.
      • tbabej
        antroy: you have to drop the hyphens, I'm afraid
      • antroy: it's a known bug
      • haircode joined the channel
      • antroy
        Oh - ok
      • You mean in the search functionality?
      • OK - looked at the ticket. Cheers.
      • tbabej
        antroy: You're welcome.
      • Kalendium joined the channel
      • haircode has quit
      • synthmeat
        is it possible, without hooks, to increase priority of tasks which have exact time, not only date, specified during creation?
      • on a glance, i don't think so, default time is "ingrained" and don't see how would one differentiate between "default" 00:00 and "specified" 00:00
      • Unode
        synthmeat: you might want to use tags instead and then you have some metadata to distinguish on.
      • synthmeat
        yeah, i'm considering that, but i'm afraid i'm going to have to go with hooks or something
      • i need that task to have topmost priority immediately before due time
      • this is bit going from "tasks" to "events", so i'm not sure am i to even criticize on this
      • i think much of this would be handled ok in my case if i were to reduce coefficient of tasks other depend on
      • they're mostly the ones making the priority list a bit wonky. will try that
      • the "blocking" is the terminology, i think
      • oh my, yes. setting it to 4.0 was it. cool!
      • besides changing weekstart to monday, this is the only "semantic" tweak i had to do over default configuration. pretty good.
      • bqf
        synthmeat: 2.4.5 will have "better urgency inheritance". tasks inherit their blocked tasks' urgency if and only if it is higher than their own. "blocked" and "blocking" priority coefficients are then redundant. i have them set to 0.0 currently, and it's great.
      • synthmeat
        oh, yeah, that makes even more sense
      • bqf
        that whole "blocking tasks have higher priority because they unblock other tasks" thing has been annoying me for a while. just because a task unlocks more tasks it doesn't deserve a higher urgency if other things are actually, you know, urgent.
      • synthmeat
        yes, making it just over any other singular task they block is exactly what i think would work
      • .01 or something over that, i think i may have seen that .01 thingie around in priorities
      • maybe it was coincidence, dunno
      • bqf
        synthmeat: if you want to play with it feel free to get and build the 2.4.5 dev branch. i can always use feedback.
      • synthmeat
        bqf: sure, why not. lemme double-check on backups
      • bqf: that's only for client, right?
      • bqf
        synthmeat: yeah, the current urgency.inherit adds 0.01 to the blocking task's urgency, but that is only done in case users run reports where blocked tasks are not hidden. i hide them, and can do so without feeling bad now that any urgent task will have its blocking task moved to the top of my "next" report.
      • synthmeat: backups are always a good idea, yeah. currently 2.4.5 is pretty stable, and by that i mean "no known problems apart from those already present in 2.4.4" - but you never know.
      • and yes, you only need to update the client.
      • synthmeat
        uuuh, yeah, another good one! i should hide blocked ones too, dunno why i show them at all
      • bqf
        synthmeat: you show them because it's the (broken, IMHO) default for "next". i use 'report.next.filter="status:pending limit:page +READY'.
      • synthmeat
        bqf: this is not criticism either, but any particular reason you guys aren't on github?
      • Unode
        synthmeat: that's a recurring question. Nowadays everyone assumes github is the place to be.
      • we go have a repository on github, but it's not the main platform. It's just a clone of the official one.
      • *we do
      • synthmeat
        i've lived through sourceforge up and down, so i understand the "perks" of self-hosting, sure
      • just that i'm pretty sure you might be missing on some exposure as well as contributors
      • Unode
        I can't say much about it. When I joined the infrastructure was already in place, so I took it as is.
      • Lynoure
        I find a commercial entity being "the place" for git kind of creepy.
      • Unode
        synthmeat: as I said, there's a repo there. Paul can probably tell you how much feedback comes from github users.
      • synthmeat
        yeah, and there's that whole "people using issues as support for usage" thing that's a lot of noise, i bet
      • Unode
        Lynoure: they do make it confortable and attractive :)
      • synthmeat
        Lynoure: yeah, it is a bit scary
      • it's ok for development, i think
      • but some package managers basically run off github repos
      • that's gonna break one day horribly, i think
      • so many node modules, homebrew formulas, ugh.
      • js and ruby ecosystem thrives mostly there
      • bqf
        synthmeat: it's not self-hosted, it's all backed by atlassian products. it used to be self-hosted, then everything was moved over to stash/jira/(...)
      • synthmeat
        ah
      • bqf
        if there was a discussion, and i'm not sure there was one, i'm pretty sure my input into it would have been "i'd prefer github, but if you want something else i can live with it", possibly with some ranting along the way.
      • synthmeat
        yeah, it's not like you're not one command from another remote upstream anyways
      • bqf
        i hated redmine when it was used (before jira), and i hate jira now. that's the way it goes with bugtrackers. haven't found one that just gets everything right, they're all equally broken in different ways.
      • synthmeat
        well, i preferred redmine over trac 4-5 years ago
      • antroy has quit
      • bqf
        synthmeat: yeah, of the two turds, redmine was the shinier one.
      • synthmeat
        bqf: urgency inheritance is not retroactively calculated somehow?
      • since i didn't note any ordinal change
      • (in 2.4.5, just built it)
      • bqf
        synthmeat: run with "task rc.urgency.inherit=on" to see the difference.
      • as i said, my suggestion would be setting these: urgency.inherit=on urgency.blocked.coefficient=0.0 urgency.blocking.coefficient=0.0 report.next.filter="status:pending limit:page +READY" in your ~/.taskrc
      • synthmeat
        bqf: everything works splendidly well on both ubuntu and os x :)
      • <3
      • d=(´▽`)=b
      • k, will note any issues i may encounter. super cool, you guys are all amazing
      • c++ is crap tho ;D
      • bqf
        synthmeat: i wouldn't call it crap. just needlessly complex and unsuited for many things it tends to be used for. but hey, feel free to start a Taskwarrior clone in Python ;)
      • synthmeat
        pfft, python!?
      • that's even worse
      • i'd go C and regret it few k LOC in
      • bqf
        Taskwarrior89, in ANSI C. has a nice ring to it.