#taskwarrior

/

      • diyfupeco has quit
      • diyfupeco joined the channel
      • xayto has quit
      • xayto joined the channel
      • dtzWill joined the channel
      • leowzukw joined the channel
      • leowzukw has quit
      • deitario1 joined the channel
      • deitarion has quit
      • bqf
        only now saw Unode's reply regarding the support emails. he's not here right now, but that report is about an event two weeks ago, and it doesn't explain the more recent missing mails at all, does it?
      • xayto has quit
      • xayto joined the channel
      • Haudegen joined the channel
      • simonwjackson joined the channel
      • simonwjackson
        Just a heads up for you Mirakel users.. I had to downgrade from the new 3.0 release to the previous version (2.8.1 i think). Taskwarrior server support appears to be broken and they don't know when/if they are going to fix it :/
      • diyfupeco
        ouch simonwjackson :(
      • simonwjackson
        yeah, and i think its the only android app that supports TW
      • tbabej
        simonwjackson: yep. Unfortunately the integration is not that good, which casued me to stop using it.
      • seschwar joined the channel
      • erikw joined the channel
      • Pest has quit
      • seschwar has quit
      • Pest joined the channel
      • seschwar joined the channel
      • Haudegen joined the channel
      • pbecking1
        bqf: Email is being dealt with today, I believe.
      • bqf
        pbecking1: i look forward to spam reaching me more consistently then :)
      • sliptonic has quit
      • sliptonic joined the channel
      • pbecking1
        bqf: I hope so, yes.
      • erikw has quit
      • danielsh joined the channel
      • chiromante joined the channel
      • danielsh
        morning, how do I write a filter to exclude tasks in proj:foo including subprojects? i.e., anything except '^foo($|\.)"
      • pbecking1
        danielsh: Which version?
      • danielsh
        pbecking1, 2.4.4 (debian testing/unstable)
      • pbecking1
        danielsh: Okay, then this: project != foo
      • the != is a partial non-match, ie not exact. !== is an complete non-match operator.
      • You'll need to probably do this: project '!=' foo
      • danielsh
        pbecking1, "The expression could not be evaluated."
      • (Also not sure I understand the distinction partial .v complete; and the man page doesn't doc !== at all.)
      • It works if I put the 'x != y' after the report name.
      • pbecking1
        dnaielsh: partial means this is a match: food = foo
      • danielsh complete means only this is a match: food == food
      • danielsh
        pbecking1, thanks, != does what I wanted,
      • pbecking1
        Partial match is a leftmost completion kind of thing.
      • danielsh
        Yes, I checked the source ...
      • pbecking1
        Source is always more reliable than me.
      • danielsh
        So != is "not startswith" and !== is "strcmp() != 0".
      • pbecking1, Amen to that.
      • Anyway, another one if you have time - http://sprunge.us/ONYV small docs patch
      • pbecking1
        Thanks, I have time.
      • bqf
        pbecking1: you have time? i'm not seeing the sort=cli commit yet...
      • danielsh
        Next one on my list, "task recent rc.report.recent.filter:'( limit:page )'ยจ gives me an error "No expression to evaluate.".
      • I worked around it by adding a second term within the parentheses that matches everything.
      • pbecking1
        bqf: sort=none is coming, but I'm in a mid-refactor and new command mess, so it waits.
      • task-git
        *** Taskwarrior - commit 3a7216c in "2.4.5", Daniel Shahaf: docs: Have report.X.annotations document its replacement
      • pbecking1
        danielsh: Yeah, that one. That's a bug, and 'limit:page' is not really a filter, it's some pseudo-attribute nonsense, and those parens are a mistake. The fix will be, eventually, to change 'limit:page' to 'rc.limit:page'.
      • danielsh
        Okay.
      • It works the same without the parens, but for my education, why are they a mistake?
      • pbecking1
        danielsh: (missed this earlier) != is "not startswith" and !== is "strcmp() != 0" --> exactly.
      • danielsh nods
      • Because parens should surround a valied expression, like (id == 1), but as limit:page is not a filter term, it evaluates to ( ), which is nonsense.
      • bqf
        pbecking1: still just teasing you, i don't particularly care when it's here. i currently have no use for it except making djp happy ;)
      • pbecking1
        The ( ) *should* be eliminated during the Dyjkstra Shunt, but I guess not, because I probably shove in an 'and' after it.
      • bqf
        it will be great for playing with experimental reports though, because the whole "display the report in a nicely accessible way" thing will be taken care of by Taskwarrior itself.
      • danielsh
        pbecking1, okay, so in short the parser needs to be made smarter about that.
      • One more, can I easily make the 'Status' column report 'Blocked' instead of 'Pending' when a task is blocked?
      • when I see 'Pending' I keep interpreting it as 'Actionable now', but blocked tasks aren't...
      • pbecking1
        bqf: Yes, I get it, it's coming soon.
      • danielsh: Yes, making the parser smarter is my calling in life. Working on it.
      • danielsh: No, you won't be able to easily make 'Pending' read 'Blocked'.
      • danielsh
        pbecking1, glad you've found your destiny :)
      • viq joined the channel
      • bqf
        danielsh: i just hide blocked tasks in my reports. would that not be an option?
      • danielsh
        bqf, I do too for some of my reports, but not in the 'per project' view.
      • pbecking1, why would it be difficult? Just change the label if .depends is non-empty... doesn't seem too hard.
      • bqf
        danielsh: yeah, i guess that makes sense. wouldn't want to think a project is mostly done when in fact it's just blocked. good thing i don't use projects. ;)
      • danielsh
        :)
      • chiromante has quit
      • bqf
        danielsh: because "status" is pending, waiting, deleted or completed. there are no other values. (unless i just missed one...)
      • danielsh
        bqf, I'm thinking first in terms of functionality, and in that layer 'blocked' is something that makes sense (to me, anyway)...
      • bqf, it's possible the data model doesn't map 1:1 to the values a user would want.
      • bqf
        danielsh: yeah, but it's somewhat hard to come up with a good representation here. what about blocked tasks that are waiting? what about completed ones?
      • danielsh
        bqf, I'd say "completed" trumps "blocked". Not sure about "waiting" v. "blocked"; they are similar.
      • Perhaps marking a task "waiting" is a form of acknowledging that it is blocked, so "waiting" should trump "blocked".
      • pbecking1
        danielsh: I meant it would be difficult as a user to do that. I could change a label, but ...
      • Here is what I would do: Add a new "format" for the status column, that indicates more than just the literal status value. If you have a compelling set of interpretations for 'status+others' then maybe.
      • nielsk_ joined the channel
      • nielsk_ has quit
      • danielsh
        pbecking1, "Blocked" is just one thing I ran into. I don't claim to have a complete set of interpretations here.
      • I'm not sure what are all the additional things users might want to see in the "Status" column.
      • pbecking1
        danielsh: Therein lies the problem. No one is asking for it, but if it's compelling, they'll like it, then djp will ask for 10 more things. But it needs that initial seed, otherwise no one will think of it.
      • c0b joined the channel
      • diyfupeco has quit
      • danielsh
        pbecking1, okay :) I'll figure out something locally for now see what I think after using it for awhile.
      • c0d has quit
      • pbecking1, bqf: thanks for the help!
      • pbecking1, I haven't forgotten about your threat re. functional programming, either.
      • pbecking1
        danielsh: Right, I'l respond to that soon
      • danielsh
        *nod*
      • By the way: that rc.sort= feature would be really useful for what I'm doing right now.
      • (writing a filter that implements my "Blocked" idea)
      • I think, at least. Pursuing another approach right now.
      • pbecking1
        danielsh: Yeha, it's a good idea, and it should be in 2.4.5. Unless I discover some horrible blocker.
      • I meant 'yeah', not 'yeha!'. Not that enthusiastic.
      • danielsh
        :)
      • hpincket joined the channel
      • bqf
        danielsh: was that second "t" in "threat" intentional? if so, i'd definitely love to know more. haven't seen pbecking1 threaten people with functional programming yet, but it sure sounds like something he'd do.
      • pbecking1
        bqf: I'll copy you on the conversation.
      • Doing that now, in fact.
      • Guest98795 is now known as maaku
      • simonwjackson
        Why are due dates mandatory for recurring tasks?
      • Vargos joined the channel
      • Pest has quit
      • pbecking1
        simonwjackson: Suppose you have two monthly recurring tasks...
      • simonwjackson: [1] Pay the rent
      • simonwjackson: [2] Pay credit card bill
      • Are those due on the same day? Does any day of the month work?
      • simonwjackson
        yeah, that makes sense. But the only down side is that this gives more priority to tasks that aren't really "due" like little reminders to take out the trash or someething
      • chiromante joined the channel
      • chiromante
        If I filter using a project that has dash in it tw thinks that I am subtracting strings. Is it a known bug?
      • simonwjackson
        i think its a known bug until 2.4.5
      • err.. fixed in 2.4.5
      • chiromante
        eheh, okay
      • simonwjackson
        not sure if this is a bug.. but i typed the year in wrong and got a strange error
      • task add schedule:2105-08-03T11:00:00 test
      • "Cannot subtract strings"
      • I wonder if that is part of the fix in 2.4.5
      • but it works if i change it to 2015
      • chiromante
        well I did "task pro:my-project" and had the same error
      • while "task add pro:my-project something" works perfectly
      • pbecking1
        chiromante: The "2105" thing is a new bug. Good catch.
      • chiromante
        pbecking1: credits for that goes to simonwjackson
      • pbecking1
        simonwjackson: Thanks. The validation for 'year' is anything between 1900 and 2100. I'll extend that a bit.
      • simonwjackson
        no worries :)
      • I cant seem to get ISO format to work here: report.agenda.dateformat=hh:mm
      • am i missing some special syntax?
      • pbecking1
        simonwjackson: H:N
      • simonwjackson: 'man taskrc', look for 'dateformat'.
      • simonwjackson
        i was looking for that, but my ultimate goal was to get a report to use 12hour syntax
      • looking at that*
      • pbecking1
        Not sure 12-hour is possible, what with it being ambiguous.
      • simonwjackson
        yeah, its just hard getting used to :) my brain works in 12h mode
      • i've been living in europe for over 2 years and i still cant get used to it
      • :)
      • pbecking1
        simonwjackson: Yeah, sorry, that's not suported.
      • danielsh_ joined the channel
      • danielsh has quit
      • awwaiid has quit