From 11/15/2009 to 12/14/2009

12/14/2009

11:42 AM Bug #356 (Closed): Report not sorting correctly on completion date
Paul Beckingham
11:41 AM Bug #356: Report not sorting correctly on completion date
I actually happened to be looking at the koji page last Friday and it showed that task-1.8.5 was built successfully. Don't know why it hasn't made it into testing or updates though. John Florian
10:25 AM Bug #356: Report not sorting correctly on completion date
Regarding 1.8.5 and F10 we will have to see when and if 1.8.5 made it into F10. Build was submitted before cut-off date, but you never know with the build-queue list... I can't check right know as the build and update system has a planned-outage f... Federico Hernandez
09:56 AM Bug #356: Report not sorting correctly on completion date
On further review, it may be that the end date is being sorted on, but possibly as an alpha-string rather than a numerical date as I saw the following sequence for one project: 12/10/2009, 12/11/2009, 12/7/2009, 12/8/2009. Hope this helps. John Florian
09:51 AM Bug #356: Report not sorting correctly on completion date
I have the following report defined: ... John Florian

12/13/2009

11:09 PM Feature #193: pcal plug-in
> ... would it not be easiest from both user perspective and programmer perspective, > to have "include all these" and "exclude all these" lists? I think providing the user with a simple filtering mechanism is most important. Listing all high... Paul Beckingham
09:38 PM Feature #193: pcal plug-in
Paul Beckingham wrote: > This one: > Has no results because a task cannot have a project with two values. > No disjoint operations yet (see http://taskwarrior.org/issues/show/142). right, not yet, ok. > > filter=status:pending,proj:foo,bar >... David Patrick
05:02 PM Bug #355: Config::load called twice
It appears that Config::load is called twice, no matter what. Why did I do that? Paul Beckingham
12:07 PM Feature #19 (Closed): TDB support for multiple files
This was actually finished a while ago. Paul Beckingham
11:22 AM Feature #193: pcal plug-in
This one: filter=status:pending,proj:foo,proj:bar Has no results because a task cannot have a project with two values. Those filter elements are all ANDed together. No disjoint operations yet (see http://taskwarrior.org/issues/show/142). ... Paul Beckingham
09:44 AM Bug #354: Unit tests won't build without ncurses
The unit tests need to be built with autotools too. Currently they expect ncurses to be installed, which is not necessarily the case. This is not user-visible. Paul Beckingham

12/12/2009

06:11 PM Feature #193: pcal plug-in
well that's dandy! just the ticket! and, if I may ask, what of filter=status:pending,proj:foo,proj:bar because filter=status:pending,proj:foo,bar would not work (?) (comma separation collision ?) Or perhaps I'm confusing behavior at the cli ... David Patrick
05:10 PM Feature #193: pcal plug-in
filter=status:pending,proj.not:foo,pri.not:l That would give you a report of pending tasks that were not project foo and had either no priority, H or M. Paul Beckingham
04:53 PM Feature #193: pcal plug-in
Paul Beckingham wrote: > Sorry, it's the other way around. The filters filter in, not out. then how would "filter=status:pending,proj.not:foo,pri.not:l" behave ? > Guess it's more of a selector than a filter, huh? I think so. more like "includ... David Patrick
01:57 PM Feature #193: pcal plug-in
Sorry, it's the other way around. The filters filter in, not out. Guess it's more of a selector than a filter, huh? Paul Beckingham

12/11/2009

11:52 PM Feature #193: pcal plug-in
In futzing with pcal and the resulting printed calendars, I have a couple of potentially effective (and/or annoying) ideas, that may spawn new #features. They are both related to the report/export intended for pcal processing (but applicable anywh... David Patrick

12/09/2009

07:29 PM Support #353: Decision required on case sensitivity in confirmation dialogs
For me: * [no] as default makes most sense * as a rule I think case sensitivity is a bit of a pain * I don't see any pressing need to perform bulk undos, but if Paul/David feel frisky... :) Cory Donnelly
05:06 PM Support #353: Decision required on case sensitivity in confirmation dialogs
I am going to advocate for [no] as the default action for bulk changes. [OOPS! Fredde, I edited, instead of quoted your post.. .. you were saying ? That sounds like a good default, especially if I can change it to [all] and even more so if it we... Federico Hernandez
04:57 PM Support #353: Decision required on case sensitivity in confirmation dialogs
I wouldn't advocate case sensitivity in confirmation dialogs, that sounds annoying. I may have _seemed_ like I was pushing for it, as I drew my mock-up dialogs with a capitalized letter, but that was to indicate the default action on <enter>, not ... David Patrick
04:48 PM Support #353: Decision required on case sensitivity in confirmation dialogs
If you make a bulk change that affects three tasks, then you have three separate undo operations, because the undo is not also bulk. That might make a nice enhancement - to make either a single or batch undo. Would it be easy? No. No it woul... Paul Beckingham
04:36 PM Support #353: Decision required on case sensitivity in confirmation dialogs
I consider myself a unix guy and would not assume a default action for this confirmation dialog. A default answer for me is @[yes]@ or @yes/[no]/all/quit@. When using @rm -i@ the _-i_ is step one in the confirmation, _y_ is step two. Using @Yes... Federico Hernandez
04:00 PM Bug #347 (Closed): Confirmation dialog is lowercase for "all"
Federico Hernandez
03:29 PM Feature #352: rc file should support includes
agreed 100% and as I'm delighted to find, the .calendar format supports them, too. include.cal ~/.task/canadian.holidays.calendar include.cal ~/.task/usergroups.calendar or just include.cal ~/.task/MyCalendar and it #includes the rest it _w... David Patrick
02:22 PM Support #353: Decision required on case sensitivity in confirmation dialogs
It's come up a few times in the past week or so -- I misinterpreted a deliberately-implemented feature for a bug in Issue #346, Federico corrected me and submitted the inverse in Issue #347, and then David was confused in Issue #350. I think we... Cory Donnelly
11:54 AM Feature #352: rc file should support includes
The rc file should support include files, like this: ... Paul Beckingham
10:50 AM Feature #351: milestones; a measure of success
A project is just a multi-step task. In project management, milestones are used to indicate completion of important stages of a project. A milestone is not a task itself, but a marker to indicate progress. Milestones could be implemented a... David Patrick
08:58 AM Feature #337: productivity rewards
tcharles yun wrote: > i want to chime in that this is a great idea; i like it; i want it. and eventually, you shall have it! my question is how milestones are visualized during a simple `task list` scenario. There are two distinct thin... David Patrick

12/08/2009

08:51 PM Feature #343: Allow spaces in Projects
I want to say no, too, but I'm not sure how much of that is based on my own usage. (I would not use it) Should it be possible ? Is there a logical reason to want this ? Is there a technical reason not to do this ? I can not say.. David Patrick
08:18 PM Bug #350 (In progress): deleting multiple tasks uses wrong confirmation dialog
Federico Hernandez
08:13 PM Bug #350: deleting multiple tasks uses wrong confirmation dialog
bulk delete with confirmation; yes in the first example above, Proceed with change? (Yes/no/All/quit) only ONE (or none) options should ever be capitalized, indicating default action on <enter> David Patrick
07:47 PM Bug #350: deleting multiple tasks uses wrong confirmation dialog
I say yes, provided it needs confirmation. If we prevented bulk deletes, I don't think it would be too long before someone asked for it. Paul Beckingham
07:29 PM Bug #350: deleting multiple tasks uses wrong confirmation dialog
The question is should we permit bulk operations for the delete command? Federico Hernandez
07:24 PM Feature #193: pcal plug-in
an excellent .calendar, including undated tasks, could easily be generated by an external script (warning; psudo-code ahead) $ task export.calendar proj:foo,bar due.any: > FooANDBar.calendar # filtering out due.none: $ echo '$TOMORROW ... David Patrick
07:09 PM Feature #193: pcal plug-in
undated tasks in export.calendar problem partially solved; solution inherited! After closer inspection of pcal documentation, it seems that if a [description] does not have an associated date, then that date is inherited from the event above. Thi... David Patrick
07:02 PM Feature #343: Allow spaces in Projects
I want to say no, only single word tags. But I don't have a good reason. I need to defer on this. David? Paul Beckingham
06:34 PM Bug #347 (In progress): Confirmation dialog is lowercase for "all"
Federico Hernandez
06:34 PM Bug #350 (Assigned): deleting multiple tasks uses wrong confirmation dialog
Federico Hernandez
06:34 PM Bug #350: deleting multiple tasks uses wrong confirmation dialog
When changing multiple tasks a 'Yes/no/All/quit' confimation dialog is shown. For example: ... Federico Hernandez
06:22 PM Feature #337: productivity rewards
i want to chime in that this is a great idea; i like it; i want it. my question is how milestones are visualized during a simple `task list` scenario. specifically, what happens when one has tasks with different: due dates, project names, et... t charles yun
05:47 PM Feature #343: Allow spaces in Projects
would this thinking apply to tags as well? task add "woo hoo" +"i get to use spaces in tags too!" t charles yun
12:40 PM Bug #348: task done a little over zealous
I do generally use the form: ... John Florian

12/07/2009

10:33 PM Bug #348: task done a little over zealous
Yes, both ... Paul Beckingham
10:25 PM Bug #347 (Assigned): Confirmation dialog is lowercase for "all"
Paul Beckingham
10:15 PM Feature #307 (Closed): taskrc.vim
after botching the uber simple install instructions, I got this in place and it looks GREAT ! way better, thank you John ! there's already a ticket re: folding, but I think we need to firmly establish the rc file layout and heading styles. David Patrick
10:09 PM Bug #293: tag usage inconsistency
David, neither of your +ghost-busters examples would yield any matches. Tags now match completely or not at all. Paul Beckingham
09:56 PM Feature #349: project descriptions
In entering a task, and in displaying a list, we will always want to use the shortest, most compact names for our projects. Sometimes, however, the project name really deserves to be more descriptive, especially if other people will be working f... David Patrick
09:43 PM Bug #348: task done a little over zealous
Maybe just clarifying the syntax ? is $ task 127 do correct ? shouldn't it be $ task do 127 ? Especially with "done" we should be picky with the structure. And isn't it a little bizarre that it accepted "do" as both short-for-do... David Patrick
08:52 PM Bug #293: tag usage inconsistency
I actually like the new syntax better. Using +/- as selectors always seemed double-duty to me. I like them for adding/removing tags and accepted them as filters simply because of their brevity. However, I don't see the new syntax as having to k... John Florian
07:58 PM Bug #348: task done a little over zealous
Consider the following: ... John Florian
06:32 PM Bug #347: Confirmation dialog is lowercase for "all"
The confirmation dialog uses uppercase "Y" for yes (as an extra step for confirmation by using 2 keys), but the corresponding "all" is in lowercase. Federico Hernandez
06:30 PM Bug #346 (Rejected): Confirmation dialog response is case-sensitive, but shouldn't be
The uppercase "Y" is used intentionally. As an extra step to confirm changes by having to use 2 keys. The bug would be the lowercase "a" in all. Federico Hernandez
11:41 AM Bug #346: Confirmation dialog response is case-sensitive, but shouldn't be
This bug existed previously, but I only noticed it in 1.8.5 because of a change made when the 'quit' option was introduced. Patch to util.cpp attached. Cory Donnelly
08:50 AM Bug #293: tag usage inconsistency
Paul Beckingham wrote: This meant adding new "word" and "noword" modifiers, and breaking the functionality that John was enjoying. then, considering the new "word" and "noword", if we $ task add blather +ghost-busters how would $ task list ... David Patrick
08:31 AM Bug #293: tag usage inconsistency
John was enjoying partial tag matches (task list +artial) but missing partial color rules for tags (color.tag.artial=...). Internally, tags are stored as: ... Paul Beckingham
07:57 AM Feature #340: deleted and completed tasks need color rules, too
good default color settings for deleted and completed items would be color.completed=invert # bg/fg reversal of colors and color.deleted=black on_white # based on the opposite of light/dark terminal defaults these should stand out hard. I... David Patrick
07:29 AM Bug #293: tag usage inconsistency
please to explain.. the problem was.. the solution is.. ? David Patrick
01:30 AM Bug #293: tag usage inconsistency
I have implemented two new attribute modifiers - 'word' and 'noword' for positive and negative whole-word matching (plus docs, help, unit tests etc). Then I used these attribute modifiers to re-implement: ... Paul Beckingham

12/06/2009

11:32 PM Bug #293: tag usage inconsistency
Yuck. The following example of John's: ... Paul Beckingham
01:05 PM Feature #345: autoshell
When a user enters an incomplete task, one where required attributes are missing, they should have the option of automatically going into shell-mode, and be prompted for the missing elements, OR seeing a message about the missing attributes. This ... David Patrick
11:33 AM Feature #344: Configurable policies
That looks like the kind of thing the shell should do. If that were used in a script, it would hang. Paul Beckingham
11:29 AM Feature #344: Configurable policies
instead of $ task add project:Audit Prepare the 2009 books priority ? (l.m.h.) > ... project ? > ... etc ? David Patrick
10:44 AM Feature #344: Configurable policies
Task could make use of policies that govern how it is used. For example, there are those who believe every task needs a priority and a project. This is an example of a policy. Others might be: ... Paul Beckingham

12/05/2009

11:59 PM Feature #343 (Assigned): Allow spaces in Projects
You're right. A single word project is inadequate. Text is text, and spaces are text too. Paul Beckingham
11:01 PM Feature #343: Allow spaces in Projects
I _could_ do that, but it's a bit constraining. That's what I was doing before I started cheating by using the '_' character. In (my understanding of the) GTD world, projects are nothing more than multi-step tasks and should always begin with a... Cory Donnelly
10:14 PM Feature #343: Allow spaces in Projects
Cory Donnelly wrote: > In the examples people throw around on the forums, projects are always single words. 'Camping' or 'Wedding'. In reality, mine look more like: > > * Wedding.Prepare_for_reception.Compile_dancing_playlist > * Wedding.Pr... David Patrick
10:03 PM Task 1.8.5 (PAUL) is released
This is a bug fix release and it is recommended to update to this version. 1.8.5 will be the last release in the 1.8 series. Further possible bug fixes will be released together with the upcoming 1.9.0 before we focus developement on the interacti... Federico Hernandez
07:24 PM Feature #310 (Closed): 'task add' with external editor
Federico Hernandez
07:04 PM Feature #193: pcal plug-in
Paul Beckingham wrote: > There are two things at play here. The first I presume is mere filtering for an export command. Easy. > The second thing I presume is capturing the output > This is provided by the shell, so it's a free feature. YAYY FRE... David Patrick
05:07 PM Feature #343: Allow spaces in Projects
I'm with Paul and David -- attribute:"quoted text" is consistent and intuitive. I also see the value in not impacting the simplicity of 'task add foo' Cory Donnelly
04:44 PM Feature #193: pcal plug-in
There are two things at play here. The first I presume is mere filtering for an export command. Easy. ... Paul Beckingham
04:12 PM Feature #193: pcal plug-in
And (because I clearly give no thought to Pauls' sanity or workload) is this (hypothetically, remotely) possible ? $ task export.calendar proj:foo MyJobs.calendar $ task export.calendar +bar >> MyJobs.calendar and then what of duplicates ? David Patrick
04:00 PM Feature #343: Allow spaces in Projects
I'm all for establishing the attribute:"you need to quote or escape spaces" pattern, with very few exceptions (eg; desc) There are just too many places where an attribute using a string with unquoted spaces is just un-parseable. David Patrick
03:24 PM Feature #343: Allow spaces in Projects
I just went searching for this but couldn't find it: we have an (apparently) unwritten agreement that we will not make any changes to the basic functionality of task. This means that the following commands will always work and do exactly what th... Paul Beckingham
03:22 PM Feature #193: pcal plug-in
Then there is the matter of tasks, which have no due:date, but you want to include them in your calendar. This makes more sense when you consider other .calendar processors, which can create page-per-day "day-timer" style pages. It's a bit of a ... David Patrick
02:54 PM Feature #343: Allow spaces in Projects
In the examples people throw around on the forums, projects are always single words. 'Camping' or 'Wedding'. In reality, mine look more like: * Wedding.Prepare_for_reception.Compile_dancing_playlist * Wedding.Prepare_for_reception.Compile_ea... Cory Donnelly
02:25 PM Feature #193: pcal plug-in
Paul Beckingham wrote: > So what would it take - an export.pcal command and some kind of template file to combine it with? First of all, the target file format (.calendar) may be the pcal default, but it is (and will be) used by a handful of oth... David Patrick
01:11 PM Feature #193: pcal plug-in
Good point - this is a specialized export. So what would it take - an export.pcal command and some kind of template file to combine it with? Then we could provide multiple templates. Paul Beckingham
01:06 PM Feature #193: pcal plug-in
this might be better thought of as $ task export.calendar [date-range][desc][attrs][tags] filename.calendar David Patrick
12:16 PM Support #342: Doc needed - a comprehensive 'setup' walkthrough
The wiki needs a really good setup walkthrough. Not the install itself, but the next steps, which include: * $PATH * .taskrc generation * .taskrc color rules, 256-color support * .taskrc aliases * shell aliases * using the man pages * ho... Paul Beckingham
11:19 AM Feature #341 (Closed): Autogenerated .taskrc should reference man taskrc
Thanks for the patch Cory. I did tweak it a little, and also added similar to the "version" command output. Paul Beckingham
11:12 AM Feature #341 (Assigned): Autogenerated .taskrc should reference man taskrc
Paul Beckingham
11:05 AM Feature #341: Autogenerated .taskrc should reference man taskrc
I've been using various flavours of linux for years and have a mac now, but it never occurred to me to check to see if a man page had been created for .taskrc (or task itself) as Paul suggested in Issue #335 -- can I respectfully propose a minor c... Cory Donnelly
10:16 AM Feature #337: productivity rewards
Paul Beckingham wrote: > The feature I was thinking of (which I don't think ever came up in conversation) was milestones. > But I could never decide if it was useful or just silly. Silly was the conclusion. VERY interesting and basically the sa... David Patrick
12:32 AM Feature #336: new command; prepend:
Paul Beckingham wrote: > In order to be a symmetrical twin of the "append" command, prepend would work like this: > > ... David Patrick

12/04/2009

11:40 PM Feature #263 (Rejected): NFS-safe file locking
Too much of a hassle. Paul Beckingham
11:39 PM Bug #315: Unexpected behaviour - marking task as done releases task ID
This is a tricky one to fix. I'm pushing it back to the 1.9.0 release to allow time for some rethinking of the shadow file call sequence. Paul Beckingham
11:07 PM Feature #340: deleted and completed tasks need color rules, too
status:completed tasks are currently almost never seen, as (I think all but "completed") reports filter them out, but being able to view those tasks, in mixed company with status:pending tasks, would actually be very useful in the analysis of proj... Paul Beckingham
11:05 PM Bug #317 (Closed): Output produced by `task completed` shouldn't use due-related colouring
There is both a bug and a feature in this one issue. That's bad, for tracking reasons. once a bug is fixed, it's number goes into the ChangeLog file permanently, so there should be no revisions on closed bugs, unless they were not actually fixed... Paul Beckingham
10:58 PM Feature #318: timestamp for status changes
Changed title - "Report Enhancements" too generic. Paul Beckingham
10:55 PM Feature #335 (Rejected): tab completion should complete projects, tags, etc., found in completed....
Paul Beckingham
10:55 PM Feature #335: tab completion should complete projects, tags, etc., found in completed.data as wel...
This is already implemented folks. Check the man page, and look for "complete.all.projects", for details: ... Paul Beckingham
10:46 PM Feature #337: productivity rewards
The feature I was thinking of (which I don't think ever came up in conversation) was milestones. It would work something like this: You first create a milestone ... Paul Beckingham
10:32 PM Feature #337: productivity rewards
Paul Beckingham wrote: > Could you include some real command line examples of using this, and how task responds? Not really and yes Not really because other that changing the reward values with something like $ task rc reward.daily.6=pri:l ... David Patrick
10:21 PM Feature #339: New date attribute required for GTD compliance: last reviewed date
Paul Beckingham wrote: > Adding a review command to update the "last-review:" attribute is a simple addition. Would there be any other use of this attribute? Would there be any other ways to exploit it? For me, the idea of a Review process fall... David Patrick
10:07 PM Feature #336 (Closed): new command; prepend:
Paul Beckingham
09:37 PM Feature #336: new command; prepend:
In order to be a symmetrical twin of the "append" command, prepend would work like this: ... Paul Beckingham
09:31 PM Feature #337: productivity rewards
Could you include some real command line examples of using this, and how task responds? It sounds very similar to an old feature I couldn't think of a name for and never implemented. Paul Beckingham
09:28 PM Bug #338 (Rejected): nitpick: task /foo/bar/ doesn't play nicely with spaces
Paul Beckingham
09:28 PM Bug #338: nitpick: task /foo/bar/ doesn't play nicely with spaces
Cory, this problem has bitten me many times. Sometimes I leave off the trailing slash which creates the same kind of mess. But this is actually a shell problem. When the shell see spaces in a command line, it considers the space to be a break... Paul Beckingham
09:20 PM Feature #339: New date attribute required for GTD compliance: last reviewed date
Adding a review command to update the "last-review:" attribute is a simple addition. Would there be any other use of this attribute? Would there be any other ways to exploit it? David, Implementing this requires maintaining per-task state. I... Paul Beckingham
05:22 PM Bug #308 (Rejected): task shell needs tab completion, history and color
There is already a feature request for this - so it can't be a bug as it isn't implemented yet. Federico Hernandez
03:36 PM Feature #339: New date attribute required for GTD compliance: last reviewed date
I really like this concept. I have a weekly recurring task that tells me to review all my tasks. It's a huge pain to go through several hundred regularly like this, but I often find tasks that I've actually completed already. If this were a sep... John Florian
12:02 PM Feature #339: New date attribute required for GTD compliance: last reviewed date
That's a very GTD suggestion, Corey! Reviewing specific time periods, and tracking what you have been over, would make Mr. Allen proud. I have already suggested a Review process, and this modality should be strongly considered, but I'm not sure ... David Patrick
10:55 AM Feature #339: New date attribute required for GTD compliance: last reviewed date
A big part of GTD is periodic (daily, weekly, monthly, yearly) reviews of ongoing projects. A simple way to begin supporting this is to add a new 'last reviewed date' attribute to tasks, allowing users who wish to indicate such that they've been r... Cory Donnelly
09:08 AM Feature #326: add command; log
In my mind, task log should: * Set entry date to current date and time * Set due date to whatever the user specifies (leaving it empty if the user doesn't provide anything) * Set completed date to whatever the user specifies (using the curren... Cory Donnelly
08:41 AM Bug #338: nitpick: task /foo/bar/ doesn't play nicely with spaces
I experience this under v1.8.4. Unless the user escapes /foo/bar/[g] with quotes, task doesn't correctly parse the command when foo or bar contain spaces. For example: ... Cory Donnelly
07:52 AM Feature #337: productivity rewards
Task will already nag; "Is that all ya got ?" (or whatever you chooses to prod yourself with) but in the same spirit, task could be taught to let you give yourself rewards for certain arbitrary levels of productivity. Let's say you decide that ... David Patrick

12/03/2009

09:23 PM Bug #315: Unexpected behaviour - marking task as done releases task ID
I swear I've seen this and I don't use the shadow feature at all. IIRC, the usage pattern may have been something like add several new tasks then mark one or more of those done all w/o any reports in between. I would have reported it at the time... John Florian
09:19 PM Feature #328: GNU readline support for shell
I agree that this would be immensely valuable to the CLI. As is, I don't use the current shell at all because the readline support in bash outweighs the hassles of all the escaping and quoting. Oddly, I never think to use my python wrapper tha... John Florian
11:46 AM Bug #160 (Closed): shrink name display
David Patrick
11:23 AM Feature #326: add command; log
Cory Donnelly wrote: > it would be valuable to be able to specify _when_ in fact the logged task was completed. In addition to a due: attr, do we need (or already have in the works) a done: attr? Have a look into pending.dat (in ~/.task/) and you... David Patrick
10:43 AM Feature #326: add command; log
Log it is! David alluded to it when he wrote, "use of dates and date references bears further discussion" -- it would be valuable to be able to specify _when_ in fact the logged task was completed. In addition to a due: attr, do we need (or alrea... Cory Donnelly

12/02/2009

01:38 PM Bug #317 (Assigned): Output produced by `task completed` shouldn't use due-related colouring
cheekily re-opened David Patrick
01:10 PM Bug #317: Output produced by `task completed` shouldn't use due-related colouring
status:completed tasks are currently almost never seen, as (I think all but "completed") reports filter them out, but being able to view those tasks, in mixed company with status:pending tasks, would actually be very useful in the analysis of proj... David Patrick
12:21 PM Feature #335: tab completion should complete projects, tags, etc., found in completed.data as wel...
I think you're right Corey, as we are creatures of habit. When the tag completion gets cluttered with actions, tags and projects we no longer care about, then it's time to purge. David Patrick
12:16 PM Feature #28: Support journalling of entries
this function should be renamed, or hidden, to avoid confusion with "task journal" David Patrick
11:56 AM Feature #336: new command; prepend:
analogous to append: , pre:<string> will prepend the string to the description string. This is desirable for symmetry and needed for some actions, like that outlined in the "enhanced shall" ticket #210 task shell add due:07/21/2010 proj:campi... David Patrick
11:29 AM Bug #255 (Rejected): create status matrix
status use is being streamlined, matrix no longer needed as behavior will be non-confusing. David Patrick
11:27 AM Bug #311 (Rejected): tab completion knows about 'waiting', not 'wait'
status:waiting is going away David Patrick
11:14 AM Feature #335: tab completion should complete projects, tags, etc., found in completed.data as wel...
Often, a project continues after all tasks found within that project have been marked as completed. This can be a periodically-recurring project: ... Cory Donnelly
11:01 AM Bug #315: Unexpected behaviour - marking task as done releases task ID
I've uncovered the root cause of this issue -- enabling the shadow feature via .taskrc effectively runs the ls/list (depending on configuration) on the users' behalf immediately after a `task done` is executed. This updates the IDs of the affected... Cory Donnelly
07:57 AM Bug #96: .css change; ChowFun for h1
At the moment, our stylized font is pasted in as images. Can we convert chowfun (our "chop sockey" fun font) into a font that can be served up by mantis to any <h1> (and maybe even <h2>) heading ? David Patrick
07:23 AM Bug #132: Task sorting
I don't really understand this one.. task sorting in reports ? that already lives withing each report definition.. ? .. or this bug should read; add "started" to sorting keys . ? David Patrick
05:23 AM Bug #332: output field "recur_ind" not valid?
yup, running the full form of the name works on my side. hooray! I'll update to 1.8.5 when i see the announcement. As always, thanks for the info. - Charles t charles yun
12:29 AM Feature #334: new attribute; note:< fileref | string >
Sometimes, you will have more information than you want to cram into a task description. With note: you will be able to add any amount of additional information, from a single word to the largest text file imaginable, by assigning a file referen... David Patrick
12:17 AM Bug #293 (Assigned): tag usage inconsistency
Paul Beckingham
12:10 AM Bug #319 (Closed): Removing tag from multiple tasks at once causes incorrect change summary
There was a logic error in the "diff" results analysis. Paul Beckingham

12/01/2009

11:55 PM Bug #317 (Closed): Output produced by `task completed` shouldn't use due-related colouring
This is already fixed in both 1.9.0 and 1.8.5. The colorization rules for due and overdue are suppressed if the status is completed or deleted. Regarding the new color config (which should be a new "feature" issue), the use of "-Z" is ambiguous ... Paul Beckingham
11:47 PM Bug #315: Unexpected behaviour - marking task as done releases task ID
Cannot replicate. Can anyone else? Paul Beckingham
10:11 PM Bug #332 (Closed): output field "recur_ind" not valid?
It's worse than I thought. The name "recur_ind" is wrong - the real name is "recurrence_indicator", and is implemented, but not documented. Similarly, "tag_indicator" is implemented but not documented. Charles, consider upgrading to 1.8.4, or... Paul Beckingham
04:44 PM Bug #333 (Closed): duplicate command should display the ID of the created task
Fixed for upcoming 1.8.5 /Fredde ... Federico Hernandez
11:22 AM Bug #333: duplicate command should display the ID of the created task
agreed David Patrick
11:00 AM Bug #333: duplicate command should display the ID of the created task
Returning the id of the source task is nice, but it would be more helpful if the newly-created task was also displayed: ... Cory Donnelly
09:36 AM Feature #331: New 'cancel' command
yup, i was thinking of the discussion you identified and the done/defer/delete discussion from the same period, iirc. - tcy t charles yun
09:32 AM Bug #317: Output produced by `task completed` shouldn't use due-related colouring
new color parameter; color.completed= ? down-the-road, visibility of completed items within normal reports (exclusions over-ride) will be toggle-able, in interactive ui, using the (Z)show-all key. cli example; $task -Z list foo David Patrick
09:29 AM Feature #331 (Rejected): New 'cancel' command
waiting for wait:<string> David Patrick
08:35 AM Feature #331: New 'cancel' command
I believe the discussion keyword you're looking for is "retired". This is a contentious one, and there seem to be two solutions; one involving use of the proposed wait:<string>, where the string in this case might be "skipped", and one involving... David Patrick
04:37 AM Feature #331: New 'cancel' command
this sounds vaguely like a discussion i started earlier, but can not find a link to it. not sure if it was pre-this-website (thus only via mail). i mention only because i think there was some good discussion. i'll dig a bit further to see if i ... t charles yun

11/30/2009

06:41 PM Feature #328: GNU readline support for shell
Not a plug-in. If this were an external program, then yes, a plug-in, but this is a library. It is linked in, like ncurses. Paul Beckingham
09:50 AM Feature #328: GNU readline support for shell
An optional feature, that requires an external program, sounds like a plug-in. Perhaps the enhanced shell plug-in will be our first. David Patrick
08:31 AM Bug #332 (Assigned): output field "recur_ind" not valid?
Charles, the 'recur_ind' column is missing from a critical place in the code. Good catch. Thank you. Paul Beckingham
05:54 AM Bug #332: output field "recur_ind" not valid?
howdy, messing around with some stuff and decided to try "recur_ind" and failed. task help indicates that recur_ind should be a viable field. But a stripped down output filter tells me: Unrecognized column name: recur_ind the filter: ... t charles yun
05:17 AM Feature #331: New 'cancel' command
tasks can either be added as a single instance or via recurrances. To keep a history of entered, but not completed/done/performed tasks a new new command could remove the corresponding entry of the task (in case of recurrances the single incarn... Federico Hernandez

11/29/2009

10:39 PM Feature #330: new color parameter; invert
in .taskrc color.tag.foo=invert to trade the current background and foreground colors this is very effective hilighting, and yet doesn't necessarily break the theme. David Patrick
10:18 PM Feature #326: add command; log
Paul Beckingham wrote: > Here are some more wacky alternatives in the hope of finding better: > > retro > already > finished > past > over > was hmmm.. they all suck worse than "did" .. and the winner is... "log" ? (of course, I'll be making an ... David Patrick
09:55 PM Feature #326: add command; log
I agree that 'log' and 'did' are the best yet, but I still don't like either. I lean towards 'log'. We've been looking at a past tense version of the verb "do". Here are some more wacky alternatives in the hope of finding better: retro al... Paul Beckingham
09:15 PM Feature #326: add command; log
it's starting to look like a race between "log" and "did".. good thing for command aliases ! David Patrick
07:25 PM Feature #326: add command; log
_completed_ is already used as a report. Federico Hernandez
07:19 PM Support #329 (Closed): How to list non-recurring tasks only?
Yeah, that's what I figured once I got it working. John Florian
07:18 PM Feature #326: add command; log
Federico Hernandez wrote: > accomplished, achieved, executed, finished, ... .. completed ? David Patrick
07:13 PM Feature #326: add command; log
_did_ is also nice and and matches _done_ quite well. Though it might be too close and could call for missunderstandings. What about some synonym for _do_ in its past tense: accomplished, achieved, executed, finished, ... Federico Hernandez
07:11 PM Feature #326: add command; log
> I like 'log' too, but the journal/logging overlap is a bit of a problem, yeah. Not necessarily a conflict; journal will be a linear, personal document that includes the days task-related actions, but will also allow free-form entries, it's lik... David Patrick
06:47 PM Support #329: How to list non-recurring tasks only?
Good point about the colon, but as it is coded now, the colon is what makes this look like an attribute versus just a search keyword. This might be one for the "when we are using a real parser" pile. Paul Beckingham
06:46 PM Support #329: How to list non-recurring tasks only?
Sorry, I got that completely backwards. Paul Beckingham
06:41 PM Support #329: How to list non-recurring tasks only?
Both of those commands show only the recurring tasks. I want just the opposite, which with the hint I was able to find that these work: ... John Florian
06:40 PM Feature #326: add command; log
I like 'log' too, but the journal/logging overlap is a bit of a problem, yeah. How about 'did?' as in: ... Cory Donnelly
06:17 PM Support #329: How to list non-recurring tasks only?
One or all of the following should work: ... Paul Beckingham
06:11 PM Support #329: How to list non-recurring tasks only?
Is there a way to do this? It would be useful when reviewing hundreds of tasks to filter out those that are recurring as I know those need to be there. John Florian
03:24 PM Feature #326: add command; log
"log" is a good name for this - I was thinking of "record". The problems are: - There is a proposed "journal" feature coming up in the future - There is logging in 2.0 Both of these overlap with the "log" command somewhat. David, s... Paul Beckingham
02:47 PM Feature #326: add command; log
My suggestion would be _log_ as it mimics a log-book function. $ task log Added more diskspace to server FOO. And then it could be aliased by people still wanting to use _x_ /F Federico Hernandez
09:45 AM Bug #322 (Closed): rc: override for shell command does not propagate
Paul Beckingham
09:25 AM Feature #328: GNU readline support for shell
The task shell would greatly benefit from integrated readline support. This would provide command line editing, history, searching and completion. This would need to be an optional feature, thus not adding readline to the list ofdependencies. Paul Beckingham
07:39 AM Feature #326: add command; log
So now all it needs is a better name than "x". Paul Beckingham

11/28/2009

12:20 PM Feature #307: taskrc.vim
David, Please give the attached files a try. You should be able to just extract them in your home directory. The highlighting generally tries to make things like "foo=bar" easy to read. If foo is a known to be good, it gets highlighted dif... John Florian
09:54 AM Bug #327 (Closed): Deleting due date on recurring task wraps to 1969
Paul Beckingham
09:17 AM Bug #327: Deleting due date on recurring task wraps to 1969
Simpler case: ... Paul Beckingham
09:12 AM Bug #327: Deleting due date on recurring task wraps to 1969
By running the following commands: ... Paul Beckingham

11/27/2009

03:11 PM Feature #162: email plug-in; alpine or sup
While both ncurses email clients, Alpine is mature and feature complete, and bigger, and sup is new, light and modeled after gmail with a search and tag metaphor and a thread-centric interface. http://www.washington.edu/alpine/ http://sup.ruby... David Patrick
12:00 AM Bug #95 (Closed): css changes; header
David Patrick

11/26/2009

11:51 PM Bug #109 (Closed): use log-in ID, not name, for issue listings
duplicate and done David Patrick
08:42 PM Feature #326: add command; log
I believe task can already do this, although I forget how. wasn't this one of my earliest requests ? that you (Paul) implemented already ? David Patrick
08:27 PM Bug #160: shrink name display
Jin-young Heo wrote: > The field name is defined in lang/en.yml file. So result of required changing will be applied whole our site. For example, in a history of wiki page, the "Author" filed is going to be shown "By". :-) Ahh well if it affects ... David Patrick
08:25 PM Feature #326: add command; log
$task log [attrs][tags] description To create a task and mark it "completed" in one step. use of dates and date references bears further discussion Cory Donnelly
08:21 PM Support #323: Document use of rlwrap
great idea. In experimentation, I was unable to get the -p colors to behave (most likely my misunderstanding of ANSI color syntax) I tried the -r (remember) flag, it didn't seem to do anything. Somewhat disturbing is when the listed output is writ... David Patrick
08:08 PM Bug #160: shrink name display
sure!! But one more thing. The field name is defined in lang/en.yml file. So result of required changing will be applied whole our site. For example, in a history of wiki page, the "Author" filed is going to be shown "By". :-) Jin-young Heo
07:53 PM Bug #160: shrink name display
Jin-young Heo wrote: > Doest it look good? :-) It DOES look good ! Now, to fully recover the screen real-estate, is it possible to change the headings AUTHOR ASSIGNED TO becomes BY TO ? this will give us maximum screen real-estate fo... David Patrick
07:34 PM Bug #160 (Assigned): shrink name display
OK. I modified code to show a combination of first character like DP, JH. Doest it look good? :-) Jin-young Heo
06:00 PM Bug #113 (Closed): css: re-order main menu tabs; alphabetical
done David Patrick
05:59 PM Bug #180 (Closed): remove redmine references to "other projects"
done David Patrick
04:30 PM Feature #325: add "task duration" parameter
We like trivial ! go with that ! It may have interesting ramifications when applied to repeating tasks, though.. djp David Patrick
04:27 PM Bug #160: shrink name display
yup! I think the screen real-estate on the ticket listings is too valuable to squander on outrageously long names like; Buckingham. If it can be done, please render them as initials. D.P. or, why not just DP k? David Patrick
04:23 PM Bug #134 (Closed): css trainwreck on resize web page width
looks fine now. the problem went away by its self. David Patrick
11:35 AM Feature #325: add "task duration" parameter
We already use durations for recurrence periods, and they look like this: ... Paul Beckingham
11:20 AM Feature #325: add "task duration" parameter
to allow estimates of how long a task might take. This is a new attribute, and establishing the parameters will take some discussion, as taskwarrior does not (yet) deal in time-of-day. The easiest (and least flexible) option would be to allow onl... David Patrick
09:58 AM Feature #20 (Closed): Obsolete DEVELOPERS file
I think there is enough online documentation, and in-source documentation that we don't need to expend any effort in describing what code is in what file. Paul Beckingham
09:56 AM Feature #324: Consider optional integration of readline
Integrating readline (optionally) would improve the task shell for all users. See http://taskwarrior.org/issues/show/323. We should consider integrating readline, and assess the difficulty and appropriateness. Paul Beckingham
09:53 AM Support #323: Document use of rlwrap
One can use rlwrap, http://utopia.knoware.nl/~hlub/uck/rlwrap/, with task shell to get full readline support. History is saved between sessions. The above suggestion from Juergen Daubert needs to be written up in the wiki. Paul Beckingham
09:47 AM Bug #322: rc: override for shell command does not propagate
This command: ... Paul Beckingham
09:45 AM Bug #321 (Closed): task shell convert all characters to lowercase
Task was converting the command to lower case to test against "quit", when it should have been only for a caseless comparison. Paul Beckingham
08:57 AM Bug #321 (Assigned): task shell convert all characters to lowercase
Paul Beckingham
04:33 AM Bug #321: task shell convert all characters to lowercase
Any uppercase character is converted to lowercase if you add data within the task shell. How to reproduce: ... Juergen Daubert

11/25/2009

11:06 PM Bug #160: shrink name display
Is this issue still valid? If it is, what about one of below styles? 1. David Patrick => David P. Jin-young Heo => Jin-young H. 2. David Patrick => D. Patrick Jin-young Heo => J. Heo 3. David Patrick => D. P. Jin-young Heo ... Jin-young Heo
10:53 PM Bug #134: css trainwreck on resize web page width
Do you mean that reducing side space of each main menu tab (Overview, Activities ....)? Jin-young Heo
10:51 PM Bug #113: css: re-order main menu tabs; alphabetical
Done. The forum is another name of board. The menu ordering is determined by "menu.push" sequences in ~/public_html/redmine/lib/redmine.rb (from line 108 ~ 150) Jin-young Heo
09:59 PM Bug #113: css: re-order main menu tabs; alphabetical
Re-ordering was completed except "Forum". Forum doesn't follow a common menu rule like other menus. I'm searching where it's generated. Jin-young Heo
09:37 PM Bug #180: remove redmine references to "other projects"
Sorry too late. Anyway the top menu and select box has been removed. !! IMPORTANT !! This modifying is temporal fix for only our source of redmine on taskwarrior.org server. If someone upgrades or changes it with another version (new versio... Jin-young Heo

11/24/2009

11:40 PM Bug #320 (Closed): color reference system needed
David Patrick
10:32 AM Visualizing task development
Take a look at the video that visualizes the development of task 2.0.0 at http://www.youtube.com/watch?v=AqGZNne-q70 Federico Hernandez

11/20/2009

07:35 PM Bug #320: color reference system needed
David, this is going to make your day. The colortest script you are running generates output like: 16: 00/00/00 17: 00/00/5f 18: 00/00/87 19: 00/00/af 20: 00/00/d7 21: 00/00/ff etc That number in the first column (16, 17, 18, 19, 20, ... Paul Beckingham
10:52 AM Bug #320: color reference system needed
On my system (ubuntu) colortest creates such a list, is there any chance that the numbers on the left might correspond to the taskwarrior color numbers ? djp@orange:~ $ colortest ************************** *XTERM 256Color Test Chart ***********... David Patrick

11/19/2009

12:10 AM Bug #320: color reference system needed
although 256 colors are now possible, it seems the color numbering system is unique. We need some method of cross referencing to a known color scheme, where sample tables exist, or we need task to generate its own reference table, something lik... David Patrick

11/18/2009

03:57 PM Bug #319: Removing tag from multiple tasks at once causes incorrect change summary
Experienced in version v1.8.3. Removing an existing tag from multiple tasks causes the 'No changes were made' message to be displayed for each task, even though an action was performed. Here the tags are added: ... Cory Donnelly
02:48 PM Feature #318: timestamp for status changes
Cory Donnelly wrote: > For all statuses, when status is changed, an associated date value should be set (statusdate:"1258403641" perhaps?) I think that with the combination of existing task and history data, and minor enhancements of existing repo... David Patrick
11:45 AM Feature #318: timestamp for status changes
Interestingly, we already have this information in the undo log. How would you like to see it presented? Just on the "info" report? Somewhere else? Paul Beckingham
11:29 AM Feature #318: timestamp for status changes
(was: All statuses should store an associated date value) Currently in v1.8.3, completing a task causes the task to be moved to completed.data, status is set to "completed", and an end value (end:"1258403641" for example) is added. This could ... Cory Donnelly
11:22 AM Bug #317 (Assigned): Output produced by `task completed` shouldn't use due-related colouring
Good point, Cory. If a task is completed, due date is meaningless. Thanks. Paul Beckingham
11:18 AM Bug #317: Output produced by `task completed` shouldn't use due-related colouring
For completed tasks, colouring them according to color.overdue and color.due no longer really makes sense. Since `task completed` is just a regular report specified within .taskrc this might be more trouble than it's worth to address -- I can't... Cory Donnelly
09:51 AM Bug #222 (Assigned): make test data set available
Paul Beckingham
09:51 AM Bug #222: make test data set available
Ah, I see. That makes a lot of sense. A sandbox. Paul Beckingham
09:43 AM Bug #222: make test data set available
I was referring not so much to your battery of software tests, as I was to a sample.data file that was packed with imaginary examples of tasks, using every feature, for a whole imaginary family; something for new users and experimenters to play with. David Patrick
12:44 AM Bug #211: usage of "update" missing from help output
Ah. Yes. When you do either of the following: ... Paul Beckingham
12:20 AM Bug #211: usage of "update" missing from help output
Paul Beckingham wrote: > Please elaborate. in an IRC discussion, if I recall correctly, you described a seemingly undocumented command "update". (existing usage, as opposed to a future, expanded implementation) did I imagine that ? if s... David Patrick

11/17/2009

11:18 PM Bug #222 (Rejected): make test data set available
All unit tests generate their own data immediately prior to running the test, then clean up after. Therefore I don't think this is important. Paul Beckingham
11:16 PM Bug #211: usage of "update" missing from help output
Please elaborate. Paul Beckingham
11:12 PM Feature #239 (Closed): i8 - stats panel
Paul Beckingham
11:11 PM Feature #238 (Closed): i7 - keys panel
No asynchronous message posting. Not needed. Paul Beckingham
11:08 PM Bug #315 (Assigned): Unexpected behaviour - marking task as done releases task ID
You're right, this is not consistent with intended behavior. It is also inconsistent with this simple test: ... Paul Beckingham
10:57 PM Bug #316 (Closed): timesheet report oddly sorted
It happened 20090922. Hmm. Paul Beckingham
10:50 PM Bug #316: timesheet report oddly sorted
The timesheet report is broken - all the group headings appear first, then the data. When did this happen? Paul Beckingham
10:50 PM Bug #315: Unexpected behaviour - marking task as done releases task ID
Consider: ... Cory Donnelly
06:31 PM Task 1.8.4 is released
See the *"task 1.8.4":http://taskwarrior.org/wiki/taskwarrior/V184* page for details. Federico Hernandez

11/16/2009

11:02 PM Bug #312 (Closed): Applying changes to one task affects another
No fix because this is intended functionality. Added warning when modifying a recurring task that all instances of the recurring task may be modified. Further identified all recurring tasks when asking the "Are you sure?" question. Paul Beckingham
10:44 PM Bug #312: Applying changes to one task affects another
After puzzling over this one I realized that the two tasks are related. 78 is the parent task of 79. This can be seen with: task info 78-79 This is therefore functioning as designed - edits to a recurring task are propagated to other i... Paul Beckingham
10:27 PM Bug #208: 'task recurring' not listing all recurring tasks
Cory's sample pending.data file does indeed exhibit the bug. Task 40 in the sample file does not have the parent:xxxx attribute, and as such doesn't qualify as a recurring task. Whether or not the filtering logic for selecting recurring tasks is... Paul Beckingham
10:02 PM Bug #313 (Closed): Edit command fails when data.location includes directories containing spaces
Nice patch. Thank you. Paul Beckingham
09:56 PM Feature #298 (Assigned): Configurable recurring task count
Paul Beckingham
09:55 PM Bug #306 (Assigned): add tags to annotations
Paul Beckingham
09:53 PM Bug #311 (Assigned): tab completion knows about 'waiting', not 'wait'
Paul Beckingham
09:53 PM Bug #314 (Assigned): tab completion knows about 'recurring', not 'recur'
Paul Beckingham
09:52 PM Bug #313 (Assigned): Edit command fails when data.location includes directories containing spaces
Paul Beckingham
09:51 PM Bug #312 (Assigned): Applying changes to one task affects another
Paul Beckingham
01:55 PM New preview movie
It shows change highlighting and multi-user access. You may need to watch on high quality because the text is small. http://www.youtube.com/watch?v=9CyCxvhvFMo Federico Hernandez
11:16 AM Bug #314: tab completion knows about 'recurring', not 'recur'
See Bug #311 for waiting/wait. Cory Donnelly
11:11 AM Bug #313: Edit command fails when data.location includes directories containing spaces
Experienced under v1.8.3 running on cygwin, likely elsewhere as well. Trivial patch attached. Cory Donnelly
08:58 AM Bug #312: Applying changes to one task affects another
Please disregard the inconsistency in the description of task 79 in the above output of `t strangeness 79` -- in testing bug #208 I mangled this output. In the lines beginning 'Task [78|79],' the descriptions of both tasks were reported correctly ... Cory Donnelly

11/15/2009

10:39 PM Bug #312: Applying changes to one task affects another
I experience this issue under v1.8.3. Updating the description (specified by task ID) causes the description to be updated for another task as well. May be related to Bug #208. Most data in the attached database were initially imported from to... Cory Donnelly
10:30 PM Bug #208: 'task recurring' not listing all recurring tasks
I'm currently experiencing this issue under v1.8.3. See the attached database. ... Cory Donnelly
06:23 PM Bug #311: tab completion knows about 'waiting', not 'wait'
Admittedly one is a report name, and the other an attribute, but I think the tab completion might be making an assumption about argument order. Reported by Cory Donnelly. Paul Beckingham
« Previous
Next »
 

Also available in: Atom