I took a break on one of my goals, duration is 1 week, to begin next monday.
Then I did retroratchet because it had to much buffer.
My break appears to have been canceled. Is this a bug?
2 Likes
Retroratchet doesnât treat flat sections created by take a break differently than ânormalâ sections. It does whatever it has to to make you N days away from eep, including destroying your vacation if it has to.
Currently, I think Beeminder views this as WAI but thereâs lots of feature requests around making breaks a first class thing that various parts of Beeminder would respect.
3 Likes
Hey @chrp! If you email support we can reinstate that break for you.
Retroratchet predates take-a-break; the current behaviour used to make more sense, back when you couldnât schedule road changes. Now it causes confusion, so as @drtail says, this is yet another vote for implementing true breaks in the road.
PS: itâs particularly good for us to hear about stuff that surprises people, because thatâs always a sign of a thing that needs more work. (Even when the behaviour is âperfectly rightâ, the affordances and expectations werenât cued up properlyâŚ) Thanks.
2 Likes
@drtall What is WAI? I too am for making breaks a first class thing: I use them to schedules my holidays, which are fixed some time in advance.
@philip Thanks but I think I will leave it as it is. It will force me to build some buffer.
2 Likes
âWorking As Intendedâ which is just to say they already know about the behavior and theyâre treating modifications like feature requests and not bug reports. Maybe the term isnât as widespread as I thought, sorry 
2 Likes
I had a similar issue. I set up several breaks for times when I would be out-of-town. The dates in Beeminder kept moving, and it took a couple of times of resetting my breaks before I connected the problem to using the retroratchet feature.
1 Like
note to self: donât try to plan too far ahead. 
I think this is on the list of likely fixes this year, since the confusion/annoyance factor is rising as more folks start ratcheting their goals and other relatively advanced beeminding.
2 Likes
Just wanted rise this again/add another voice here. I feel pain pretty much each time I need to schedule a break 
. The incompatibility with retroratchet is only one thing.
There is also the fact that I canât really remove or edit the breaks Iâve added, which often leads to the total mess in the schedule when I try to âeditâ things by adding additional overrides on top of previous ones. Also, the way the breaks are represented is not really easy to understand.
5 Likes
I rely on the visual road editor for breaks, which I donât exactly find easy to use but does at least make it possible to modify breaks, and to check that Iâve set things up the way I intended before I pull the trigger. But it doesnât solve the retroratchet problem.
2 Likes
You can actually remove and edit breaks if you get to grips with the manual road editor, or try the visual road editor as @oulfis suggested. I think a lot of people think the road matrix is only modifiable by us, and itâs not true â you can edit it as well, as long as you donât want to make the road easier within the akrasia horizon.
Re: the representation of breaks, what do you mean by that? When I put in a break using the break tool, it says thereâll be a rate of 0 from x-date to y-date â how do you think that could be clearer? Are we relying too much on people understanding graphs? (Because to me, a rate of 0 is obviously a flat line which obviously means no data will need to be added to be on the right side of the line, and also obviously means âa requirement of 0 per time periodâ â and Iâm not much for thinking mathematically. So itâd be good to know where this is failing.)
1 Like
Well, one problem I always have, over and over again, is fencepost errors: Iâm never sure whether the listed begin/end dates also have 0 rate. Iâve gotten burned by this â thinking my break was longer than it turned out to be, because the beginning/end date wasnât actually set to 0.
1 Like
For people who put it in via the take a break function on the stop/pause tab, the flat spot begins on the first date (the from date) and ends on the second date (the to date). That seems logical and as expected to me?
The âupcoming breaksâ wording could be confusing, I guess? E.g. here
" Upcoming changes:
Through 2018-11-09 increase at rate 5
Through 2018-11-17 continue at rate 0
Through 2029-04-18 increase at rate 5"
To me itâs obvious that means that âup to and including 9th November, the rate is 5/time periodâ, âup to and including 17th November, the rate is 0/time period (flat)â and then âup to and including 18th April 2029, the rate is 5/time periodâ. (Time period being whatever youâve set, day/week/month/year.) Would that be clearer wording, maybe (with the system inserting the actual time period chosen not just saying time period)? 
What might also be confusing if you look at the road matrix is that the entries there arenât âa new rate starts from this dateâ. Instead, they mean âthis is the date when the stated rate endsâ. So e.g. âend date 2018-12-10, end value [blank], rate (per time period) 5â means the rate of 5/time period ends on 10th December (and then the new rate starts on the following day).
1 Like
Graphic road editor is my savior. Didnât know I can do that 
Regarding the break representation: I share the @lanthala confusion there.
I think the âFrom: date. To: dateâ format would be less confusing. If you can have a specialized graphic representation of upcoming breaks would be even better
1 Like
If this is true (no reason to doubt it if you say so), I find it very confusing as well. I would expect any intervals to include the defined starting date.
At least in real life, when I hear someone say âNo more smoking from January 1!â I would think that she would smoke her last cigarette no later than December 31. And â20 push-ups daily from tomorrowâ includes doing 20 push-ups tomorrow.
I guess the cleanest solution would be [from : to] with both dates inclusive (so, the next âfromâ equals the last âtoâ + 1), but if only one date is given, I would always assume it to be inclusive.
1 Like
Honestly, I see no reason for an ordinary user to need to look at the road matrix; the âupcoming changesâ should be informative enough, or even the road itself. The road matrix is what specifies the line of your yellow brick road on the graph, so Iâm fairly sure itâd be difficult to change how it works (and, I could be wrong, but I think thereâs little point in doing so since the primary use case is staff fixing your graph, and we understand it!).
If Iâm wrong and there are people using the road matrix who think itâs confusing, then do let us know! There are other easier ways to check what your road is doing, and even edit your road, but if this is something people are using we might have to give it some thought!)
1 Like
Maybe I was getting a bit ahead of myself. I thought this applied to the regular (i.e. on the goal page) pause editor as well. But itâs pretty clear in the Stop/Pause tab of the goal page with the âthroughâ being inclusive.
You can view the road matrix on the goal page as well (under settings), but itâs buried enough that I definitely wouldnât expect it to be the first port of call!