The Feedback Points System

Bugs and Feature Requests
#1

As you may know, in Feedback you can indicate which cases are the five that are most important to you. This then adds points to those cases and the total of those points from all users interested in the case helps inform us as to the relative level of importance of that case.

There could be other ways to do this. We could simply have a like button. Rather than being limited to five cases, you’d simply like cases that are important to you. The upside is that this is simply and obvious. The downside is that users would need to understand that with every like they are lessening the impact of all their other likes. I’m not sure that would be obvious. Liking everything is the same as liking nothing.

I would like it to be simple and obvious and perhaps a like button accomplishes that and if there’s a downside, it’s worth the simplicity.

Any thoughts on this? Do you like the current top 5 system? What do you think about replacing that with a Like button?

1 Like
#2

But hopefully that doesn’t mean that the points set so far expire and the feedback cases slip back in the ranking, does it?

#3

I’d probably go with something a little different. When viewing a case, have an “assign points” button. When you click that, you’d get a window that shows where your current points are assigned, how many you have remaining, how many you want to assign to the current case, etc. That way if I wanted to assign all my points to one case, I could. Or I could assign 1 point to 30 cases. Let me remove assigned points from cases in that window too, so I can do it all from one place. (edit: when something is fixed, removed, whatever, your points become available automatically)

But even if you only have a “like” button, having it more obvious than the current way of My Top Cases is an improvement.

6 Likes
#4

@Geoff_Perlman The most important thing is:

  • give a clear overview of the “highest priority cases” from the Xojo side.
    – It’s impossible to know where xojo is headed, the roadmap is not clear and feedback is not showing anything about what xojo is working on. Having a way to see what’s being worked on would give customers a better overview of “xojo supporting it’s customers”.
  • give a clear overview of the “highest priority cases” from ALL customers together.
    – The current feedback ranking is almost invisible, it’s very hard to find the correct view for all rank listings.

A like button system is not a good idea i think. Personally i’d give customers an option to have a top 100 listing instead of top 5. That way you can better see what’s the most value and lesser value for each customer. We can have a good list (for our self) that shows what we want.

The top 5 is a thing that i update once, and keep for as long as it’s implemented or fixed, now given Xojo takes a LONG time to fix (my) issues or implement (my requested) features, a TOP 5 is way to few options.

3 Likes
#5

It may be more important to highlight with a new release which bugs you picked to be fixed due to the points. So we see that the points actually have an effect.

My top item is case 59997, as it may be a bug affecting my users. How can we be sure that such cases get attention?

3 Likes
#6

@Geoff_Perlman, thanks for seeking to address one of the weaknesses of the Feedback app.

I like @Bill_Gookin’s idea of every user having a certain number of points. How many you get could be determined by license type or some other metric. If the new web-based Feedback app was able to let you assign a variable amount of points to an issue that would be great. It would also be super useful if when an issue is verified, the points allocated are released back into a users “pool” automatically as I find at the moment that I often forget to change my top cases.

May I make a couple of other suggestions about how either the existing Feedback app (or the in-development web-based app) could be improved?

  • Add better formatting options to the case. It would be nice when reading current issues if we had syntax highlighting for example. Perhaps add Markdown support when writing issues? You are welcome to use my open source Markdown parser (written in Xojo) if that is a barrier. Markdown has proven popular on this forum and your users are familiar with it.
  • If a bug has been verified by Xojo to be reproducible (and therefore needs fixing) perhaps a weighting or special points should be added to it so that it bubbles up the priority task list? Maybe add ten points for every three months it goes unfixed for example? This would mean that old verified bugs will eventually move to the top of the list to fix and not be forgotten about.
1 Like
#7

I don’t think any point or ranking system is needed. There’s a few problems with trying to rank cases using any system.

  1. For features, the major cases will always bubble to the top just like they have now. That won’t change. They also don’t help Xojo prioritize, because those major feature requests require such significant resources that Xojo must plan very carefully when to tackle each. And of course, they’re already on Xojo’s radar. Android support, preemptive threads, arm support, command line compile, and so on.
  2. For bugs, Xojo’s engineers are infinitely better at determining case importance. A saving bug with no points is dramatically more important than a feature request for popovers for example.

So what purpose will points really serve? I’d rather see Xojo focus on getting cases consistently reviewed and acted upon, than ranking them. I know from personal experience just how daunting of a job that is. That’s why I suggested in another thread Xojo keep a metric of case status age. How many cases have been in “needs review” for x days? y days? z days? Kind of like an accounts payable view. This doesn’t necessarily need to be public information, but I think it’s a better focus of effort than trying to rank cases.

11 Likes
#8

I kind of agree with Thom there. The points are cool and all but it really is more like eye candy. I do believe that any user should be allowed to vote up on an issue if it affects them (so the “impact” of the issue is better understood), but that is where the ranking should end.

I think the key thing is the feedback system itself is broke. That one had to install a new feedback every single time the IDE changed (or almost every time) was such a non-starter - I mean it is supposed to be a feedback app, not a life-support system - how come it has to change all the time. And that many Linux users are disenfranchised if they choose to only use Linux because the app is buggy in Linux seems like counter-productive - how can their issues be counted if the system won’t even allow them to use it. That is where a web-based system would rule. If the feedback system is meant to better the product at the expense of the users, then it should be dead-simple to use. If the system is just meant to serve as eye candy then any system will suffice - heck the new feedback forum section is good enough for that.

Maybe a few conversations with the @Susan_Fennema lady could help in reshaping the feedback system (not just the points, but the whole system).

#9

just 5 entrys is not enough, i like to order all my open feature requests.
largely all are important for me or others.
i not like to search/find/vote other peoples entries.
.

#10

Hi @LangueR - I’d be happy to help! But, I’m not sure how much flexibility the system has - what’s been customized vs. what is out of the box. That would for sure be an @Dana_Brown question.

#11

Or… in reading up the thread, an @Geoff_Perlman question. (sorry for the oversight there!)

#12

I don’t like replacing it with a “Like” button.

I like Bill Gookin’s idea of being able to spread a pool of points to cases.

I also like GarryPettet’s idea that verified bugs get points added as they age, forcing them up the list.

One of my gripes with the feedback system was that it did not separate features from bugs, as Thom pointed out feature requests always bubble to the top. At one point there were only three bugs listed in the top 20. I always felt we should have two lists, bugs and features with points to spend on both. I know there is other adjustments made to the ranking internally but from our point of view it looked like bugs took a back seat to feature requests.

#13
  • Allow Like/Points for Feature Requests only.
  • Weight Reported Bugs on how often this/similar Bugs have been reported. (I know this is not as easy as it sounds…)
#14

We already do that in that multiple users can put the same report on their top 5 list.

#15

I know age seems like a good idea as a criterion but I’m not convinced it is. Often we find old bugs have been fixed as a side effect of some other change. For example, it’s unlikely that any bug reported against the old Web Framework exists in the new one.

As for feature requests versus bug reports, imagine a user that has no bugs that are bothering them but they want 5 features so they put those on their top 5 list. That’s an accurate reflection of their interest. To take an extreme example, imagine that no one asked for a single feature request but only put bugs on their top 5 list. Under those conditions, only fixing bugs would clearly be the mandate. Having both on one list is a way for us to be able to gauge the relative importance of all cases.

#16

Points don’t expire. They of course don’t count once a case is closed but you can then remove that case from your list and put a different one on the list.

Perhaps we just need to make it easier and more obvious as to how to choose a case for your Top 5. Imagine you are looking at a specific case and there was an easy way to see which chases you already have on the list and add this one to the list without having to leave the screen. That might make it more likely to be used.

#17

Right, but in that case why would you want to leave the bug report open anyway?

2 Likes
#18

imagine a user that has no bugs that are bothering them but they want 5 features so they put those on their top 5 list. That’s an accurate reflection of their interest.

Ok, in the combined list they get to spend their points on 5 features. Two months later they find a bug. Now they have to decide which feature to not vote for in order to add a point to the bug. Did one of the features become less desirable? No. It also has the affect of showing feature as less desirable because points were removed. In a spit list they simply spend their bug points and don’t affect the features list.

#19

I also think 5 items is way too low; I thought it since the beginning of Feedback.
There are so much features in Xojo and you only have to pick 5 improvements…
Not sure how much I’d think is ideal though. At least 10, definitively not more than 50.

My opinion.

#20

In that case they should be marked as fixed. That is not always the case. It would certainly help with the optics of the situation, after all, the team have done the hard work and squashed it. It seems like your QA guys/girls could do with some sort of rolling audit system.

1 Like