Proposal Discussion Bugs & Issue Tracker

My brother is a code monkey himself. He told me that some bugs are easier to fix than others and sometimes even finding the cause of a bug is a massive and time intensive challenge, let alone finding a way to fix it without breaking even more stuff
A rational approach would lead to the same conclusion but, as we know, rationality is severely frowned upon on the internet!

It is funny, though (and quite baffling nevertheless) that when a patch is released, a major, glaring bug is not fixed, while the notes report plenty of fixes similar to "fixed a bug where a sound wasn't playing correctly when deploying the landing gear of the Hauler during a Thargoid fight near icy rings", but I guess it's just someone who forgot a semicolumn in the code, or something!
 
I really don't understand why "fixing bugs" and "creating new content" need to be mutually exclusive in Elite 🤷‍♂️
It doesn't NEED to be, but it seems to be nevertheless. Otherwise I like bug tracker - it just needs to be used properly. Right now I don't see any bugs being squashed, either from bug tracker or the forum. No bug collecting system helps if nobody works on the actual bug removal.
 
My brother is a code monkey himself. He told me that some bugs are easier to fix than others and sometimes even finding the cause of a bug is a massive and time intensive challenge, let alone finding a way to fix it without breaking even more stuff

There's a fun little sorting effect happening with the "voting" system, as well. Other than critical-severity bugs which rocket to the top, bugs will tend to accumulate more votes the longer they've been around, so things which are near-impossible to fix will end up stuck there forever regardless of their actual importance to players.

If you ignore the bugs reported since U14 as at least potentially able to get a fix in 14.02 or 15, the top 10 most-voted bugs (and all of the five-bar ones) are:
1) Anti-aliasing doesn't work. That's probably one symptom for a huge pile of inter-related graphics issues
2) The game sometimes hangs on exit. Not only is this probably hard to track down (there are some things which seem to guarantee it, but they're such bizarre things that I can't begin to guess why they should) it's also a pretty minor inconvenience in practice with an easy workaround, and doesn't technically affect gameplay at all. And it's the second-highest non-recent bug :) Predates Odyssey but it'd be really hard to argue - despite its votes - that this is what Frontier should drop everything to fix.
3) Joystick buttons limited to 32 per device. This isn't even a bug, it's a feature request. Predates Odyssey.
4) Ship lateral thrusters are weaker in Live than in Legacy. Since it interacts with the rotational correction setting anything could be going on here.
5) Telemetry data exposure. Another feature request. Predates Odyssey.
6) An issue with Navlock and interface in Wings.
7) "Things other than CPU or GPU utilisation can cause low FPS". That's more a statement of fact than a bug; if taken as a proxy for "Live performance is still a lot worse than Legacy" that's still not just one bug.
8) BGS tick rollout is inconsistent. Certainly is, might well require completely rearchitecting how their servers and the client deliver and cache data to fix it, though.
9) Star lighting colour doesn't work like it used to. Another for the "Odyssey's display routines have a big pile of issues" list
10) PWA isn't working properly. This might be yet another for the "display problems" list or maybe not

So that's:
- two feature requests which shouldn't even be there
- three or four for the giant "Odyssey graphics problems" heap which probably aren't going anywhere without yet another Odyssey-scale rewrite of the whole thing
- one absolute triviality that has just accumulated more votes than almost everything else by being extremely old
- one which is probably deeply tied into their server design
- two or three which might actually be fixable

No wonder that total votes is basically completely irrelevant to fix chance.
 
A rational approach would lead to the same conclusion but, as we know, rationality is severely frowned upon on the internet!

It is funny, though (and quite baffling nevertheless) that when a patch is released, a major, glaring bug is not fixed, while the notes report plenty of fixes similar to "fixed a bug where a sound wasn't playing correctly when deploying the landing gear of the Hauler during a Thargoid fight near icy rings", but I guess it's just someone who forgot a semicolumn in the code, or something!

Hah, I remember him cursing that missing semicolumn and comma are actually the WORST, because you just don't notice them when looking for the bug :LOL:

That would actually be an interesting topic for a Dev Q&A in the Elite Livestream: What was the most annoying bug you've fixed (or still haven't fixed)? :D
 
What's that, then?
Expose enough real-time movement data from the game so that motion simulator setups can apply pitch, roll, yaw and simulate impacts without having to either estimate it imperfectly from control inputs and journal events or by {redacted} which breaks every time there's a new client.
 
E.D. Is the only simulator of its type that I am aware of in that it was based on nasa data to 1-to-1 represent our home… that alone, without any thargoid fluff gives ED value to more communities than just gamers. The science community and students could potentially be using ED forever… So when they announced that they would no longer be working on any of the planetary tech regardless of known bugs, I realized they was planning to shift their focus away from ‘simulation’ and focus on making ED more of a ‘game’. Obviously you have to make a galaxy fun to bring in the younger crowd! But the ‘timeless simulation’ aspect of ED should have always been their focus (in my opinion) because a reliable, smooth running (and updated… ex. James Webb, like the original designers did with discovery-1. And continued potential exobiology) Milkyway galaxy can never get old… it is our home, and the only VR experience we have of exploring our home!

Put the ‘game’ on hold and fix the backbone ‘simulator’, if the galaxy is proper, you can have decades of new players for storylines and excitement to play with. I came to ED just to fly through the galaxy. I appreciate the thargoid story and the burning cities are fantastic, but I think it only works because we love our home and the time we have invested in our home galaxy. BGS, having space jobs like mining, trucking, a good ‘trade-market’, the political parties, PVP, pirating, exobiology etc…. Normal galaxy life should be priority fixes overall. Those are the things that make ED our home and worth defending when the side-stories hit. At least that is what ED is for me.
 
Last edited:
Agreed. They fix plenty of bugs - including some with pretty trivial severity, presumably if the fix is also trivial - but the public issue tracker appears to have no meaningful connection to what they fix (which may partly be because they often forget to mark fixed bugs as being fixed there, and even when they do it lacks the "notify the reporter" feature that every other bug tracker in existence has)


Hard to say. For the games that they develop as well as publish, and only putting as much effort into identifying duplicate or invalid reports as Frontier do:

Resolution rate is Fixed+By Design+Invalid+Duplicated / Reported
Fake Fix Rate is assuming that all bugs which are not Confirmed are the collective delusions of only 10 Elite Dangerous players or 3 Other Game players.
GameReportedCurrently ConfirmedFixedBy Design, Invalid or DuplicatedFix rateResolution RateFake Fix Rate
Elite Dangerous (excluding separate Alpha/Beta reports)38649227128850333%16%85%
Planet Coaster (PC/Mac + Console)435 + 35372 + 845 + 4331 + 10811%29%52%
Jurassic World Evolution (1+2)642 + 139015+8057 + 147208 + 43210%42%68%
Planet Zoo8353851636632%10%66%
F1 Manager1469141003927%33%87%
All self-developed games except Elite Dangerous1264227455518344%19%67%
So ... their fix and resolution rate for Elite Dangerous isn't that much less than their fix and resolution rates for the average of their other games, but in absolute terms they've fixed and resolved over twice as many ED bugs as the rest of their games combined. ED is of course older than any of the other games - though not older than the combined age of all their other games - which will skew this a bit. (Is it just more full of bugs than the average Frontier product? Well, again, it's also bigger than the others and I can't see a way to tell if it has more "bugs per square feature" or whatever you want to measure)

On the "fake fix rate", Elite Dangerous is considerably better than almost all their other games (F1 probably benefiting from being new and still in the "early patches" stage) - if a bug does fight through the obstacles to get to Confirmed, it has a good chance of getting fixed. Of course, if they allowed ED bugs to be confirmed with 3 confirmations rather than 10, like the rest of their games, the number of Confirmed bugs would rise substantially and the fake fix rate would probably fall.

On that basis I'd say (and I'm surprised by this first point):
- report bugs to the bug tracker if you have the organisation (or alts) to confirm them without much effort; despite Frontier's best efforts to portray it as pointless Confirmed reports do actually seem to do some good over and above forum complaints
- ignore the voting system [1], the vast majority of those fixed bugs will never have got any significant vote count, it's not necessary

[1] As an aside on this, if you open the issue tracker for all games, then sort the Confirmed bugs by Top Voted, the first non-Elite bug is a Planet Zoo crash bug at 29th (and you have to get to 108th before a third game appears in the list). Elite Dangerous bugs get into the "only 1 bar of votes" stage by about 40th.
Thanks for the very interesting post and your conclusions.

I can't believe there really are ~32000 unfixed individual bugs. It's more like it feels wrong than actual knowledge. But I would guess the game doesn't even have 32000 different "reportable functions" so it can't be correct.

So I guess the system isn't very precise when it comes to duplicates or invalid reports...

Personally I think it would be better to move bug reporting back to the forums, including different sub forums, moderation and direct feedback.
 

rootsrat

Volunteer Moderator
Thanks for all the comments everyone, it's great to see the discussion :) Let's hope FDEV will notice this thread and take the feedback onboard, even if they don't reply (hopefully they will though!).

One thing that would personally encourage me more to log issues is if the BACK button (the one within their webapp, I am not talking about Previous Page browser button) would take me back to the previous page of my report without wiping everything I've put in.

This is by far the biggest functional annoyance for me (systemic ones aside, like voting or 10 votes for ED vs 3 for all the other games). Having worked as a software tester and in IT support in general, I try for my reports to be as detailed as possible, as I know how helpful that can be. So it takes me 15-20, sometimes 30 minutes to fill everything. But quite often after going to the confirmation page and re-reading everything I will realise there is another thing I forgot to mention.

I hit BACK and I am taken to a completely blank form. I cannot count how many times I rage-quit logging the issues because of that, and this is why I honestly just can't be bothered with the issue tracker most of the times nowadays.
 
I can't believe there really are ~32000 unfixed individual bugs. It's more like it feels wrong than actual knowledge. But I would guess the game doesn't even have 32000 different "reportable functions" so it can't be correct.

So I guess the system isn't very precise when it comes to duplicates or invalid reports...
Oh, if Frontier cared to they could massively boost both their fix and resolution rates by paying someone to actually curate the reports, which would move a lot of the "unfixed" reports into the Duplicate/Invalid box where they actually belong (or indeed into Fixed, as they're unmarked duplicates of bugs fixed months or years ago).

There's only about 25-30 new bug reports a day for ED on average, so someone looking at them full time could take about 5 minutes per bug to assess it and either mark it as duplicate, invalid, or plausible enough that it should be "Confirmed" for further investigation (most of them wouldn't take as long as 5 minutes once this person had got a little experience).

Since they don't pay for this - or even do standard bug tracker things like "don't put closed bugs on the default view" - I assume they see "Over 32000 Bugs!" as a selling point.
 
From my point of view, they only fix the easy stuff. Everything else is ignored, or excused away. It seems pretty evident to me they don't care about player retention. All that matters is new content to bring in new players. Just simple, short-term goals.
 
Hah, I remember him cursing that missing semicolumn and comma are actually the WORST, because you just don't notice them when looking for the bug :LOL:
Usually those issues are the easiest to find, as syntax errors stop code from even compiling (and the IDE highlights them).
 
Back
Top Bottom