FDEV - How do you prioritize bug-fixes?

[Northpin]

I felt like they were doing really well with commincation at that point but then it all got derailed by the console dev messaging delay.

Well... the order in which stuff happened is a bit wrong.

First half of the FY2022 ended up with FDev at a loss. Since they're a PLC, their main responsibility is to make money for the shareholders.
And if the costs are bigger than the revenues, usually some of the costs are cut.

So, consoles* dev - cut, infrastructures costs - some cuts (why have more servers dealing with carrier jumps when you can queue the jumps and have a carrier eventually jump in 55 minutes instead of 15) and so on
And based on the lastest announcement, the comms will be toned down (only one stream per week, no more discovery scanner streams, etc)


*actually it's fair to say that Odyssey engine optimizations were Cut, which means Odyssey has no chance to run on Consoles, which means no Console port

 

[nanite2000]

Do you really need to ask?

Because fixing this:
'Fixed some incorrect text in a Sabotage mission.'

Is easier than fixing this:
'Tiling Planetary Features.'

Everyone with a text editor can fix the former while it requiers a PhD to fix the latter. Should they stop fixing small bugs when it only takes a few minutes of work of an intern just because the other problems are more complicated?

But if you look at the list of bug fixes I shared, most of them required a dev to go into the source code and change something more than just a text string.

Take the optimisations for example: they’ve resulted in little to no measurable improvement in performance - so why did they bother? What did it accomplish? Those devs could have been fixing the lighting engine, for example, resulting in immediate and obvious improvement.

like, imagine if FDev said “Sorry - not many bug fixes this quarter. Our team are working on a new antialiasing engine”. The community would go wild.

 

[Northpin]

But if you look at the list of bug fixes I shared, most of them required a dev to go into the source code and change something more than just a text string.

Because they were stilly trying to optimize the engine to increase frames-per-second, which would eventually lead to releasing the game on consoles, as planned initially.

That will happen at a much leisurely pace from now on, if ever.

 

[Screemonster]

• https://issues.frontierstore.net/issue-detail/3018
• https://issues.frontierstore.net/issue-detail/45736
• https://issues.frontierstore.net/issue-detail/45741
• https://issues.frontierstore.net/issue-detail/1763

In order:
1: a non-trivial network issue.
2: a one-off event for which a workaround exists
3: literally fixed in update 10
4: a non-trivial engine issue

 

[babelfisch]

But if you look at the list of bug fixes I shared, most of them required a dev to go into the source code and change something more than just a text string.

Take the optimisations for example: they’ve resulted in little to no measurable improvement in performance - so why did they bother? What did it accomplish? Those devs could have been fixing the lighting engine, for example, resulting in immediate and obvious improvement.

like, imagine if FDev said “Sorry - not many bug fixes this quarter. Our team are working on a new antialiasing engine”. The community would go wild.

How do you know some of the fixes were not applied when they investigated something else?

 

[nanite2000]

How do you know some of the fixes were not applied when they investigated something else?

Maybe they have. That would be a very unfocused way of working and would explain why they are so slow at fixing things.

 

[Erei]

Well, it's an easy, professional and efficient process !

First, the bug report need to go through a popularity contest
Then, if the contest is won, the bug is officially acknowledged as a bug
Following that, the report is put through another popularity contest
After that, if the report win enough votes, the Lords of Fdev will hear the barking of the peasants, and read said report
Then, they'll decide if they fix it, depending on various parameters, ranging from feasibility, time, amount of coffee left in the machine, and if they are not too busy working on another project
Finally, if the report went through all that process, and the Lords decided to fix it, then it will be fixed. And another bug will be created to keep the balance in the universe.

Should the report stay for too long in any of this stage, it will disappear, since, as we know, time heal all.

As you can see, it's most possibly the best way to report bugs ever invented by humanity ! I cannot possibly imagine a better one !

Praise Lord Brebus.

 

[Ian Doncaster]

In order:
1: a non-trivial network issue.
2: a one-off event for which a workaround exists
3: literally fixed in update 10
4: a non-trivial engine issue

It's certainly a pretty good measure of how disconnected that bug tracker is from Frontier's actual development processes that it's only about a 50-50 chance that bugs they (believe they) fix will get marked as fixed in the tracker...

 

[Northpin]

One of the problems i have with bug fixes is a lot more fixes gets in than they get mentioned in the Patch notes. A lot more

The other problem is bug regression (which is way worse than the other one)
They fix one in U6, gets bugged back in U7, gets fixed again in U9

 

[Thatchinho]

Because it's clearly not prioritized based on the issue tracker you keep asking us to use for logging bugs and issues.

For example, below is a selection of some of the bug-fixes that were applied in Update 11:

• Fixed some incorrect text in a Sabotage mission.
• Fixed corpses clipping through the floor.
• Added and adjusted ammo crate positions to improve distribution.
• Missing images when accessing some articles via the Insight Hub were replaced.
• Fixed misaligned text on Fleet Carrier nameplates and decals.
• Fixed scatter rock appearing black on planet surfaces from a distance.
• Fixed the cockpit view rendering over the Galaxy Map.
• Checkerboard rendering is now automatically applied to low, medium, and high quality graphics settings. This option will reduce the cost of the planet terrain shader when in close proximity to a landable planet.
• Text colouring on the ship boarding panel was fixed.
• The layout of the search results in the Galaxy Map when missions exist in bookmarked systems was fixed.
• Text resizing to the item/emote wheel so letter wrap correctly was added.
• Focusing behaviour for the staff list on terminals was fixed.
• Filters at the Pioneer Supplies vendor when buying/selling a weapon on suit were fixed, preventing new items which don’t match the filters being shown.
• An issue on the multicrew report popup that would cause multiple crew members to be selected at once was fixed.
• Fixed cases of shadows flickering from directional lights.
• Ensured planet-related systems do not perform unnecessary work when in space.
• Miscellaneous optimisations to lighting and colour grading.
• Fixed some incorrect text in a Theft mission.
• Fixed an issue where a faction would sometimes target themselves for a Hacking mission.
• Fixed an issue where some missions expire but not correctly fail.
• Fixed Salvage missions failing to have the relevant item for the mission.
• Fixed the spawning of mission givers for the hand-in of missions taken from a settlement.

Why were these given higher priority over the top 20 issues in the issue tracker? They seem pretty ineffectual compared to the top 20 items in the issue tracker, so why were they given higher priority?

I doubt this hasn't been covered already but...

There will be multiple criteria that determine priority. It's never been implied that the rank on the issue tracker was the sole determinant of priority.

Also your fundamental assertion is wrong.

Just because something was fixed first doesn't mean it was given higher priority. The top priority can still take a lot of time and effort to fix. Things that get fixed first can be much lower priorities, they could just be fixed first.

 

[Thatchinho]

Finally, if the report went through all that process, and the Lords decided to fix it, then it will be fixed*.

*Actual level of fixing may be less than implied.

FTFY

 

[Mogen Teras]

*actually it's fair to say that Odyssey engine optimizations were Cut, which means Odyssey has no chance to run on Consoles, which means no Console port

EVen if it is proving to be very difficult for them, I doubt they would cut out optimization; it would seriously restrict the reach of the game if only a small subset of an already niche community is even able to run the new client at all. And after already taking the revenue hit by cutting consoles out, I doubt it's in their best interest to cut even MORE of their potential revenue.

 

[thistle]

Take the optimisations for example: they’ve resulted in little to no measurable improvement in performance - so why did they bother? What did it accomplish? Those devs could have been fixing the lighting engine, for example, resulting in immediate and obvious improvement.

That's not true, my game was very slow on release, unplayable at times in the early patches, and runs perfectly fine now.

And they did put time into the lighting, fixing for example the incorrect and glaringly wrong atmospheric light. Have they fixed all the lighting problems? Clearly not.

like, imagine if FDev said “Sorry - not many bug fixes this quarter. Our team are working on a new antialiasing engine”. The community would go wild.

Some would go wild with excitement, some would go wild with apoplectic rage. Just the usual.