Rabidfox's wall of text- old bug reports

Jaspar

Well-known member
I'll be sure to start new threads for future bugs. That initial wall of text was just things I mass pulled from bug reports I had submitted over the years, but I could also break those up into individual posts too if it would be more helpful.

Its been talked about a bit but I dont know if bug reported yet - mobs casting scorching ray and frost ray are doing easily 200% more damage thsn they should.

Im assuming mobs got the damage boost to those spells which we got - by accident.
 

The Blonde

Catalogues Bugs
355473 Celestial warlock's lvl 15 ability Ameliorate doesn't apply any effect on non-undead. The description says that undead that aren't killed by initial effect and non-undead will take 10 MRR debuff for 15 seconds. Undead take this debuff correctly, but non-undead mobs don't get debuffed by the ability.
More detailed/expanded bug report on Ameliorate:

Ameliorate (Celestial Warlock Lv15 ability) has the following bugs:
- Typo: "Undead who fail this save.." should instead be "Undead who succeed on their save.."
- Non-undead enemies are not affected by the MRR reduction portion of the ability.
- Red named/Purple named undead that suceed on their save are not affected by the MRR reduction portion of the ability.
- It doesn't show up as a debuff on enemies that have had their PRR/MRR reduced (The only enemies that are correctly being affected by the MRR reduction portion of the ability are non-boss undead who make their save, but this MRR reduction is not indicated by a debuff on them).
 

seph1roth5

Well-known member
This has been fixed internally, and will be in an upcoming patch. Thank you!

As a note, if you have unrelated bugs to share on the forums, I do definitely prefer one new thread for bug or category of closely-related bugs instead of a big thread with many disparate bugs in replies. It just makes things a lot easier for me :) Thank you!
Wouldn't it be easier if someone (rabidfox) did a megathread and just kept it updated in the first post? We'd let him out to eat a couple times a day, maybe use the bathroom.
 

Igognito

Well-known member
So! This is indeed the same bug as Magic Fang, and in the short term, is getting the same band-aid - a max caster level of 20. However, behind the scenes, the real bug is still looming... unfixed... and ready to strike.

The real bug is hidden deep within the Progressions system within the game. A Progression is a piece of data in DDO that associates a level value with a result. In the case of these misbehaving spells, the Progression associated with the duration of the spell is what's breaking. What's happening is that the association data for "Spell Duration, 5 Minutes Plus 1 Minute Per Level" only scales to level 40. Any level beyond that, and the progression is returning null, which is a fancy word for "no data" and in a practical sense means that the spell computes its duration out to 0. This causes the effect to fail to apply correctly.

So, in the long term, we're looking at the work to fix up the Progression system to return the highest value when the index goes out of bounds on the upper end. We still want the lower bounds to remain null (after all, we want to be warned if someone is asking for a negative caster level, that seems very bad!) but we think it might be safe to clamp the index of the top end without causing too much disruption. The tricky part here is that Progressions are the backbone of a bunch of unrelated and hard-to-test systems... like, for example, all of Random Loot. This trickiness is why we're opting for a band-aid now and a comprehensive fix a little while later, once we're sure we have the bandwidth to do some regressions on the larger systems. We're reasonably certain that this fix is going to fix a ton of seemingly-unrelated minor bugs, but it's better safe than sorry when it comes to something so comprehensive.
A question of pure Curiosity: DDO for progression keeps a pre-computed structure with the data of each possible level?
So when you happen to ask level x progression out of the pre/computed ones you get a null value on that data?
When you mention you band-aid a Max level, I assume you just modified the index to fall within bounds (lvl 20 in this case).
I'm assuming that the bounds of each progression are different and you have no cheap/efficient way to know the bounds of a specific progression which could provide you a general solution.
You could though, create a batch job that would probe all the progressions and duplicate each current max to a fixed high level let's say 2^31 (assuming 32bit index available) and default to that index when null occurs... I'm obviously making many assumptions here, mainly that your progression table is an O(1) retrieval structure that won't also require values for lvl X until default.
Another solution would be to have a second progression structure with the default values. When you get null with a positive index, you jump to the default values.
Fun little hassle you got here 😅
 

Redgob

Well-known member
The real bug is hidden deep within the Progressions system within the game. A Progression is a piece of data in DDO that associates a level value with a result. In the case of these misbehaving spells, the Progression associated with the duration of the spell is what's breaking. What's happening is that the association data for "Spell Duration, 5 Minutes Plus 1 Minute Per Level" only scales to level 40. Any level beyond that, and the progression is returning null, which is a fancy word for "no data" and in a practical sense means that the spell computes its duration out to 0. This causes the effect to fail to apply correctly.
But why not simply extend all of the progression arrays(or lists or whatever you use) to something like level 60 with just a repeat of level 40 values? I agree that's still a bandaid but if level 40 is going to be last... This can probably be automated too. Altering the systems at unknown point in the future might not be that much better.
 

Igognito

Well-known member
But why not simply extend all of the progression arrays(or lists or whatever you use) to something like level 60 with just a repeat of level 40 values? I agree that's still a bandaid but if level 40 is going to be last... This can probably be automated too. Altering the systems at unknown point in the future might not be that much better.
My best assumption is Memory consumption. Also it doesn't appear to be just an array (certainly not a list that would have an expensive traverse and could just have a pointer to the last element to ensure you don't go out of bounds) it is either a dictionary/hash or something else. It is a fun little riddle for us too. To figure out what is there 😅.
The max progression of things is not as fixed as the character level. For example, how about cannith crafting progression. How about monster progression? That is certainly not bounded by level 40.
Again we are making a lot of assumptions. Which is fun 😅.
 

PaleFox

Well-known member
So! This is indeed the same bug as Magic Fang, and in the short term, is getting the same band-aid - a max caster level of 20. However, behind the scenes, the real bug is still looming... unfixed... and ready to strike.

The real bug is hidden deep within the Progressions system within the game. A Progression is a piece of data in DDO that associates a level value with a result. In the case of these misbehaving spells, the Progression associated with the duration of the spell is what's breaking. What's happening is that the association data for "Spell Duration, 5 Minutes Plus 1 Minute Per Level" only scales to level 40. Any level beyond that, and the progression is returning null, which is a fancy word for "no data" and in a practical sense means that the spell computes its duration out to 0. This causes the effect to fail to apply correctly.

So, in the long term, we're looking at the work to fix up the Progression system to return the highest value when the index goes out of bounds on the upper end. We still want the lower bounds to remain null (after all, we want to be warned if someone is asking for a negative caster level, that seems very bad!) but we think it might be safe to clamp the index of the top end without causing too much disruption. The tricky part here is that Progressions are the backbone of a bunch of unrelated and hard-to-test systems... like, for example, all of Random Loot. This trickiness is why we're opting for a band-aid now and a comprehensive fix a little while later, once we're sure we have the bandwidth to do some regressions on the larger systems. We're reasonably certain that this fix is going to fix a ton of seemingly-unrelated minor bugs, but it's better safe than sorry when it comes to something so comprehensive.

Why not just set the max progression to 60? Once we're at level 40 there prob will be ways to get MCL and CL way higher than the level the character is at.
 

Tonquin

of Lightning Hammer fame!
More detailed/expanded bug report on Ameliorate:

Ameliorate (Celestial Warlock Lv15 ability) has the following bugs:
- Typo: "Undead who fail this save.." should instead be "Undead who succeed on their save.."
- Non-undead enemies are not affected by the MRR reduction portion of the ability.
- Red named/Purple named undead that suceed on their save are not affected by the MRR reduction portion of the ability.
- It doesn't show up as a debuff on enemies that have had their PRR/MRR reduced (The only enemies that are correctly being affected by the MRR reduction portion of the ability are non-boss undead who make their save, but this MRR reduction is not indicated by a debuff on them).
This has been fixed internally, and will be in an upcoming patch. Thank you!
 

Nimrud

Well-known member
Hey, @Tonquin, could you please take a look at 2 item bugs:
- Legendary Necronomicannon (Fey runearm) has Epic Trap The Soul instead of Legendary version (when you equip the item, the icon for the Epic version shows up in your buff bar) - and it doesn't work at all in legendary content.
- Legendary Solar Gem of Threat Decrease (+20% Artifact Bonus to Threat Reduction) doesn't work for ranged damage, it only reduces melee threat.
 

Arkat

Founder & Super Hero
- Legendary Solar Gem of Threat Decrease (+20% Artifact Bonus to Threat Reduction) doesn't work for ranged damage, it only reduces melee threat.
Agreed, and there is no separate Ranged Threat Reduction Solar Gems (or Lunar Gems for that matter).
 

Tonquin

of Lightning Hammer fame!
Hey, @Tonquin, could you please take a look at 2 item bugs:
- Legendary Necronomicannon (Fey runearm) has Epic Trap The Soul instead of Legendary version (when you equip the item, the icon for the Epic version shows up in your buff bar) - and it doesn't work at all in legendary content.
- Legendary Solar Gem of Threat Decrease (+20% Artifact Bonus to Threat Reduction) doesn't work for ranged damage, it only reduces melee threat.
This is going to sound crazy but I'm getting the sense that if I keep replying in this thread players will keep posting here instead of making new threads for their bugs, so if you would do me the silly favor of just starting a new thread, we can pick these up there :p sorry for the strange request but having one big thread full of bugs is much harder for me to work with than new threads for each bug or small related list of bugs. As always, the best way to get bugs to the dev team is to report them to us with the bug report tool!
 

droid327

Hardcore casual soloist
This is going to sound crazy but I'm getting the sense that if I keep replying in this thread players will keep posting here instead of making new threads for their bugs, so if you would do me the silly favor of just starting a new thread, we can pick these up there :p sorry for the strange request but having one big thread full of bugs is much harder for me to work with than new threads for each bug or small related list of bugs. As always, the best way to get bugs to the dev team is to report them to us with the bug report tool!
I think everyone believes you at your word... But then why is it that bugs seem to get squashed most efficiently from mega threads, somewhat efficiently from individual threads, and submitting a bug report feels like putting a message in a bottle?

If the bug report tool is really the preferred and best method, why don't we see more bugs from it on patch notes? Most of the big ones any given patch can be traced back to squeaky wheels on the forums or maybe discord
 

The Blonde

Catalogues Bugs
I think everyone believes you at your word... But then why is it that bugs seem to get squashed most efficiently from mega threads, somewhat efficiently from individual threads, and submitting a bug report feels like putting a message in a bottle?

If the bug report tool is really the preferred and best method, why don't we see more bugs from it on patch notes? Most of the big ones any given patch can be traced back to squeaky wheels on the forums or maybe discord
they-hated-jesus-because-he-told-them-the-truth-meme-template-regular-d4f6ceea.webp
 

waxinglyrical

Well-known member
So! This is indeed the same bug as Magic Fang, and in the short term, is getting the same band-aid - a max caster level of 20. However, behind the scenes, the real bug is still looming... unfixed... and ready to strike.

The real bug is hidden deep within the Progressions system within the game. A Progression is a piece of data in DDO that associates a level value with a result. In the case of these misbehaving spells, the Progression associated with the duration of the spell is what's breaking. What's happening is that the association data for "Spell Duration, 5 Minutes Plus 1 Minute Per Level" only scales to level 40. Any level beyond that, and the progression is returning null, which is a fancy word for "no data" and in a practical sense means that the spell computes its duration out to 0. This causes the effect to fail to apply correctly.

So, in the long term, we're looking at the work to fix up the Progression system to return the highest value when the index goes out of bounds on the upper end. We still want the lower bounds to remain null (after all, we want to be warned if someone is asking for a negative caster level, that seems very bad!) but we think it might be safe to clamp the index of the top end without causing too much disruption. The tricky part here is that Progressions are the backbone of a bunch of unrelated and hard-to-test systems... like, for example, all of Random Loot. This trickiness is why we're opting for a band-aid now and a comprehensive fix a little while later, once we're sure we have the bandwidth to do some regressions on the larger systems. We're reasonably certain that this fix is going to fix a ton of seemingly-unrelated minor bugs, but it's better safe than sorry when it comes to something so comprehensive.
What an interesting and informative communication which is in stark contrast to the historic wall of silence and outward appearance of denial. So well done and long may the new informative relationship continue. Welcome back by the way :)
 
Top