I can't do this anymore

GrizzoMtGPQ

Brigby said:

@Mburn7 I totally understand why players would be upset, especially if it's a card they use frequently. Seeing it function in a manner that doesn't follow its wording can be confusing and frustrating, and I apologize for that inconvenience.

As I mentioned before, this is purely my speculation as to how they test. If that is the correct test method though, then I believe bugs like Verix Bladewing's token are an unfortunate result of testing overall mechanics, as opposed to individual cards.

The ability to deal damage on entrance works correctly. Damage dealt by that ability works correctly. The ability to convert gems to a specific color works. Once you combine all three together into Verix Bladewing's token though, then that's where the bug occurs. 

I honestly believe that is one of the biggest challenges the team faces. How do you efficiently test ~2,000 cards each time you receive a new QA build, especially when MtGPQ is a live game; not a standalone one.

@Brigby you write automated regression tests to do this. It is a well known thing in software development. Every time you produce a new build you re-run the tests to ensure your new code/cards didn't break anything. I've reached out the development team but never been able to make a connection. 

Brigby

GrizzoMtGPQ said:

Brigby you write automated regression tests to do this. It is a well known thing in software development. Every time you produce a new build you re-run the tests to ensure your new code/cards didn't break anything. I've reached out the development team but never been able to make a connection. 

True, but wouldn't that be limited to the combinations of interactions the development team can come up with, or is there a way to systematically be like, "Ok computer. Here are all the cards in the game, and all of their respective abilities. Now test any and all possible combinations. Go!"

(If you can't tell, I only did black box testing, not white-box testing, when I was in QA. )

FindingHeart8

Brigby said:

GrizzoMtGPQ said:

Brigby you write automated regression tests to do this. It is a well known thing in software development. Every time you produce a new build you re-run the tests to ensure your new code/cards didn't break anything. I've reached out the development team but never been able to make a connection. 

True, but wouldn't that be limited to the combinations of interactions the development team can come up with, or is there a way to systematically be like, "Ok computer. Here are all the cards in the game, and all of their respective abilities. Now test any and all possible combinations. Go!"

(If you can't tell, I only did black box testing, not white-box testing, when I was in QA. )

Unknown

This content has been removed.

HarryMason

I brought punch and pie ! The number of bugs is definitely getting epic. There's a multitude of cards that don't do what they say. There's very little consistency to how cards work . Some support destruction will auto cast with no targets. Some won't, but cards like hour of destruction, which don't target , can't be cast with no creatures on board. Lots of 'kicker' cards drain different amounts of mana than stated. It feels like there's no attention to detail . I've been over it for a while, but I also haven't spent any money on the last 3 sets . I'm just trying to have fun in spite of the bugs. Broken game is broken. Meh 

DumasAG

My understanding was that we always had to be patient on bugs until they caught up to Dominaria, then the focus would switch to smoothing out all the technical problems with the game. Is that no longer correct?

Tilwin90

@DumasAG
Not really because M19 core is coming. And let's be serious, there will always be new sets and new content to be made. It's a neverending story unless the team changes strategy and keeps a healthy mix between new content, technical stabilization (refactoring and such) and bug-fixing.

FindingHeart8

DumasAG said:

My understanding was that we always had to be patient on bugs until they caught up to Dominaria, then the focus would switch to smoothing out all the technical problems with the game. Is that no longer correct?

exactly, I'm not trying to be an ungrateful snobgoblin here, but we're not really seeing significant progress on a new set or bug fixes.  Again (broken-record here) it ultimately boils down to poor communication on Octagon's part if they are actually working on some big changes at the moment.  But it's challenging to be patient when you really have no idea what's going on.  Brigby's attempts to calm the flames of player impatience are nice, but ultimately he's in the dark too.  I thought Diane works for Octagon, but I haven't seen a comment from her since she first made a profile here.

I need a drink.

Brakkis

Brigby said:

Now consider MtGPQ, its collection of almost 2,000 unique cards, and the small, but passionate QA team we work with! If we were to only factor in just the interactions of the 10 cards in a single deck, the number of possible permutations would be 2,000 factorial to the 10th number. That's a big number; one that would be impossible for any team to manually test.

While the response you gave was very helpful in understanding why there's such a high possibility of bugs slipping past, it fails to account for the current scenario involving supports.

With 2.7.1, one of the bug fixes was to make certain support destruction cards stop targeting indestructible supports. Now, they no longer target those supports for sure, but if there is an indestructible support on the field, they don't target anything. These cards will now either not be cast-able, even if they have a secondary part to them - i.e. Fiery Intervention won't let you target a creature for 3 damage if an indestructible support is on the board; or, they will just fizzle out and not do anything at all - i.e Demolish will cast and do nothing... at all.

That's not a matter of thousands, hundreds, or even tens of cards. The bug fix was for a very select few cards and their interaction with an indestructible support in play. How could this have not been found through a single testing of the fix?

Additionally in 2.7.1, the change to supports giving priority support destruction over treasures was introduced. Now we have Vraska's first ability taking out token supports - including Servo's and Clue's - before the support with the highest shield count. Likewise, Antiquities War now gives +x/+x where x is the shield count of the highest token support - again, including Servo's and Clues - rather than +x/+x where x is the shield count of your highest shield support.

These are also not a matter of thousands, hundreds, or even tens of cards. This is an interaction between a single PW and at most, 4 cards. A token support generator, a creature, a regular support, and antiquities war.

Neither of these bugs can be glossed over as hard to catch amongst the interactions of multitudes of cards. These were fixes designed around very specific cards and interactions and they were very obviously not tested. Coded in, but never tested.

Brigby

@Brakkis To preface what I'm about to say, please note that I'm not a developer, nor do I know exactly how the game's inner workings are structured. I am merely speculating and hypothesizing based on my previous experience as a QA tester, and my basic knowledge of programming.
-----
If there are cards that no longer function correctly after this change, then of course that is a bug which should be fixed. Support destruction cards not targeting anything after the change, even if they have a secondary effect, is a clear example of this. This bug fix though was not specifically targeting a couple of cards, and why I say that can be explained by the way I believe abilities are programmed in the game.

Since there are thousands of cards in the game, programming each individual card to have specific abilities would be a very tedious and time-consuming process, especially if changes are made to the ability afterwards. Instead, I believe that abilities are programmed as Classes, and if a card has that ability, then it will reference that Class.

Picture an ability (let's say First Strike) as paint, and each card that has First Strike as an artist. Instead of going one by one to hundreds of artists and giving them a can of paint each, you place a giant tub of paint (the ability's Class) in the center of the room and tell all the artists to dip their brush in that tub, whenever they need to use that color.

Now if you want to change the color of the paint, all you have to do is fill the tub with a different color, and then voila! Every single artist is using that color paint now. No need to go one by one and replace each artists' can of paint with new ones when you want to make a change.

The problem you're seeing with this structure is that inevitably there will be edge cases where some cards have other effects that rely on this ability interacting in the old way. Those would be the ones that developers would need to debug and fix individually, however I speculate the time and effort required to do that is far smaller amount than having to program each card on an individual basis.

ArielSira

Even after the paint example, it just baffles me that the changes of priority in regards to targetting treasures has the *opposite* effect on Vraska's ability or Antiquities War (from not targetting treasures to targetting them first and from selecting highest shield to selecting lowest).

Really hope it gets fixed asap, she's lost her main trick now.

Brakkis

Brigby said:

@Brakkis To preface what I'm about to say, please note that I'm not a developer, nor do I know exactly how the game's inner workings are structured. I am merely speculating and hypothesizing based on my previous experience as a QA tester, and my basic knowledge of programming.
-----
If there are cards that no longer function correctly after this change, then of course that is a bug which should be fixed. Support destruction cards not targeting anything after the change, even if they have a secondary effect, is a clear example of this. This bug fix though was not specifically targeting a couple of cards, and why I say that can be explained by the way I believe abilities are programmed in the game.

Since there are thousands of cards in the game, programming each individual card to have specific abilities would be a very tedious and time-consuming process, especially if changes are made to the ability afterwards. Instead, I believe that abilities are programmed as Classes, and if a card has that ability, then it will reference that Class.

Picture an ability (let's say First Strike) as paint, and each card that has First Strike as an artist. Instead of going one by one to hundreds of artists and giving them a can of paint each, you place a giant tub of paint (the ability's Class) in the center of the room and tell all the artists to dip their brush in that tub, whenever they need to use that color.

Now if you want to change the color of the paint, all you have to do is fill the tub with a different color, and then voila! Every single artist is using that color paint now. No need to go one by one and replace each artists' can of paint with new ones when you want to make a change.

The problem you're seeing with this structure is that inevitably there will be edge cases where some cards have other effects that rely on this ability interacting in the old way. Those would be the ones that developers would need to debug and fix individually, however I speculate the time and effort required to do that is far smaller amount than having to program each card on an individual basis.

I do like your explanation, and it helps me further understand the way things are done in terms of the development.

Again though, this is a matter of a lack of testing. They changed the paint, but they didn't bother to see if the color worked.

I completely understand that bugs do happen. They are inevitable and no game is ever perfect and free of them. It's the simplicity with which some of these things can be tested, and aren't, prior to release that is what I think is driving people crazy with the game.

Machine

When the "Legendary" icon was introduced, I can imagine every card in the game was inspected to check if it should get this evergreen icon. That would also have been a great opportunity to match (and correct) card texts. Cards that have the same effect are sometimes worded differently and other cards do completely different things compared to how they are described, e.g. Dreadwaters. It just drains all mana from all cards in opponent's hand. No trap gems anymore.

The same thing applies for spelling errors. I see more and more of them and that just feels wrong for a game with this magnitude. I get that mistakes are made, we all do, but double checking cards, objectives and abilities isn't that hard.

Unknown

This content has been removed.

Tilwin90

Machine said:

When the "Legendary" icon was introduced, I can imagine every card in the game was inspected to check if it should get this evergreen icon. That would also have been a great opportunity to match (and correct) card texts. Cards that have the same effect are sometimes worded differently and other cards do completely different things compared to how they are described, e.g. Dreadwaters. It just drains all mana from all cards in opponent's hand. No trap gems anymore.

The same thing applies for spelling errors. I see more and more of them and that just feels wrong for a game with this magnitude. I get that mistakes are made, we all do, but double checking cards, objectives and abilities isn't that hard.

Unfortunately I would have to disagree here. Your solution falls under "trying to do too many things at the same time". 

Whenever a change is applied, there is always the risk of messing something up inadvertently. Now chain multiple changes at once, and suddenly you don't know which of the changes led to the problem. This increases investigation time (which is already skyrocket long if we look at the current bug list) and can only lead to MORE problems afterwards.

In your example, I think it was wise to only add the legendary "ability" to cards without performing other changes in the process simultaneously. But based on community feedback they should compile a list of inappropriate card texts and fix those specific cards as a separate issue.

Laeuftbeidir

I'd love to contribute to such a list. If the devs actually would ask for help - or accept the offered help - I know several people who'd jump in immediately.

I think @Machine was referring to card texts only, those should have no effect on the way a card works and are pure "cosmetical".

Although I like @brigby s explanations for being interesting I don't see how this situation could ever improve with the focus on adding new sets (that's the focus we know about. Maybe they're actually trying to fix the game without telling us?) . Bugs will get worse and more complex with every release. I understand how several bugs won't get fixed, since they might be hard to track and occur not that often (like the overload cascade wrong board side effect), but I don't understand how totally simple and easy to track bugs get ignored for multiple months. It's one thing when QA regulary fails to find a bug, but another thing to ignore reported bugs.

BTW,

The bug subforum is a total disaster and could really need a redesign to account for the sheer amount of bugs reported.

James13

Laeuftbeidir said:

The bug subforum is a total disaster and could really need a redesign to account for the sheer amount of bugs reported.

I used to post observed bugs only in the bugs subforum.  But now I skip the forum and submit actual tickets in the app.  That way I know it goes into a review queue.

Mburn7

James13 said:

Laeuftbeidir said:

The bug subforum is a total disaster and could really need a redesign to account for the sheer amount of bugs reported.

I used to post observed bugs only in the bugs subforum.  But now I skip the forum and submit actual tickets in the app.  That way I know it goes into a review queue.

Funny, I do the opposite.  I gave up on submitting tickets since the bugs never got fixed and started submitting them here.

wereotter

Mburn7 said:

James13 said:

Laeuftbeidir said:

The bug subforum is a total disaster and could really need a redesign to account for the sheer amount of bugs reported.

I used to post observed bugs only in the bugs subforum.  But now I skip the forum and submit actual tickets in the app.  That way I know it goes into a review queue.

Funny, I do the opposite.  I gave up on submitting tickets since the bugs never got fixed and started submitting them here.

I post it if it’s not acting correctly to player detriment, like Seismic Shift doesn’t destroy enough gems. I just started reporting bugs players can exploit to troll opponents using the in-app feature as I don’t trust everyone not to exploit the system if they know they can. 

Unknown

This content has been removed.