Jump to content

Alexey Kryshen

Members
  • Content Count

    8
  • Joined

  • Last visited

Community Reputation

2 Neutral

About Alexey Kryshen

  • Rank
    (1) Prestidigitator
  1. You are absolutelly right. And I said that not as user, but as developer.
  2. Programmed ? Of course! Released ? Never, except exactly one. For just a couple of reasons: Spend a lot of time to make perfect architecture Defensive coding Competent QA team Good architecture leave no places for most stupid "bugs", defensive coding style & techniques forced across team by team lead(s) with apropriate monitoring (code review, etc.) leave no places for pesky "bugs". In most cases defensive coding allows to detect these "bugs" at first run after they are was introduced (in simplest cases via the appropriate assertions and invariants checks across the code). I personaly not even count these as a bugs because it mostly never even reach the QA team and detected by programmer himself during write-run-check cycles. For me bugs are the errors which are detected by QA team (or, the worst case - by user) and which are really not obvious errors in code. It hard to believe but my team never released any "game-braking" bugs. Actually we never released any major bugs, thanks to 3 points I have mentioned above (well, it is also because quality of the products was a priority for our company) I have only one relatevely serious bug released in my practice. It was threads synchronization bug. It reveal itself once in a 2-3 month (for a QA machine run in 24/7 mode, just to say), so it was really tough beast. It was never fixed because whole subsystem were refactored and rewritten (not due to the bug but for other purposes). So, PoE situation is not normal under any circumstances. It is not normal to have first-day patch, it is not normal to have tons of obvious bugs detected by users in fisrt 3 days after product release. I would say it is unacceptable, especially for bugs in game mechanic, which are essential for such game types. Does we complain about the annoing but harmless UI bugs ? No! We complain about broken abilities which could be detected by QA team before release, we complain about broken combat math which could be perfectly self-tested by set of good unit-tests or via other techniques. So we complain agains the things which never should occur for released product. I hope you get my point now. Hey Alex, What was the name of the game your company released, if you can tell us? At least describe it Like the core system (Epic's Unreal engine, Gamebryo, etc.) and the target platform (PC, Xbox, etc.) What was the size of the dev and QA team, and what was your ballpark budget? I would prefer not to disclose such information. I hope you will not be offended.
  3. Why to be so offensive ? If you've never seen reliable programs, it does not mean they do not exist, right ? Moreover, you deliberately distorted the meaning of my words in favor of the position you are defending, which is not good. The Poe was implemented very badly, why not recognize this obvious fact and do not attack people who simply stating this fact. When you buy broken car you have the way to compensate this. When you buy broken software you have no way to compensate this. In a case of PoE Obsidian could just stop fixing the endless bugs in the game and we all will end up with useless piece of data on HDD, money will be just throw away. This is why it is absolutelly necessery to force software developers to provide reliable software at release date. Please note that I am the developer himself, so I said it not because I am angry user. You are right about the delayed project & related cost of it, but for PoE in this case it would be far more fair in respect to players to release under the Early Access terms. Obsidian would take the money to finish the development and users would be not offended by false promising. Isn't ? I think most of us here would agreed that PoE in current state is most suitable for Early Access than for Production Release ?
  4. Programmed ? Of course! Released ? Never, except exactly one. For just a couple of reasons: Spend a lot of time to make perfect architecture Defensive coding Competent QA team Good architecture leave no places for most stupid "bugs", defensive coding style & techniques forced across team by team lead(s) with apropriate monitoring (code review, etc.) leave no places for pesky "bugs". In most cases defensive coding allows to detect these "bugs" at first run after they are was introduced (in simplest cases via the appropriate assertions and invariants checks across the code). I personaly not even count these as a bugs because it mostly never even reach the QA team and detected by programmer himself during write-run-check cycles. For me bugs are the errors which are detected by QA team (or, the worst case - by user) and which are really not obvious errors in code. It hard to believe but my team never released any "game-braking" bugs. Actually we never released any major bugs, thanks to 3 points I have mentioned above (well, it is also because quality of the products was a priority for our company) I have only one relatevely serious bug released in my practice. It was threads synchronization bug. It reveal itself once in a 2-3 month (for a QA machine run in 24/7 mode, just to say), so it was really tough beast. It was never fixed because whole subsystem were refactored and rewritten (not due to the bug but for other purposes). So, PoE situation is not normal under any circumstances. It is not normal to have first-day patch, it is not normal to have tons of obvious bugs detected by users in fisrt 3 days after product release. I would say it is unacceptable, especially for bugs in game mechanic, which are essential for such game types. Does we complain about the annoing but harmless UI bugs ? No! We complain about broken abilities which could be detected by QA team before release, we complain about broken combat math which could be perfectly self-tested by set of good unit-tests or via other techniques. So we complain agains the things which never should occur for released product. I hope you get my point now.
  5. I am game developer with more than 10 year of experience in this area, working as programmer & as system architect. And I am asure you as an expert that all these bugs in PoE are not due to the project "magnitude" but solely due to the low-quality programming. It could take a lot of time to explain my position in details but simplest example is bug with double-clicked applying of armours & items. In properly designed & written code such kind of bugs are just impossible. It is still possibly to have bug with incorrectly applying modifiers provided with items (because it is really easy to make mistake in math in code) but definitely impossible to have different effects for different ways of performing the same action. Imagine text editor. You have main menu entry for "Cut" action, you have pop-up menu with same "Cut" action and you have keyboard shortcut "Ctrl+X" for the same action. These three different ways of activation lead to the same effect. As it should be in correctly designed & implemented code. But with PoE we had two different code-paths for single action "Apply item", and one of this code-path was broken. And there are many other things which are just obvious to ones who have good programming experience & especially experience in code design & architecture.
  6. Hello, Start playing as a Ranger, got Marked Pray but can't see any effect of them in the battle. The damage with and without mark on the target seems equal. The description of ability said "+20 Damage", so effect should be clearly visible, right ? Is it the bug ? Or maybe ability description is wrong ? How could one check if ability working ? I was trying to inspect the battle log carefelly but got no any explicit signs of this ability effects. Thanks in advance!
  7. I am interesting in this too. Play as Chanter, got Phantom at start and Skeletons at Level 2, badly surprised when it realised that only one summon could be active.
  8. This problem also affect Fullscreen mode on multi-monitor system. The behaviour is exactly as stated by topic starter. The Alt+Tab is required to make cursor work correctly again.
×
×
  • Create New...