RAILJACK ARMOR CHANGED since Deimos update

[Buff00n]

On 2020-11-27 at 9:32 AM, Aadi880 said:

Computer code moves in, strange, mysterious ways.

3 hours ago, nslay said:

On the other hand, I hear Warframe is also a bunch of Lua scripts! Yeah... you can definitely make typos, misuse types/functions and having missing dependencies in a scripting language and have it run correctly most of the time. Just by the very nature of scripting languages, an interpreter reads the individual statements and runs them at runtime. As it branches through conditions, it may never encounter statements with typos or misused types/functions that are present in a script. That is very scary... You can have a script run flawlessly 99% of the time and have it run for hours and then it encounters some unusual combination of conditions that reach problematic code that would never compile in a real native statically-typed programming language.

3 hours ago, nslay said:

This is awful! I can't stand it... if Warframe is actually a very large collection of Lua scripts, I am simply amazed that it doesn't have more problems! There is so much that can go wrong and probably exponential numbers of rare code paths with problematic code that can't be analyzed because scripting languages afford no apriori guarantees about typing which ties into consistent meaning, correct use of variables, functions, etc...

As someone who writes software for a living, there's nothing mysterious about it.  It does what you tell it to do.

Most modern scripting languages are compiled to some extent, maybe not to native code but usually to some intermediate "byte code".  Python and Lua, and even Javascript these days, are no exceptions.  This means there's usually some amount of build-time or run-time checking of all code.  Dynamic typing introduces problems those compilers can't catch, however, and the only real solution is a full set of test cases.  But even with C you can have rare code paths and scenarios that you need test coverage for.  It's a universal problem.

It's hard to devote development resources to writing test cases in the best of times, and I can't imagine how hard it is under a game dev crunch.  The fact that they're willing to publicly document all their bug fixes with every update means they're at least willing to admit the problem, which is still better than some other developers I've dealt with professionally.

I'm not sure if I had a point, but I should probably take my Railjack out for a spin.  It's been quite a long time.  If there's space dust issue then I'll just keep to the open spaces until they can work it out.  If there's constant hull breaches then it'll finally give those lazy crew members something to do.

[nslay]

39 minutes ago, Buff00n said:

It does what you tell it to do.

It's the devil. It does what you tell it to do even if you didn't write it or intend it for that purpose. Isn't that the devil trope? You make a deal with the devil and he gives you the thing you wanted by some other interpretation of what you actually wanted.

3 hours ago, Aadi880 said:

That will be an error. You are essentially doing "3"+4 on the return statement. This is not a real statement in Python as it will result in a TypeError. (You can't add an integer to a string via addition operation).

Oh well... you can do stuff like this:

x = np.zeros([2])

x[0] = "5.1"

You used to be able to do stuff like this:

x[1.0] = 2

I had a a large complicated detector code base that did some math somewhere to get an integer index for an array... only it was a float. So an older version numpy took that and a newer one did not. This would occur rarely after a couple hours of training the detector. So you'd launch the training when you left the office and then when came back the next day, you might find a script error and that no real progress had been made.

[Aadi880]

1 hour ago, Buff00n said:

As someone who writes software for a living, there's nothing mysterious about it.  It does what you tell it to do.

Well, ofc it would. I was kidding around with with that statement based on my experience when i started learning programming (namely VisualBasic) for the first time. :)

[Corvid]

Just did a test run, and had two separate hull fissures in a mission that I usually get through with zero, so I can corroborate that something has changed.

That said, Shields still prevent these fissures from appearing (as with all other Railjack "procs").

[DarthK3v]

6 hours ago, Corvid said:

Just did a test run, and had two separate hull fissures in a mission that I usually get through with zero, so I can corroborate that something has changed.

That said, Shields still prevent these fissures from appearing (as with all other Railjack "procs").

I just did two test runs as well, and can corroborate. Even on the easiest Earth mission, once I let my shields go down, I almost immediately got a hull rupture. Note this is not a hull breach, which only occurs when the Railjack loses all its health. So this only seems to be a change to how easily ruptures can occur. And again, as Corvid said, shields will still prevent any ruptures from occurring. According to the wiki, hull ruptures are caused by plasma procs, so this suggests the status change of Grineer fighters and crewships has been increased to nearly 100%. I say nearly as the ruptures did not always occur on the next hit received after shields went down. Even when allowing for shield gating, there sometimes was a space of a few shots before the rupture occurred.

However, as for the piloting issues, I cannot offer anything to support those claims. I had no issues flying around in empty space, and when in close proximity to debris, I was still able to get just as close as I've always been.

In summary, I think this just reinforces the importance of shield health and pilot skill. Keep your shields up and you won't have issues, and the more enemy fire the pilot can dodge the less shield damage you'll take.

[Krilexis]

Tested that out in Ruse War Field which I do from time to time and this a literally a hell. In like 30sec had a hull rupture and just after sealing it another one. A bit later in the mission had to seal 5 in a row non stop never seen this. There was also electric + fire at the same time which I believe should not happen. I ended up ignoring the last hull rupture after sealing like 10 and spending a lot of revolite and strangely Cy didnt mention it and there was no impact.

Last time I did missions in Veil there was not as much hull ruptures maybe 5 during the whole mission but there maybe 10 or more.

[Butterfly85]

Same, ran gian point last night. As soon as I lost sheilds it was instant hull rupture/fire/electrical issue, which would repeat the instant I fixed it 

All i can say is thank frak for lv10 engineering intrinsics because rj would be near unplayable solo without it in this state

[minininja77]

11 minutes ago, Butterfly85 said:

Same, ran gian point last night. As soon as I lost sheilds it was instant hull rupture/fire/electrical issue, which would repeat the instant I fixed it 

All i can say is thank frak for lv10 engineering intrinsics because rj would be near unplayable solo without it in this state

I have started to just leave the ruptures alone and just handle the fires if they happen, leaving the first rupture alone seems to stop any further from popping up

[Aadi880]

I wonder if DE thought people were not building for shields at all and just made ruptures more common so that people consider using and upgrading shields more.

[Galuf]

Can you tell me about your rig setup? Are you playing with a zekti shield? I'm too lazy to launch the game right now, but my theory is that it should still be fine during shield gating. If you're using one, I'm as lost as you.

Le 28/11/2020 à 18:10, Buff00n a dit :

It's hard to devote development resources to writing test cases in the best of times, and I can't imagine how hard it is under a game dev crunch.  The fact that they're willing to publicly document all their bug fixes with every update means they're at least willing to admit the problem, which is still better than some other developers I've dealt with professionally.

In 2020 and after all the implementations issues they had in the past, I'm baffled they did not used TDD for the implementation of interactions within RJ. They loose actually way more money by doing things by ear. Not to mention the guys who release broken features and get yelled at or even worse, fired, because at the end of the day their n+1 sucks probably more than them... For the transparency, yes we have to praise them quite a bit, there is transparency backlog wise but the work methodology is certainly to be criticized.

[Leqesai]

On 2020-11-26 at 9:45 PM, Kaiga said:

The railjack host last night got into a fight with one of the players because his maxed out railjack "was made of swiss cheese", and hit catastrophic failure after getting hit by a singular veil proxima crewship shot.

 

That's completely insane. Is DE that bad at coding that they cant completely throw some other unrelated part of the game?