Jump to content

Patch 1.4.3 to be released next week!

UomoCapra

By UomoCapra
in 2018

 Share

Recommended Posts

Gargamel

Gargamel

Posted
Posted (edited)
8 hours ago, klgraham1013 said:

Good news.  Just say the banner saying they're looking for QA engineers.

Pretty sure that's been around for a long long time.....

 

Edit: Whoops, noticed it changed. 

Edited by Gargamel
Link to comment
Share on other sites
Majorjim!

Majorjim!

Posted
Posted
14 hours ago, RoverDude said:

twiddling done to get their thrust center improved

Your use of the word 'improved' and not 'fixed' is concerning. Do all engines have that 'floating point' offset error? Are they as 'improved' as the code will allow?

Link to comment
Share on other sites
LameLefty

LameLefty

Posted
Posted
17 hours ago, RoverDude said:

Sorted.

Fantastic news. With luck, this will drastically improve my encapsulated parachute-drop payload landings on Laythe. Previously, using MechJeb's landing guidance module, I could para-drop payloads within one - two dozen meters of each other. Now I've got some overshooting my 2 - 3 kilometers, and others undershooting by almost as much. 

Link to comment
Share on other sites
klesh

klesh

Posted
Posted

I guess the bugtracker doesn't get updated very well then?  I still see the majority of what has been posted there from 1.4.1 and 1.4.2 marked as "New", "Investigating", "Being Worked On", including one of the bugs I posted which the OP confirms will be included as a fix in 1.4.3.  That particular one, regarding the orientation of the MH spawnable launchpad, shows a progress of 30%, but we're told here its complete?  Either there will be a 1.4.4 or all of those issues are mislabeled, I guess.

How does anything get marked as "done" on the bugtracker, when even the stuff that has been fixed is left as "investigating" etc?  They must have some method of preventing one another from looking onto bugs that a colleague has fixed, but didn't close down or update the ticket?

 

 

Link to comment
Share on other sites
steve_v

steve_v

Posted
Posted (edited)
34 minutes ago, klesh said:

How does anything get marked as "done" on the bugtracker, when even the stuff that has been fixed is left as "investigating" etc?  They must have some method of preventing one another from looking onto bugs that a colleague has fixed, but didn't close down or update the ticket?

It just gets unusably messy, then gets a "cleanup" later on, at least that's what happened last time. I suspect there's another bugtracker that we don't get to see, and select reports are pulled from the public one.
It seems like dev attention to the public bugtracker isn't much greater than dev attention to forum bug posts... That'd be cool if it meant they were busy fixing bugs, but ya know, joysticks on Linux.

It'd sure be nice if the bugtracker was updated properly, preferably with other useful stuff like target releases and meaningful progress percentages, then I could stop complaining about the "total silence" treatment for certain bugs...

Edited by steve_v
Link to comment
Share on other sites
Lisias

Lisias

Posted
Posted
2 minutes ago, steve_v said:

It seems like dev attention to the public bugtracker isn't much greater than dev attention to forum bug posts... 

Neither they should. One best practice on Project Management is shielding the development from the stake holders (us). :-)

You don't want "stalking" holders messing with your dev guys! 

Link to comment
Share on other sites
steve_v

steve_v

Posted
Posted (edited)
7 minutes ago, Lisias said:

One best practice on Project Management is shielding the development from the stake holders

Sure, but that's not the same as not providing any status updates on reported bugs.
How about a quick "We're working on it" message and setting "target version" to "1.4.4", rather than just setting the status to "investigating" an leaving it at that for eternity?

Besides, in other (though generally open-source) projects I have been involved with, direct dev interaction on bugtrackers is not at all uncommon. Seems to work just fine.

Edited by steve_v
Link to comment
Share on other sites
Lisias

Lisias

Posted
Posted (edited)
32 minutes ago, steve_v said:

Sure, but that's not the same as not providing any status updates on reported bugs.
How about a quick "We're working on it" message and setting "target version" to "1.4.4", rather than just setting the status to "investigating" an leaving it at that for eternity?

Something like this? :-)

3lc6fqnSumBM2vHTPB6L3VrN3SQYkHho3IY-_hSU

 

32 minutes ago, steve_v said:

Besides, in other (though generally open-source) projects I have been involved with, direct dev interaction on bugtrackers is not at all uncommon. Seems to work just fine.

Except when it doesn't. I'm on the Open Source "biz" for 20 years now (not exclusively). I know uncountable cases of terrible situations made worst by a lapse of a developer. In almost all the occasions, the dev was right and did the right thing but didn't had the social skills to recognize and prevent falling in a trap, and the after math was a hell of a P/R nightmare for the project. Sometimes, a very costly nightmare.I had seen people loosing jobs on this.

People don't hire developers by their social skills. People hire them due their development skills - and these skills are not mutually inclusive. (to tell you the true, they are almost mutually exclusive - I know, I'm one of them)

41 minutes ago, steve_v said:

That'd be cool if it meant they were busy fixing bugs, but ya know, joysticks on Linux.

You mean this? :-)

QWPViBz4F7I0ZId1GuZZM3_PC6bHLbKOQ9senqb7

Edited by Lisias
Link to comment
Share on other sites
steve_v

steve_v

Posted
Posted (edited)
14 minutes ago, Lisias said:

Something like this? :-)

Not really, though it's better than not closing them. Again, for clarity:

40 minutes ago, steve_v said:

a quick "We're working on it" message and setting "target version" to "1.4.4"

My point is that no indication is given as to the progress on a bug or in which version it is likely to be resolved. Which results in me being annoyed that there is another patch announcement with serious bugs going unmentioned.
In this particular case*, after "investigating" for a month I would expect  that someone has an inkling as to when it might be fixed.

14 minutes ago, Lisias said:

You mean this? :-)

*Indeed. Note the lack of a target version, any comment from the devs, any official updates at all in the last 30 days, and the arbitrary "20%" set when the status was set to "investigating". Not exactly informative WRT when the game is likely work properly again on GNU/Linux.

Edited by steve_v
Link to comment
Share on other sites
Lisias

Lisias

Posted
Posted
3 minutes ago, steve_v said:

My point is that no indication is given as to the progress on a bug or in which version it is likely to be resolved. Which results in me being annoyed that there is another patch announcement with serious bugs going unmentioned.

Oh, now I see. You have a point, indeed.

lAMxiyjUOTK-cLgyx674dyEnNDuKj3tfTMAG4s_s

 

6 minutes ago, steve_v said:

In this particular case, after "investigating" for a month I would expect  that someone has an inkling as to when it might be fixed.

I think we are getting close to the real issue. There's a problem on the "Expectation Management".

Link to comment
Share on other sites
steve_v

steve_v

Posted
Posted
5 minutes ago, Lisias said:

Oh, now I see. You have a point, indeed.

Whoop-de-do, I can filter on target version... What does this have to do with anything?

 

6 minutes ago, Lisias said:

There's a problem on the "Expectation Management".

Expectation: Regressions this obvious are dealt with swiftly, or ideally, caught long before release.
Expectation: Two patches (one of which made other problems worse) do not go by without any kind of update on basic functionality being restored on one of the supported platforms.
Expectation: Restoring said basic functionality takes higher priority for a patch than introducing additional content.

Sorted, defined and managed. Not to your liking perhaps, but that's not my problem.

Link to comment
Share on other sites
Lisias

Lisias

Posted
Posted (edited)
55 minutes ago, steve_v said:

Whoop-de-do, I can filter on target version... What does this have to do with anything?

There's no 1.4.3 target of the filter (essentially what you are complaining). Dude, I'm not your enemy - drop your guns, please.

 

55 minutes ago, steve_v said:

Expectation: Regressions this obvious are dealt with swiftly, or ideally, caught long before release.
Expectation: Two patches (one of which made other problems worse) do not go by without any kind of update on basic functionality being restored on one of the supported platforms.
Expectation: Restoring said basic functionality takes higher priority for a patch than introducing additional content.

You are forgetting one: "Paid bills". The *real* management problem is how to accomplish all of that and have the bills paid at the end of the month.

55 minutes ago, steve_v said:

Sorted, defined and managed. Not to your liking perhaps, but that's not my problem.

Drop your guns, please.

(Mainly because your sorting is far from being ideal from the Project Management point of view)

Edited by Lisias
far is "far", not related to the mod! :-)
Link to comment
Share on other sites
steve_v

steve_v

Posted
Posted (edited)
23 minutes ago, Lisias said:

There's no 1.4.3 target of the filter

That does not prevent SQUAD from adding it, does it?

 

23 minutes ago, Lisias said:

You are forgetting one: "Paid bills".

And I'll be happy to help them with that by buying the DLC, after this blindingly obvious bug is addressed.

 

23 minutes ago, Lisias said:

Dude, I'm not your enemy

No, but

23 minutes ago, Lisias said:

Management...
management...
Management

Those ratbuggerers too often are.
I'm far more interested in fixing that which should be fixed than appeasing managers, beancounters or marketing departments.

 

23 minutes ago, Lisias said:

far is "far", not related to the mod! :-)

Too late, this ridiculous forum software has already attached it :P

 

 

Observe the previous iteration of this bug: Set to "confirmed" by SQUAD staff, no updates at all until a user reports that it is fixed, then no further updates for nearly a year, after which it is simply closed.

Nobody assigned, no target version set, no indication of when a fix is ready to test... This is not how you bugtracker.

Edited by steve_v
Link to comment
Share on other sites
Lisias

Lisias

Posted
Posted
15 minutes ago, steve_v said:

That does not prevent SQUAD from adding it, does it?

Nope. But now it's too late, we already know they didn't planned 1.4.3 as they usually does (and should). This is the danger of open bug-tracks, they usually tells us more than they wanted us to know. =P

17 minutes ago, steve_v said:

And I'll be happy to help them with that by buying the DLC, after this blindingly obvious bug is addressed.

It can be a problem, or a solution. Or both! :-)

The "real" Squad's client, since 2016, is TTI. We are TTI's clients now, not Squad's. So, in order to help Squad, we need to deal with TTI's first.

The cash cow is, usually, the best way to influence business decisions - but it's also the reason they use for shutdown products. It's tricky to reach the ideal balance.

But it's also the way I'm adopting - I almost had bought the Mission Builder at launch. Yes, I'm very happy with KSP - besides the bugs - it's some time since a game give me so much fun as the last weeks (i bought it at 1.4.0 times). But then they launched 1.4.2, what disrupted (seriously) my way of editing crafts (man, how I hate when things "snaps" by itself even by the snap button being disabled!), and I decided that buying the DLC would be a bad strategical decision as a customer: So I'm also waiting the 1.4.3 version to see what happens.

 

28 minutes ago, steve_v said:

[about Managers]

Those ratbuggerers too often are.
I'm far more interested in fixing that which should be fixed than appeasing managers, beancounters or marketing departments.

They are the man with the guns, I mean, money. There's not way to live without dealing with them. :-)

I'm a technician. I'm a developer. I value the ending result, I work for the product. And yeah, I had my load on attracting some Management rage over me. Then I decided to learn about these guys, and took some PM lessons. Well, I'm still a technician, I still work for the product - but at least know how to deal with them: follow the money. It's always about the money.

(Yeah. It <use your non-forum-policy-conformant-favorite-interjection-here>).

36 minutes ago, steve_v said:

Observe the previous iteration of this bug, set to "confirmed" by SQUAD staff, no updates at all until a user reports that it is fixed, then no further updates for nearly a year, after which it is simply closed.

Nobody assigned, no target version set, no indication of when a fix is ready to test... This is not how you bugtracker.

Point taken. I agree with you. Too. :-)

Some common ground between the internal management needs and the public communications needs must be worked out.

Link to comment
Share on other sites
N70

N70

Posted
Posted

As a modder, can we get better CommNet APIs, like being able to custom set plasma blackout, or making custom icons on planets similar to launchpad icons without pesky on_gui madness?

Link to comment
Share on other sites
GDJ

GDJ

Posted
Posted
17 minutes ago, N_Molson said:

Another launch site with runway ? Very cool ! Kerbin needs moar places to land at !

Oh yes! Tre' Cool!

 

Link to comment
Share on other sites
asus legion

asus legion

Posted
Posted
---------------Detectar idiomaAfrikáansAlbanésAlemánAmáricoÁrabeArmenioAzerbaiyanoBengalíBielorrusoBirmanoBosnioBúlgaroCanarésCatalánCebuanoChecoChino (Simplificado)Chino (Tradicional)CingalésCoreanoCorsoCriollo haitianoCroataDanésEslovacoEslovenoEspañolEsperantoEstonioEuskeraFinésFrancésFrisón occidentalGaélico escocésGalésGallegoGeorgianoGriegoGuyaratíHausaHawaianoHebreoHindiHmongHúngaroIgboIndonesioInglésIrlandésIslandésItalianoJaponésJavanésJemerKazajoKirguísKurdoLaoLatínLetónLituanoLuxemburguésMacedonioMalayalamMalayoMalgacheMaltésMaoríMaratíMongolNeerlandésNepalíNoruegoNyanjaPanyabíPastúnPersaPolacoPortuguésRumanoRusoSamoanoSerbioSesotho meridionalShonaSindhiSomalíSuajiliSuecoSundanésTagaloTailandésTamilTayikoTeluguTurcoUcranianoUrduUzbekoVietnamitaXhosaYidisYorubaZulúEspañol
 
  
WE'LL SQUAD NEVER LEAVE KERBAL SPACE PROGRAM THANK YOU FOR ALL THE WORK SINCE 2011 WORKING DESERVE MY RESPECTOS THANK YOU FOR ALL THE FUNNY THAT YOU GAVE US AND YOU WILL CONTINUE GIVING THANK YOU VERY MUCH !!!!!!!!
Link to comment
Share on other sites
GeneCash

GeneCash

Posted
Posted (edited)

I dunno man... they've fixed so many bugs (at least on Linux) that after completely not playing 1.3, I'm doing 8 hour KSP sessions again.

I haven't had the mysterious crash-to-desktop that used to happen at least once or twice an hour. Rovers are not wildly flipping beasts, and I have not had a single one suddenly fly into the void and corrupt my save.

The new high-latitude launch platform has invigorated Mun & Minmus launches as I figured out the best way to get into low inclination moon orbits.

My joysticks work fine on Linux and the only bugs I've seen is that DSN stations cause Kerbals to explode, and the new pad ladder isn't... but now all the ladders around KSC work, even the helipad ladders on top of the VAB and the spherical tank ladders.

I'm happy.

Edit: oh yeah, and contracts finally beyond landing on the moons. I finally got a contract to do things around Duna and even Dres! Plus I was surprised by a Sentinel contract, and contracts to do things with Kerbals and rovers already landed!

Edited by GeneCash
Oh yeah
Link to comment
Share on other sites
AVaughan

AVaughan

Posted
Posted (edited)
13 hours ago, steve_v said:

Expectation: Regressions this obvious are dealt with swiftly, or ideally, caught long before release.
Expectation: Two patches (one of which made other problems worse) do not go by without any kind of update on basic functionality being restored on one of the supported platforms.
Expectation: Restoring said basic functionality takes higher priority for a patch than introducing additional content.

You need to calm down.  You need to get a sense of perspective.   You post is arrogant, rude and toxic.  Posts like that drive developers away from the forums because they make reading the forums an emotionally draining and unpleasant experience. 

First lets cover your expectations line by line.  "Regressions this obvious are dealt with swiftly, or ideally, caught long before release."   If a bug only happens on some systems, then the bug is only "obvious" if you have one of the affected systems.  Since GeneCash reports his joystick is working on linux it is quite possible that joysticks work on whatever systems Squad uses to test the linux version, which means that there is no reason for this regression to be "obvious" to them.  (This would also make it harder to diagnose and fix the bug if they can't reproduce it). 

"Two patches (one of which made other problems worse) do not go by without any kind of update on basic functionality being restored on one of the supported platforms".  Basic functionality is build, launch and fly rockets, planes and rovers.  Also core ancillary functions like load and save, and the science and mission systems.  Joystick is more optional functionality, since you can play the game without it.  Yes it's supported functionality, and yes it should be fixed.  (And yes I can see how for some people joystick support might be central to their ksp gameplay experience, but many people play ksp without having a joystick, so in the greater scheme of things it isn't core functionality).  It would be nice if the developers communicated more on the bug forums, but given how toxic certain members of the community are, I can see why they they don't.  (I do agree that 1.4.2 seems to have been rushed out without adequate testing).

"Restoring said basic functionality takes higher priority for a patch than introducing additional content."  I've already covered my opinion about the basic functionality bit.  You seem to imply that this is a case of do A or B.  It's not.  We don't know who created the extra missions we are getting with 1.4.3, but I see nothing to suggest that they were created by someone who would be otherwise have been working on a linux specific bug.  (It's also possible that Squad have already decided it's not a bug in their code, and have forwarded the bug upstream to Unity and are waiting for a fix from them).

Personally I think that the root of the bug might be a client-side configuration issue exposed by the change from Unity 5.04 to Unity 2017.1.  Amongst other changes Unity changed from using the legacy joystick API to using SDL2 and the modern evdev API.  

See https://wiki.archlinux.org/index.php/Gamepad#evdev_API for instructions on testing if your joystick is working with evdev and SDL2.  For some possible configuration issues and fixes see https://forum.unity.com/threads/no-joystick-detected.475870/ and https://discourse.libsdl.org/t/sdl2-requires-root-privileges-to-use-joysticks-in-linux/20181 .  Both of those problems might result in joysticks working in 1.3, but not working in 1.4.   

Edited by AVaughan
Fixed link
Link to comment
Share on other sites
steve_v

steve_v

Posted
Posted (edited)
45 minutes ago, AVaughan said:

You need to calm down.

I am quite calm.
 

45 minutes ago, AVaughan said:

Since GeneCash reports his joystick as working on linux it is quite possible that joysticks work on whatever systems Squad uses to test the linux version, which means that there is no reason for this regression to be "obvious" to them.  (This would also make it harder to diagnose and fix a bug if they can't reproduce it). 

The bug is set "investigating" not "need more info", and no comment WRT to reproduction or affected systems has been made there, from which I read they're not looking for more info. If they were, I'd gladly provide it.

 

45 minutes ago, AVaughan said:

See https://wiki.archlinux.org/index.php/Gamepad#evdev_API for instructions on testing if your joystick is working with SDL2.

The Arch wiki is not particularly relevant for Gentoo or Debian, the two systems that I have reproduced this on. However, sdl2-jstest works perfectly with my device on both, and I fixed permissions on the input device long ago.
ed. I can now add Artix to that list as well.

 

45 minutes ago, AVaughan said:

You post is arrogant, rude and toxic.

I began as accepting and patient. Patches came and went, no mention was made. My patience has expired.
A little bit of (or any) communication would have gone a long way.

My complaint is not that this is not fixed yet, it's that there's been no communication and no apparent progress.
If I see "We hope to have it sorted for 1.4.4", I'll be happy.
Ask me "Can you test [X] and report back" and I'll hop to it.
Give me deafening silence, and I'll be annoyed.

Edited by steve_v
Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...