SF2 Rework, bug fixes and request topic

[Stratos 3,191]

I open this post in order to get the needed lists for both, bug Fixes and for Features Request. I please to start with bug fixes, specially requesting the help with the users that know better the bug the game has. Thanks!

[Stratos 3,191]

RESERVED for BUG FIXES LIST

[Stratos 3,191]

RESERVED for FEATURES REQUEST LIST

[+Gunrunner 314]

Let's start with something that's kind of a corner case but irritates me greatly on my Madagascar/Mozambique install.

Stock Frequency : The frequency at which the bug can be reproduced in a full stock game, from 0 (Never) to 5 (always when following the conditions).

Specific Frequency : The frequency at which the bug can be reproduced in the specific game setup, from 0 (Never) to 5 (always when following the conditions).

Severity : How bad the bug is.
    1 (occasional annoyance),
    2 (Recurring annoyance),
    3 (Immersion breaking, game can be played, missions won, but things are weird enough you're constantly aware of it),
    4 (Gameplay breaking) or
    5 (Game breaking/crashing/freezing/CTD).

Test environment : Either stock, a link to the necessary files, on demand so people interested in testing and reproducing can contact the bug reporter
to obtain the necessary files, or closed if for some reason the bug reporter can't provide a test environment.

Environment :
July 2013 patch-level, SF2NA.
Only in Single Missions, on NavalMap.
In non-stock installs.

Conditions :
If the only bases available to player are carrier groups and off-map bases.
If there are multiple friendly nations with player available planes.
If the player tries flying an ESCORT mission for a service other than FriendlyNation001.
If FriendlyNation001 only has planes using a MinBaseSize larger than the carrier generated for the player.

Expected result :
The engine generates a flight of the same nation to be escorted by the player, based on the same carrier group.

Result :
The engine generates a flight to be escorted only from FriendlyNation001, but failing to assign a plane with the right MinBaseSize (because it doesn't
exist), assigns one from the next available size (effectively trying to have MinBaseSize=MEDIUM planes take off from a CarrierBaseSize=SMALL carrier)
leading to the flight failing to populate and a mission that can only be failed by the player (since the ESCORTed flight never takes off, it can never
reach its objective, automatically making it impossible for the player to achieve his own).
Other missions non-player flights are similarly afflicted but since they are non-essential to the player's own mission, their failure to populate
has no impact on the player apart from skies mostly empty of friendlies.
Land bases in similarly constrained situations do not seem to suffer from the same problem as the engine seems not to limit itself to the planes
available to FriendlyNation001, more controlled testing needed to validate it.

Stock Frequency : 0
Specific Frequency : 5
Severity : 4
Test environment : On demand

Potential fixes :
- Instead of defaulting to FriendlyNation001 for the escorted flight, default to the player's nation.
- Provided there is another friendly carrier station on the map, populate another carrier group from FriendlyNation001 to serve as a base for the escorted flight.

 

Edited July 10, 2018 by Gunrunner

[+gerwin 451]

Can I point out the following.

1) There may be a tiny chance that the developer FEELS like patching up a few little things. But a developer needs to be wary of his audience, as they can behave like they are entitled to more then they actually are. . (Example: IF I would even accept donations for TFDtool, chances are someone gives a donation then asks for a feature to be added in.  Thus, he does not give the donation for what was already done, but for what he wants to be done. That is a big difference ) If you ever contact a Developer with requests, he will immediately be wary of what other expectations follow: Will patching one little bug start an avalance of similar requests and expectations. I figure that is why the Strike Fighters PC series are kinda buried on the website (with warnings), as advertising and selling them results in expectations from the buyers.

2) A reply to Stratos in particular. I understand your feeling regarding the fact that Strike Fighters is out of active development for 5 years now, and closed source as well. I dislike that as well. But this situation is actually the norm for commercial software.  So if you find it unacceptable then I could say:  use linux and GNU software instead. In addition, in the previous topic there were some misunderstandings about what is possible without the source code. Which is similar to your request+answer form a few years ago. These are just two examples of what I would call misunderstandings, and it would not help communication with a developer with such misunderstandings on the table. I don't know what to suggest, I don't mean to insult, but: Learn about it first  and/or  Let somebody else edit any communication with a developer first  and/or Let someone with a better understanding of that matter handle such.

[Stratos 3,191]

20 minutes ago, gerwin said:

2) A reply to Stratos in particular. I understand your feeling regarding the fact that Strike Fighters is out of active development for 5 years now, and closed source as well. I dislike that as well. But this situation is actually the norm for commercial software.  So if you find it unacceptable then I could say:  use linux and GNU software instead. In addition, in the previous topic there were some misunderstandings about what is possible without the source code. Which is similar to your request+answer form a few years ago. These are just two examples of what I would call misunderstandings, and it would not help communication with a developer with such misunderstandings on the table. I don't know what to suggest, I don't mean to insult, but: Learn about it first  and/or  Let somebody else edit any communication with a developer first  and/or Let someone with a better understanding of that matter handle such.

That's why I wanted to go my way, I'm not an expert in communication, and english is not even my first language, also I don't feel TK will agree with us about more developing, or If we can get the money he wants for it. That's why I feel inclined to see what I can do with the exe and the locked files, and that's it. Even in the case this idea here in this community goes forward I wouldn't want to be the person that will communicate with TK, I did in the past, but my time for that passed.

[+Menrva 4,200]

11 hours ago, Gunrunner said:

Let's start with something that's kind of a corner case but irritates me greatly on my Madagascar/Mozambique install.

Stock Frequency : The frequency at which the bug can be reproduced in a full stock game, from 0 (Never) to 5 (always when following the conditions).

Specific Frequency : The frequency at which the bug can be reproduced in the specific game setup, from 0 (Never) to 5 (always when following the conditions).

Severity : How bad the bug is.
    1 (occasional annoyance),
    2 (Recurring annoyance),
    3 (Immersion breaking, game can be played, missions won, but things are weird enough you're constantly aware of it),
    4 (Gameplay breaking) or
    5 (Game breaking/crashing/freezing/CTD).

Test environment : Either stock, a link to the necessary files, on demand so people interested in testing and reproducing can contact the bug reporter
to obtain the necessary files, or closed if for some reason the bug reporter can't provide a test environment.

Environment :
July 2013 patch-level, SF2NA.
Only in Single Missions, on NavalMap.
In non-stock installs.

Conditions :
If the only bases available to player are carrier groups and off-map bases.
If there are multiple friendly nations with player available planes.
If the player tries flying an ESCORT mission for a service other than FriendlyNation001.
If FriendlyNation001 only has planes using a MinBaseSize larger than the carrier generated for the player.

Expected result :
The engine generates a flight of the same nation to be escorted by the player, based on the same carrier group.

Result :
The engine generates a flight to be escorted only from FriendlyNation001, but failing to assign a plane with the right MinBaseSize (because it doesn't
exist), assigns one from the next available size (effectively trying to have MinBaseSize=MEDIUM planes take off from a CarrierBaseSize=SMALL carrier)
leading to the flight failing to populate and a mission that can only be failed by the player (since the ESCORTed flight never takes off, it can never
reach its objective, automatically making it impossible for the player to achieve his own).
Other missions non-player flights are similarly afflicted but since they are non-essential to the player's own mission, their failure to populate
has no impact on the player apart from skies mostly empty of friendlies.
Land bases in similarly constrained situations do not seem to suffer from the same problem as the engine seems not to limit itself to the planes
available to FriendlyNation001, more controlled testing needed to validate it.

Stock Frequency : 0
Specific Frequency : 5
Severity : 4
Test environment : On demand

Potential fixes :
- Instead of defaulting to FriendlyNation001 for the escorted flight, default to the player's nation.
- Provided there is another friendly carrier station on the map, populate another carrier group from FriendlyNation001 to serve as a base for the escorted flight.

 

Very interesting post! Are you using my Madagascar terrain? Because in the latest available version I have disabled certain mission types (including ESCORT) to prevent all known bugs and CTDs. If we know that the terrain crashes with certain mission types, TK has enabled us the opportunity to disable them. I find it very handy in those situations. Still, much thanks for the detailed report. I wish everyone posts so much details when bug-reporting to us modders.

[+Gunrunner 314]

Menrva, your terrain revision is the base, but I know my way around, it also happens with Iceland or any terrain I tried, it's definitely an engine bug*.

 

* Well, it's a corner case, the code was never meant to be faced with such a configuration, so it's not a game bug, just a case of the engine taking a shortcut because it knows which limited cases it will face and bypass the necessary steps to face other cases, it's a time-saving, cost-saving measure that's not future-proofed, but given the budget and manpower constraints of TK's operations, that is perfectly fine, it's only for us tinkerers that it becomes an issue.

Edited July 11, 2018 by Gunrunner

[+Menrva 4,200]

Another game-breaking bug I want to point out is the following:

Environment :
July 2013 patch-level, SF2NA
In Single Missions, and in Campaigns
In stock installs

Conditions :
The player flight is assigned an escort mission.
The escorted flight reaches its waypoint for its ground attack.

Expected result :
After engagement, the escorted flight reaches the egress waypoint, prompting a mission successful for the player's mission.

Result :
For reasons still unknown, it happens from time to time that the escorted flight is flying straight instead of engaging its ground target, and it never heads back home. By never reaching the egress waypoint, the player's escort mission ends up in a loop, it can't be completed.
Sometimes, simply restarting the mission is enough to trigger the correct behaviour of AI planes. Still, the bug is frustrating and it should not happen. I don't recall it happened in SF1 games.
Another solution, if possible, would require that a "mission complete" statement is triggered when the AI planes fly over their ground target waypoint, instead of the egress waypoint. No matter what, in that case the mission would be complete even if the AI planes keep flying straight. 

Stock Frequency : 4
Specific Frequency : 0
Severity : 4

Edited July 12, 2018 by Menrva

[+Gunrunner 314]

Environment :
July 2013 patch-level, SF2NA.
Only in Single Missions, on NavalMap.
In non-stock installs.

Conditions :
If using a land-based plane.
If the enemy is the side with a carrier group.
Rarely.
On any mission type.
Might only happen with some planes.
Possibly related to improperly configured INIs, but no common error found in terrains and planes exhibiting this bug.

Expected result :
The engine generates a mission of the chosen type, for the chosen aircraft type.

Result :
Randomly, instead of generating the expected player mission, the player will be assigned to one of the NAVAL_SEARCH missions generated, for his service,
even if the plane is non-player flyable (no cockpit available). The expected mission the player chose is not generated.

Stock Frequency : 0
Specific Frequency : 1
Severity : 4
Test environment : On demand

Potential fixes :
- Uncertain given the unclear conditions.

Menrva, if you are using the frequency and severity scale I suggested, you would be using 4 at max, as the bug doesn't happen every time, and it's only making it impossible to finish the mission, not crash the game, unless I misunderstood the bug.

Stratos, if you want this to pick up, I'd suggest contacting old timers and modders, who have been working around bugs for up to a decade now and probably have some very interesting bugs and ideas to contribute, consider also this not being simply for a future patch, but a catalog of known bugs for modders to know about, a part of the knowledge base to ease the life of modders, so they know what to avoid, get an idea how to avoid running into that bug or be reassured the problems comes from a known bug and not a flaw in their work.

Edited July 12, 2018 by Gunrunner

[+Menrva 4,200]

4 hours ago, Gunrunner said:

Menrva, if you are using the frequency and severity scale I suggested, you would be using 4 at max, as the bug doesn't happen every time, and it's only making it impossible to finish the mission, not crash the game, unless I misunderstood the bug.

Alright, fixed.

Although I don't think it matters that much at this point. I don't think we'll ever see interest from TK about all this. Just my point of view.

[Stratos 3,191]

7 hours ago, Gunrunner said:

Stratos, if you want this to pick up, I'd suggest contacting old timers and modders, who have been working around bugs for up to a decade now and probably have some very interesting bugs and ideas to contribute, consider also this not being simply for a future patch, but a catalog of known bugs for modders to know about, a part of the knowledge base to ease the life of modders, so they know what to avoid, get an idea how to avoid running into that bug or be reassured the problems comes from a known bug and not a flaw in their work.

I'm sure everyone interested already seen this post. And I agree with Menrva, TK does not care anymore.

[+Sundowner 4,819]

I think that the lack of response to this topic is a sure sign that most of us believe it's a non starter, it's easy for any number of folks to say they would pay TK to get back on board.....but when it comes to actually coughing up the dough just how many would? I wouldn't.

While we think there my still be lots to be had from this game, TK certainly doesn't or he would never have dumped it in the first place.

 

 

[TempestII 262]

On 10/07/2018 at 11:40 PM, Menrva said:

Another game-breaking bug I want to point out is the following:

Environment :
July 2013 patch-level, SF2NA
In Single Missions, and in Campaigns
In stock installs

Conditions :
The player flight is assigned an escort mission.
The escorted flight reaches its waypoint for its ground attack.

Expected result :
After engagement, the escorted flight reaches the egress waypoint, prompting a mission successful for the player's mission.

Result :
For reasons still unknown, it happens from time to time that the escorted flight is flying straight instead of engaging its ground target, and it never heads back home. By never reaching the egress waypoint, the player's escort mission ends up in a loop, it can't be completed.
Sometimes, simply restarting the mission is enough to trigger the correct behaviour of AI planes. Still, the bug is frustrating and it should not happen. I don't recall it happened in SF1 games.
Another solution, if possible, would require that a "mission complete" statement is triggered when the AI planes fly over their ground target waypoint, instead of the egress waypoint. No matter what, in that case the mission would be complete even if the AI planes keep flying straight. 

Stock Frequency : 4
Specific Frequency : 0
Severity : 4

+1. This frustrates me too. 

 

I think that it's due to the weapon loadout of the strike package. Purely AI flights no longer seem to be able use LGBs/EOGBs so will not drop their ordnance on the target, leading to the flight continuing on until they get shot down, crash or reach the map's edge.

I've even seen it happen with "dumb" bombs when there's been more than one mark of bomb - IE if they're all carrying MK82s you should be fine but having both MK82s and MK83s on the aircraft too can lead to the same issue. 

(if I'm incorrect then feel free to correct me, I'm basing this off my own observations and several posts I've seen on this forum). 

[Stratos 3,191]

15 hours ago, Sundowner said:

I think that the lack of response to this topic is a sure sign that most of us believe it's a non starter, it's easy for any number of folks to say they would pay TK to get back on board.....but when it comes to actually coughing up the dough just how many would? I wouldn't.

While we think there my still be lots to be had from this game, TK certainly doesn't or he would never have dumped it in the first place.

 

 

And I knew that cause It happned before. Now that is certified, can we close the topic?

[+Menrva 4,200]

15 hours ago, TempestII said:

+1. This frustrates me too. 

I think that it's due to the weapon loadout of the strike package. Purely AI flights no longer seem to be able use LGBs/EOGBs so will not drop their ordnance on the target, leading to the flight continuing on until they get shot down, crash or reach the map's edge.

I've even seen it happen with "dumb" bombs when there's been more than one mark of bomb - IE if they're all carrying MK82s you should be fine but having both MK82s and MK83s on the aircraft too can lead to the same issue. 

(if I'm incorrect then feel free to correct me, I'm basing this off my own observations and several posts I've seen on this forum). 

Good to know! I didn't know this was the cause of such bug! I think it's rather important to keep in mind this bit of information, at least for us modders and SF2 enthusiasts. 

2 hours ago, Stratos said:

Now that is certified, can we close the topic?

I don't want to close it. TK won't react, but if there are more modders willing to share their infos on bugs, etc., a topic like this would be useful.

[+gerwin 451]

17 hours ago, Sundowner said:

I think that the lack of response to this topic is a sure sign that most of us believe it's a non starter, it's easy for any number of folks to say they would pay TK to get back on board.....but when it comes to actually coughing up the dough just how many would? I wouldn't.

While we think there my still be lots to be had from this game, TK certainly doesn't or he would never have dumped it in the first place.

 

 

I don't think dumped is the right word here. My take is that Third Wire either supports a game or not. There is no in between. And it is primarily because customers do not understand and do not accept half-support, but they surely understand unsupported. As soon as TW openly shows a limited support for the SF2 games, like a Windows 10 update, there will be customers complaining he does not support it enough.

Also I have some things I could write down here as bug-fix wishes, but I just get cought up in my memories of the last years of SF2 development, where TK wrote how things should be simpler instead of realistic. Game and not Sim. IIRC he at some point suggested automatic landing gear up instead of manual. Then there were many complaints about missile effectiveness in SF2NA which was addressed by making player missiles stronger then AI missiles in June 2012 level. I am on May 2012 level because of that.

I do not blame TW to have priorities that differ from my wishes. I think TW priorities were at that point not personal anyways, but about customers and sales. Or maybe they were personal a little, because the original course was just to hard for Devs. Modders and bolt and nut counters make very annoying customers already, but when beyond that there are not enough sales with increased costs , then one has to change course.

With that in mind I can only imagine some small patch for SF when it is totally clear that it does not imply much, and it is just for fun. But wheter or not it is fun for TK, no idea. It is not my business either.

 

[+gerwin 451]

Anyways, here are two bugs I would wish  to be gone:

1) A terrain with navalmap=true where nation 001 can access an Aircraft Carrier. The enemies of this nation 001 have only one Strike objective in single missions: The carrier.  Nothing else.

2) In June 2012 player missiles were made more damaging compared to the same missiles laucnhed by AI craft: There should be some line in the options.ini to get rid of this 'cheat'.

[+Sundowner 4,819]

Just maybe a little off topic, but has anyone ever compiled a list of what we lost and gained at each patch level?

[+Menrva 4,200]

8 minutes ago, Sundowner said:

Just maybe a little off topic, but has anyone ever compiled a list of what we lost and gained at each patch level?

It's much on topic, I'd say. Something like this could be very useful. To get our hands on all different .dll files of each patch level (in order to investigate bugs and changes) is another story, though.

[KJakker 901]

Ground objects do not recognize when an aircraft target has been destroyed. I have seen ships waste six or more missiles on the falling fuselage of an aircraft that has already been destroyed until it hits the ground or denigrates. As a result this lets the rest of the engaged flight approach unmolested.

[+gerwin 451]

1 hour ago, Menrva said:

It's much on topic, I'd say. Something like this could be very useful. To get our hands on all different .dll files of each patch level (in order to investigate bugs and changes) is another story, though.

There was the above topic, but it does not hold much info.

It is hard to consider doing without the SF2NA Naval warfare goodies, These are great fun!

( Even considering SF2NA was never 100% Finished + never 100% Functionally Integrated with the other games. IIRC TK once advised not to merge SF2NA like the other games. Unfinished I mean: There are hardly distance lods for the new objects, so framerate can suffer. That would be most noticable with parked planes, but SF2NA itself does not have a problem with this, since the big RED airbases are off-map. Then there are 'Jammer' plane types that do nothing, and the criticised new terrain... )

Maybe older game versions can be interesting for optimal gameplay on terrains without Naval warfare... I don't know the particulars.

Edited July 13, 2018 by gerwin

[+gerwin 451]

On 13-7-2018 at 1:07 AM, TempestII said:

I think that it's due to the weapon loadout of the strike package. Purely AI flights no longer seem to be able use LGBs/EOGBs so will not drop their ordnance on the target, leading to the flight continuing on until they get shot down, crash or reach the map's edge.

I've even seen it happen with "dumb" bombs when there's been more than one mark of bomb - IE if they're all carrying MK82s you should be fine but having both MK82s and MK83s on the aircraft too can lead to the same issue. 

(if I'm incorrect then feel free to correct me, I'm basing this off my own observations and several posts I've seen on this forum). 

Good point! I had not really noticed before, but just flew some SEAD mission with an AV-8B and one wingman. If my wingman has a Rockeye+AGM-65 loadout, he at some point keeps making dives to targets without releasing the remaining stores, neither will he use guns. Make it a rockeye only loadout and he behaves properly + uses his guns when out of stores... This is not even a pure AI flight as mentioned above.

Edit: Tried a similar setup but with the A-10: Wingman has no problems using different types of weapons.

Edited July 13, 2018 by gerwin

[TempestII 262]

3 hours ago, gerwin said:

Good point! I had not really noticed before, but just flew some SEAD mission with an AV-8B and one wingman. If my wingman has a Rockeye+AGM-65 loadout, he at some point keeps making dives to targets without releasing the remaining stores, neither will he use guns. Make it a rockeye only loadout and he behaves properly + uses his guns when out of stores... This is not even a pure AI flight as mentioned above.

Edit: Tried a similar setup but with the A-10: Wingman has no problems using different types of weapons.

I think the EOGR issues are due to AI data such as release speeds, altitude etc but I haven't done enough testing to say for certain. I plan to do some experimenting this weekend. 

Like you, I've never had issues with A-10s due to its low & slow attack runs (which is why I assume it's such a good CAS asset in real life) with AGMs. It'll fire all 6 or so Mavericks and then start using any other weapons it has. But fast air like F-15s, F-16s, F/A-18s etc might fire one missile (if you're lucky) and then just orbit over the battlefield if on CAS or fly off into the sunrise if on a strike. 

[fallout3 12]

Wonder if trimming of the three axle could be made possible, the autotrim we have now feel kinda of...Strange, no need to step on the ball while turning.

Also, wish AI could have two different ways of handling the aircraft in and out of engagment, have been playing around with The Last of Lightning missions, AI fly in a very peculiar way and it is not possible to fly close formation with an AI fighter even in a simple pattern.

Better Navaid and Weather Maybe?

Regards.