Seeing absolute silence since the open beta started isn't my idea of an improvement

 

So I'll help you out a little bit, there is a problem (2009.007) re-assigning aircraft missions (probably any intercept ship/sub/air). Once a plane group is told to intercept a certain group it is very difficult in many cases to assign that plane group to attack different target or otherwise change tasks.

 

What is needed:

1. Test scenario of the situation (small, tiny,...)

2. Saved game with planes in the air ready to have a new target assigned.

3. Instructions on recreating the issue.

4. The above three items summarized in a single report at the Issue Tracker, http://harpgamer.com/harpforum/index.php?autocom=bugtracker

5. Once the report exists, someone needs to verify the report and add any additional information. Once confirmed it becomes an actionable item for me when I have my programmer hat on.

 

 

Let's give it a go folks. Thank you

Is this a test for the Open-Beta-Potential-Recruits or can the Old China Hands jump in too?

Is this a test for the Open-Beta-Potential-Recruits or can the Old China Hands jump in too?

 

We'd like to hear from the new people. They have to start somewhere.

But in the interests of time let's now open it up to the old hands as well

addressed in 2009.008, too slow folks

Just in case those new to the testing procedure are still unsure about how to go about it, I'd like to present the basics of how I test a bug, step by step, in a series of posts. This post will deal with creating a Test Scenario. (Tony, I'll probably be repeating a lot of what you've already said, but I like to be thorough.)

 

Part One: Writing A Test Scenario

 

A Test Scenario is created to try to replicate a bug that you or someone else has found and reported. A Game Save from a Test Scenario is preferable to a Game Save from a Combat Scenario because a Test Scenario can be stripped down to the bare minimum to test the bug, while a Combat Scenario usually has so much action going on that it may be more difficult for the house programmers who have to repair the bug to run the problem down. The Test Scenario will be created using the Scenario Editor.

 

In this case, the problem concerns a difficulty in Beta2009.007 with making aircraft that are an Intercept mission change to attack a new target . To cover all the bases, I eventually decided the minimum I needed was a Blue airbase with a dozen multipurpose aircraft, a Red airfield with at least 2 aircraft, two bases and two ships. Also, I prefer to have enemy targets that don't shoot back to keep things less complicated.

 

Open the Scenario Editor. I tend to use the EC2003 battleset as my testbed as I gotten more familiar with it. I used the arrow keys alongside the Group map to put Iceland in the center of the screen. Now click Groups in the options line and Create New Base. Choose the Blue side and go down the base list to "Keflavik, Iceland, Blue". After Keflavik pops onto the map, scrolling down the Edit Airfield list allowed me to select the F-4S Phantom as my multipurpose aircraft and to add 12 of them to the base. (Screenshot 01 shows the process to this point).

I gave them a 100% chance of inclusion and closed the Edit Aircraft box and the base selection box. To cut down on the waiting time to find the enemy, I also like to include an AWACS, so I added an E-3 Sentry RSIP to the base. If everything worked, clicking the Blue airfield shows that it is BLUE Airfield, AAa at Keflavik, Iceland with 13 Planes. As a final step, I clicked the Ready Air and changed the loadout on 8 of the F4s from Intercept to AAmr.

 

Now you need targets. I figured that I would need 2 bases, 2 ships and an airbase with at least 2 aircraft so the the attacking Blue aircraft have targets to switch off against. Using the same procedures as for Blue, I created Red Land targets at Crossroad Echo, Iceland; Antenna Farm, Iceland; and Dispersement 1, Iceland. At Dispersement 1, add 2 Russian Mi-24K Hind G2 helos.

 

My live air targets will be those 2 Hinds. To get them airborne, I clicked Dispersement 1, then Launch Air. I launched one helo to Patrol about 60nm due west of the airfield, and the second to patrol about 60nm due east. I selected a Delta 1 minute launch for both.

 

Finally, my Sea targets are a couple of Russian Okean intelligence vessels. I posted them separately about 60nm west and northwest of Keflavik.

(Screenshot 02 shows everything done up to this point.)

 

Finally, I clicked File, then Enter Game Limit. 1 Day will be more then enough to complete this test.

 

As far as I can figure, this scenario now contains everything needed to test the fix. To save it, click File one last time, then Save Scenario As. In the Save User Scenario As screen, select a Folder the scenario will be saved to. Under File name, replace the * with a filename of no more than 8 characters; it will automatically be saved as an EC2003 battleset. My file is named "INTTEST".

 

My completed test scenario is attached here for documentation purposes. If you wish, print out this section and uses it a a step-by-step guide for using the Scenario Editor to duplicate it.

INTTEST.zip

Well, Wombat, I for one learned something here. I never knew I could get away with skipping the ViCond's headaches when writing a scenario strictly for bug testing purposes.

 

Thanx,

 

Buddha

Part Two: Working Through The Testing Process.

 

"...there is a problem (2009.007) re-assigning aircraft missions (probably any intercept ship/sub/air). Once a plane group is told to intercept a certain group it is very difficult in many cases to assign that plane group to attack different target or otherwise change tasks."

 

This was the task at hand that lead to the creation of the Test Scenario in Part One. Now I'm going to use that Test Scenario to try to generate some Saved Games that catch the bug in the act.

 

The most common way to dispatch aircraft is to use the Launch button. Pressing this at any airbase gives you the option to Attack a target; to Patrol at a certain point; or to Ferry aircraft from one place to another. After thinking through the process, I decided the best way to approach the problem was to a) launch an Attack mission against the helos and see if I could switch targets between them or Land and Sea targets; Launch a Patrol mission and try to attack air, sea or land targets; c) launch a Ferry mission and try to attack air sea, or land targets.

 

AH HA! The more astute are saying. How can you Ferry aircraft when you only have one base. Double AH HA! Where are the subs? Answer one: I screwed up. I forgot all about the ferry option until I actually started testing, so I modified INTTEST into INTTEST2 to add another. Answer two: I personally find it difficult to test anything involving subs. My pious hope was that if the fix works for Land, Sea and Air targets that it will also work for Subs. If not, I'll deal with it then.

 

So, let the testing begin. After selecting the EC2003 GIUK battleset, load the User Scenario INTTEST2.SCm and up it comes. Before you do another thing, though, you have to prepare the board. First, under Settings, then Game Options, Set the Save Every to 0 minutes. This allows you to save the game at the point you want without it being automatically overwritten during the testing process. Second, use the ALT-I key combination to label all the icons on the Group and Unit maps. This makes tracking everything so much easier. Finally, get the AWACS airborne with radar active. All the Red targets, including the helos in about a minute, will be visible. Since this is going to be the baseline from which all the tests will be run, I saved this as SETUP.hpm so I can open it again and again as needed. Screenshot 3 shows this in operation.

 

Now I'll start running my first series of tests. I'm going to Launch 2 groups of 2 F4S with AArm loadouts from Blue airfield AAa. One group will Attack Red helo ZTH; the other, Red helo ZUH. En route, I will try to get them to switch targets with each other. If they do or they don't, I will try to get them to Attack the Land targets at ZZb and ZYb. If they do or they don't, I will finally try to get them to Attack the Sea targets at ZVS and ZWS. I'll run the test at low time compression, no more than 10 or possible 30 seconds so as to try not to miss anything.

 

Screenshot 4 shows this test underway.

ADA is to attack ZTH and AEA is to attack ZUH. (This is a good place to make a save because the test in underway in its initial stages and can be opened again and again as needed. Having the aircraft up and assigned also cuts down on the time that the programmers have to spend on unnecessary fiddling around. This test was saved as Air2Air1.hpm.) As the Phantoms flew, I tried several times to get them to switch off on their assigned targets. Evertime I tried this, they complied without problems. Then I tried to get ADA to Attack Land target ZZb and AEA to Attack ZYb. This time, they did not comply no matter how hard I tried to get them to attack either target. I then tried to get them to go after the Sea targets. Again, nothing would dissuade them from their single-minded pursuit of the helos. I let them have their way and destroy the helos. Surely now, I thought, they would let me assign them to Land or Sea attacks, but surprisingly, they would not. ADA would obey no order but Land Air. AEA, though, still with AAM and a target available was willing to attack it (as Screenshot 5 shows), after which it too would only obey the Land Air order to either Blue airbase.

 

By now, I'm satisfied that there is a bug and that I'm fairly confident that I have at least some basic information to work with. Now I'll start running whatever variations on problem I can think of on the theory that it's better to have too much than too little to work with. Rather than bore you all to tears, I'll just end at this point as they are all going to be summarized in the next installment, when I write up the formal bug report. If you can't wait, both INTTEST2 and SETUP are contained in the .zip file.

Part_2.zip

Thanks Wombat. Very interesting, and echoing Buddha, I didn't know you could so easily dispense with the VC's. Screwing around with VC's has been the long pole in my scenario making tent all along.

A serendipitous discovery on my part.

I'm still not sure I can get my VCs to work properly...

Interesting problem.

 

I have never come across this before.

 

Most of you know by now that I have spent most of my time with Harpoon with Harpoon Classic, yada, yada. However, I did test drive the demo of HCE and I did not notice this problem.

 

Now the following are anecdotes, but they are accurate and, I believe, repeatable.

 

Example: WestPac 3 (Philippines).

 

My platform is the F-35B Lightning II. It carries a variety of different weapons, but for present purposes it is loaded with some number of AMRAAMs plus some number of anti-surface weapons, such as 2 ARMs plus 2 bombs. I typically used every weapon before rearming the plane, which is to say that I would shoot down a couple of aircraft, launch the ARMs at ships that had SAMs, and bomb some freighters. The chronological order of the attacks was not relevant. There would always be at least 2 target groups involved, and sometimes as many as 6 different target units. The point is that I never had any difficulty in attacking multiple targets, although I generally shot at one target before acquiring another (i.e. I didn't assign the plane to some target and then change my mind).

 

I typically launch with "patrol" orders, because I typically use groups with only one or 2 units, and the "patrol" default is one unit, while the "attack" default is all units of that type, so the former means I have to click the mouse less. But I also launch with "attack" orders, and get no problems that way either.

 

More details are available in my AAR in the Matrix forum http://www.matrixgames.com/forums/tm.asp?m=2002545

 

Example: WestPac 2 (Pirate War).

 

My platform is an attack helicopter (various types). It is carrying some number of anti-surface weapons. The helicopter would engage one target after another, ending only when it ran out of weapons. Generally, all targets would be in the same group.

 

More details are available in my AAR in the Matrix forum http://www.matrixgames.com/forums/tm.asp?m=1998012

 

Question: Could the problem have anything to do with the specific weapons the Phantoms are carrying? For example, the anti-armor loadout may not include weapons that can be shot at ships or land bases.

 

Question: Can you try clearing the course (F3 or CTRL-3) before you send the aircraft after a new target?

First, under Settings, then Game Options, Set the Save Every to 0 minutes. This allows you to save the game at the point you want without it being automatically overwritten during the testing process.

 

I was looking for this (turning off the autosave). I had interpreted the "save every 0 minutes" option as "save with a time interval of 0", which is what all the other options mean, i.e. save continuously.

 

Now I know.

If you're using .008 or .009, the problem should not, of course, be happening as it was actually fixed there. (The problem hadn't occurred until the earlier 2009 betas probably as an unintended consequence of trying to fix something else.) That it was definitely broken in .007 made it a good basis for the how-to and why-did posts.

AArm is a favorite test loadout for me because the aircraft weapons are usually some long-range AAM and the Mavericks, which can be used against either Land or Sea targets.

Clearing the course actually didn't occur to me. I was too focused on trying to switch targets on the fly. I'll have to see if that works and amend the post.

I also forgot to mention that sometime in the past the Computer Gods would take the "Save Every 0 Minutes" and do a continual save. I haven't seen this lately. If it should happen again and the ritual sacrifices do not placate them, "Save Every 60 Minutes" gives you more than enough time to find a good save point.

Happy Bug Hunting.

Part Three: Preparing the Bug Report

 

Following a mind bout of flu, I completed the testing phase begun in Part Two, including the suggestion that Victor in the Pacific made, and ended with a huge amount of information. The final task is figure out how much of it to use and then to organize it into useful form.

 

Over time, a standard report format has evolved. Here it is with all the blanks filled in, after which it will be explained:

 

Defect Name: Retargeting Problems

Build: 2009/007

Repeatable: Yes

Operating System: Windows Vista SP1

DB used: HCDB-090128

Scenario used: Test Scenario "INTTEST2.SCm" for EC2003 GIUK

Long Description: Launch Air attacks cannot be changed to Land or Sea attacks.

 

Sequence to reproduce: 1) Open and run saved game "Air2Air" in EC2003 GIUK battleset.

2) Blue Plane group ADA was launched to attack Red Helo group ZTH; AEA was launched to attack ZUH.

3) Click ADA and try to have it attack any of the three Land targets. Repeat with AEA.

4) Click ADA and try to have it attack either Sea target. Repeat with AEA.

 

Expected behaviour: ADA or AEA should follow orders to attack the Land or Sea targets.

Observed behaviour: ADA or AEA cannot be changed to Land or Sea attacks once they have been ordered to attack aircraft.

 

Additional information: 1)ADA and AEA can repeatedly be switched between either Helo target.

2) Attempting to Clear the course will change the target to None on the Unit screen (see "Screenshot.jpg"), but you still can't attack Sea or Land targets.

3) Launch Patrol and Launch Ferry missions also cannot be diverted once assigned.

 

Attached File(s): TestFiles.zip

 

 

Defect Name: You have to call it something; something self-explanatory is best.

Build: Which beta is being tested.

Repeatable: Hopefully Yes, because a non-repeatable bug is that much harder to track.

Operating System: It is possible that the Computer Gods will let something misbehave in one OS and not another.

DB Used: Some problems are database-specific. Though most players will be using the most current version of the standard Harpoon database (currently HCDB-090128), others will use one of the speciality databases.

Scenario used: As mentioned earlier, a specially-written test scenario is preferred. For ease of use, also state the battleset it runs with.

Long description: Long enough to get the point across.

 

Sequence to reproduce: Here's where you can go into detail about getting the bug to show. Be very specific: The more information you can provide, the more

helpful it is for bug hunters.

 

Expected behavior: What you think is supposed to happen.

Observed behavior: what actually did happen.

 

Additional information: Any items that could be of possible use that didn't fit anywhere else, including all that extra information. Strictly speaking, the Launch Patrol and Launch Ferry mentions should be split off into two seperate bug reports, but again I hoped that fixing one would also fix the rest. If not, the issue can be taken up later.

 

Downloads: All the files necessary to demonstrate the problem have to be copied into a .zip file and uploaded to the post. In this attachment are the

original test scenario INTTEST2 (just in case); InitialSetUp (the first save with the AWACS up); Air2Air (the save with the F4s launched and targeted); and

Screenshot (showing the Clear course failure).

 

I'm pretty much satisfied with the results of my efforts. This all now would get wrapped up into a post to the Issue Tracker section (if it wasn't a

tutorial for an already solved problem) where an interested party will pick up on it and either confirm or debunk the findings.

TestFiles.zip