3.0 Bugs

[Steve Walmsley]

I am a complete newbie but on my first installation of 3beta while trying to find my way around in the pop and production screen, research - avaiable projects I clicked on reduced-size lazer/20xrecharge, the descriptor then refers to missile tech, not lazers!

Only a minor bug, assume its a database prob?

Yes, cut and paste problem . Corrected for v3.1

Steve

[Steve Walmsley]

Another little bug.  Actually two.

The first is related to the Political Stability Modifier.  The modifier is dropping on my home planet, in spite of the fact that the Population & Production Summary tab shows an adequate level of protection.  The Requested protection level is 1020, while the Actual Protection level is 1330.  The Actual Protection level was 1000 before I appointed a fleet commander with a cracker-jack public relations officer that boosted the actual protection level to 1330.  I suspect that Aurora is still using the original protection level.  

Just to check, do you mean a Task Force when you mentioned the public relations officer?

Steve

[Kurt]

Another little bug.  Actually two.

The first is related to the Political Stability Modifier.  The modifier is dropping on my home planet, in spite of the fact that the Population & Production Summary tab shows an adequate level of protection.  The Requested protection level is 1020, while the Actual Protection level is 1330.  The Actual Protection level was 1000 before I appointed a fleet commander with a cracker-jack public relations officer that boosted the actual protection level to 1330.  I suspect that Aurora is still using the original protection level.  

Just to check, do you mean a Task Force when you mentioned the public relations officer?

Steve

Yes, I was referring to the task force to which the planetary defense PDC's were assigned.  

I ended up taking the PDC's out of the task force they were in, and putting them in a new task force with no officers assigned.  That way the actual protection level went down to the amount provided by the PDC's, with no officer bonus.  I then built more PDC's to boost the actual protection level, and this problem went away.  I think there was something wonky with the way the pulbic relations officer assigned to the task group that the PDC's were originally assigned to boosted the actual protection factor.  

This isn't a fatal error, though, just a minor annoyance.

Kurt

[sloanjh]

I'm getting a divide-by-zero in PopulatePopGU.  I think it's because I've got a colony world with engineers but no factories.  It hits whenever I make the colony active in the F2 screen, or when I hit a 5-day update while the colony is active (I think).  Doesn't seem to be causing any harm - just something you probably want to clean up.

If you had no actual population (i.e. people) either then I found the bug :-)

John

PS - I'm pretty sure the colony was auto-generated when I gave an order to a TG to drop automated mines there.

[sloanjh]

I'm getting a path not found on "Flags" when I open the galactic map (F11).  I do have a "Flags" directory in my install directory (I'm using the same directory that I did for 2.4)

It sounds like a particular flag is missing from the directory. If you can figure out which race is causing the problem (by seeing which flag doesn't appear), see what happens if you open the race window. Its possible that if you have some flags missing from the directory, the create race code setup a flag that doesn't exist.

Yep - that was it.  "My" race didn't have a flag.  I thought I dropped the last set of new flags in, however.

John

[sloanjh]

Not sure if this is broken or working as intended, but it looks like auto-assign refuses to assign talentless officers to ships.  I have ships without commanders (cheap pickets that only require R1) and commanders without assignment, and the auto-assign now button doesn't change this.  I suspect that if the "score" for the commander taking up the assignment is zero, i.e. if the training rating is 0, then the commander isn't assigned.

That is working as intended. I didn't want talentless officers taking up a post for a whole tour when new officers might be more suitable.

Aha!  Not a high priority request, but could you make this an optional behavior?  I like to stick talentless officers (especially those with good Governor skills) in a command slot to give them the opportunity to acquire on-the-job training and become talented.

I've also found that I've been overriding the auto-assign (which I love, btw) for important command slots like warships or ground unit HQ.  If a new-construction cruiser comes along, I'll look for a DD commander to promote (who's more talented than the auto-assign), even if his or her tour isn't over.  I'll then promote someone else into the DD hole, etc.  One advantage to this is to breaks up the huge "start time + 2 years" cohort of tour expirations present at start-up.

One side effect of disallowing talented commanders that I think I've noticed (at least early on) - freighters soak up all the talented officers in the initial assingments at startup, so new warship construction ends up without a commander (hence my promotion strategy).  This isn't a big deal for me, but here are some ideas if you want to use them:

1)  Have a "garbage scow" checkbox for each class on the F5 screen that allows talentless officers to be assigned (or you could just use the existing "conscript crew" flag)
2)  Have a numerical rating for each class that is set on the F5 screen that  indicates the prestige of the command.  A very talented officer in a low-prestige command would be considered (possibly with a "breaking tour" penalty) for higher-prestige commands when they come available.  this is essentially what I'm doing by hand.  For example, the score of an officer in a command could be multiplied by (1 - prestige_difference/higher_prestige) to represent breaking tour, which would disallow purely horizontal "promotions".  This could also be extended to staff positions, a lot of which are irrelevant (fighter ops for logistics command?) but are still sucking up highly qualified officers.

Like I said, though, I'm fine if you don't do any of this.
John

[Charles Fox]

Getting a bug in 3.0 where I can't design any ship classes. I select the pulldown tab for "Type" and try to select "Ship", but that leads to an error box popping up titled "Error in cbo_Type_Click" with the following text: "Error 381 was generated by Aurora. Invalid Property Array Index." That box is followed by another reading "Error 3078 was generated by DAO.database. The Microsoft Jet Database engine cannot find the input table or query 'and ComponentID IN (Select SDComponentID from ShipDesignComponents where TypeAllowed =2)' Make sure it exists and the name is spelled properly."

I tried deleting and reinstalling Aurora, and made sure that my VB runtime was fully updated. Any other suggestions?

[sloanjh]

Getting a bug in 3.0 where I can't design any ship classes. I select the pulldown tab for "Type" and try to select "Ship", but that leads to an error box popping up titled "Error in cbo_Type_Click" with the following text: "Error 381 was generated by Aurora. Invalid Property Array Index." That box is followed by another reading "Error 3078 was generated by DAO.database. The Microsoft Jet Database engine cannot find the input table or query 'and ComponentID IN (Select SDComponentID from ShipDesignComponents where TypeAllowed =2)' Make sure it exists and the name is spelled properly."

I tried deleting and reinstalling Aurora, and made sure that my VB runtime was fully updated. Any other suggestions?

Did you hit the "New" button on the F5 screeen yet?  When the game first starts, IIRC, there are zero classes in the DB and you get these kinds of errors.

John

[Hagar]

The Technology Report screen (CNTL-F7) does not have Shields, Thermal Sensors, or EM Detection Sensors available in the Category pull down list.

[Charles Fox]

Did you hit the "New" button on the F5 screeen yet?  When the game first starts, IIRC, there are zero classes in the DB and you get these kinds of errors.

John

That did the trick. On the other hand, now I'm getting overrun errors after a few time intervals. Those can probably be traced back to my hardware.  

[sloanjh]

now I'm getting overrun errors after a few time intervals. Those can probably be traced back to my hardware.  :-)

John

[sloanjh]

I have noticed the occasional "zip across the system" problem as well but unfortunately because it's intermittent it's very difficult to pin it down. I will keep looking.

That's odd - it was fairly reproducible for me.  It seems like whenever there's a big asteroid belt it kicks in (at least that's where I noticed it), and it fixates on the ones it wants to zip about to.  It's acting like the sort routine is ignoring the position of the ship after the 1st nearest body is calculated.  If I clear the bad orders, survey the nearest, then let it look for the next "5 nearest", it finds a new (presumably correct) "nearest" to survey first, but 2-4 are the same ones it wanted to hit before.

John

[Charles Fox]

Do you have any Task Groups with movement orders and 0 or 1 velocity?  That's usually where I get overflows (if that's what you mean by "overrun") - it takes a looooooooong time to get anywhere when you're creeping along at 2000 mph.

John

Nope, no Task Groups, or any ships at all for that matter. It happens about 20 - 30 intervals in, before I have a chance to do too much at all.

[ZimRathbone]

I have noticed the occasional "zip across the system" problem as well but unfortunately because it's intermittent it's very difficult to pin it down. I will keep looking.

That's odd - it was fairly reproducible for me.  It seems like whenever there's a big asteroid belt it kicks in (at least that's where I noticed it), and it fixates on the ones it wants to zip about to.  It's acting like the sort routine is ignoring the position of the ship after the 1st nearest body is calculated.  If I clear the bad orders, survey the nearest, then let it look for the next "5 nearest", it finds a new (presumably correct) "nearest" to survey first, but 2-4 are the same ones it wanted to hit before.

John

Whats probably hapenning is that the code is evaluating the 5 nearest to the start position of the TG, in one query.  These may well be in different directions from the origin point, resulting in the zigzag across the system.

What needs to happen is for the routine to evaluate the 2nd nearest using the origin point of the 1st, then the 3rd using the origin point of the 2nd and so on.  This will probably be prohibitive in processing (5 queries instead of one).

I suspect that the real solution is not to use the "next 5", and have each unit evaluate its next target as it finishes the current one.  Note that if you do this you will tend to have a wave of surveyors moveing through the system, all of them turning up at one planet and surveying 1 or 2 moons then all moving to the next, rather than send one surveyor to each planet, who then surveys the planet & moons, this increasing the efficiency of the survey (which is what I suspect Steve was tring to do with the "next 5" idea.)

To manage this I usually have two Default  orders for large survey groups  - 1 survey nearest planet, 2 survey nearest body.  For singletons I usually manually assign the most promising planet and then 1 survey nearest planet or moon , 2 survey nearest body.

It gives the best avaliable coverage without manually assigning everything

Slainthe

Zim

[James Patten]

I've noticed that when giving a task group cyclic orders that take less than the amount of time you're advancing, that you get a report at the end of the time period that the TG has completed its orders, and it does not repeat the orders until the next time period.

IE: I have freighters shipping something which takes 4 days to do, and I click on the 5 day button - or a task that takes 20 days to do and I click the 30 day button.  The freighters complete the task and wait for the next time period before resuming.