HoF Map Finder/Generator Utility 2016-10-05

[Moonsinger]

I've had a similar but slightly different idea about how the columns can be narrowed. The columns are much wider than they really need to be, mainly due to the width of the text in the header row. How about replacing the text with icons that have the names as their alternative text. I've produced an example as follows :

You'll just have to imagine the cursor hovering over the sugar icon.

That's a cool idea, Dianthus! I have the new version ready, but I will delay it until I include your idea in the report (soon).

 

[Tzar Sasha]

I've had a similar but slightly different idea about how the columns can be narrowed. The columns are much wider than they really need to be, mainly due to the width of the text in the header row. How about replacing the text with icons that have the names as their alternative text. I've produced an example as follows :

You'll just have to imagine the cursor hovering over the sugar icon.

I like the idea. It should shorten up the width quite a bit. Could you, however, test it on a screen res of 1024x768? If there is still a sidewise scrollbar on that test, could you please add a 2nd iteration# column on the right side?
Thanks....

 

[Moonsinger]

could you please add a 2nd iteration# column on the right side?

This obsession with the iteration# has to be stopped. I will post the new version in a few minutes (with a 2nd iteration# as requested).

 

[Moonsinger]

Update: MapFinder 2005v2.3

This baby will now do "lake" as we have discussed previously and some new stuffs.

 

[Dianthus]

Now, that's strange! Since I use only static variables (no pointer, link list, or dynamic array of any type), I don't really see how it could possibly running out of storage memory.

I missed telling you about a symptom. I checked Task Manager while Map Finder/Civ were still running and I still had 450Mb of physical memory free. This was at the point I could even paste into Paint.

Of course, God know what may be going on in the JPEG compression module...I will have to study this problem further. In the mean time, would you try to increase the Time Delay in the Extra setup option to see if it solves the problem.

Sure, I'll try that. I'm just about to start MapFinder before I go to work.

 

[Moonsinger]

I missed telling you about a symptom. I checked Task Manager while Map Finder/Civ were still running and I still had 450Mb of physical memory free. This was at the point I could even paste into Paint.

You are probably right! I will check the ScreenCapture subroutine...Just curious, what is the speed of your processor?

Since it takes the screen shot ONLY for acceptable maps, that may explain the different between 700 iterations the first round and 400 iterations the second round. Would you increase the delay time for saving by a couple more seconds (since it also saves ONLY acceptable maps, that won't slow anything down).

//Edit: I forgot to ask...was it with C3C or PTW?

 

[Dianthus]

I think I set the times to 1 or 2 seconds, I'll let you know when I get home from work. My PC is a 2.25GHz P4, no hyper threading. 1Gb memory. I'm running C3C 1.22 on XP Pro.

 

[Moonsinger]

I think I set the times to 1 or 2 seconds, I'll let you know when I get home from work. My PC is a 2.25GHz P4, no hyper threading. 1Gb memory. I'm running C3C 1.22 on XP Pro.

I think your setting is a little bit too fast. I have run MapFinder with over 5000 iterations on one shot without any problem before. In fact, I'm currently running the latest version at home right now (will see how far it will go). My system is actually a P4 2.8 GHz with 1Gig of PC4400 on XP Pro. I usually overlocked it to around 3.4 GHz, but when I'm not home, let set it back to normal at 2.8 GHz.

 

[Dianthus]

I've just realised I haven't reported on MapFinders last run! MapFinder was still running when I got in from work. Unfortunately C3C wasn't . It had found 36 good maps and rejected 2489 though, so it processed a lot more than previously.

My settings for that run were:
Time (ms) delay for Loading Game: 8000
Time (ms) delay for Saving Game: 3000
Restart C3C after every 50 interations
Time (ms) delay before restarting C3C: 15000

Time (ms) delay/sleep between keystrokes: 500
Time (ms) delay/sleep between Civ Dialogs: 2000

Note that I'm generating Tiny maps, and those times for Loading/Saving the game are WAY higher than needed. I set them high to try and get it running for longer.

 

[Dianthus]

I like the idea. It should shorten up the width quite a bit. Could you, however, test it on a screen res of 1024x768? If there is still a sidewise scrollbar on that test, could you please add a 2nd iteration# column on the right side?
Thanks....

I've created an example doc using images for all except the 1st 4 column headers. It fits in 1024 on my computer. I've also made it repeat the header every 20 rows, and I can show the header+20 rows with 768 height, so allowing the header to be seen no matter where the scrollbar is. Take a look and see what you think, it's <here>

Feel free to use it Moonsinger. Please . You might be interested in the html as I changed it a bit to make it smaller, just couldn't help myself. I'm such a sad geek!

 

[Moonsinger]

I've just realised I haven't reported on MapFinders last run! MapFinder was still running when I got in from work. Unfortunately C3C wasn't . It had found 36 good maps and rejected 2489 though, so it processed a lot more than previously.

My settings for that run were:
Time (ms) delay for Loading Game: 8000
Time (ms) delay for Saving Game: 3000
Restart C3C after every 50 interations
Time (ms) delay before restarting C3C: 15000

Time (ms) delay/sleep between keystrokes: 500
Time (ms) delay/sleep between Civ Dialogs: 2000

Note that I'm generating Tiny maps, and those times for Loading/Saving the game are WAY higher than needed. I set them high to try and get it running for longer.

That's good! Your settings seem to be alright; however, if it's still giving you trouble, try this:

1. Restart C3C after every 30 ilterations. This would virtually eliminate all errors that may cause by C3C.

2. Increase the Time (ms) delay/sleep between Civ Dialogs to 3000.

3. Increase Time (ms) delay for Saving Game to 9000. Since it would save ONLY when it finds acceptable map, no harm in pummping that number up.

That's it...I'm sure (99.99% sure) that you should be able to run MapFinder none-stop "forever" or for at least a week without any problem. I have done further testing on individual subroutines and found absolutely no bug of any type. Therefore, I'm 99.99% sure that MapFinder is bug free at the moment.

PS: After I get home tonight (about 18 more hours to go), I will change the format of the report to what you have suggested. Thanks a lot for your hard work!

 

[Moonsinger]

Update: MapFinder 2005v2.4

Changed the report style to what Dianthus had suggested. FYI, you can re-create your previous report by using "Generate Report" from the File menu.

 

[Moonsinger]

I've created an example doc using images for all except the 1st 4 column headers. It fits in 1024 on my computer. I've also made it repeat the header every 20 rows, and I can show the header+20 rows with 768 height, so allowing the header to be seen no matter where the scrollbar is. Take a look and see what you think, it's <here>

Feel free to use it Moonsinger. Please . You might be interested in the html as I changed it a bit to make it smaller, just couldn't help myself. I'm such a sad geek!

I thought you can see more than 25 rows with 1024x766 resolution, so I set the column headers to be repeated in every 25 rows instead. Please let me know if you can see it ok with 25 rows (for some reason, I thought 25 was a cool number). Sorry, It was too early and I didn't see the number "20" on your post before I posted the update.

 

[Dianthus]

I thought you can see more than 25 rows with 1024x766 resolution, so I set the column headers to be repeated in every 25 rows instead. Please let me know if you can see it ok with 25 rows (for some reason, I thought 25 was a cool number). Sorry, It was too early and I didn't see the number "20" on your post before I posted the update.

I'm sure 25 rows will be OK on my 1600x1200 . I'll leave it to Tzar Sasha to shout if it doesn't fit in 1024x768.

 

[Moonsinger]

Around noon yesterday, I set MapFinder 2005v2.3 to run on my old computer here in the office. It's just a P3 1GHz with 512K of RAM, WinXP with C3C 1.22. My Time Delay for Loading was at 13000ms, Time Delay for Saving at 15000ms, C3C to restart at 30 iterations with a restart delay of 16000ms, keyboard delay at 500ms and dialogs delay at 5000ms. My criteria was for huge empire map with minimum domination limit of 4400, 1 source of fresh water and at least 1 food bonus. It has been running non-stop for almost 20 hours now and I'm going to let it run through the weekend...this version seem extremely stable....probably becuase I rebuilt, redesigned, and recoded its main interface from scratch. Anyway, I will post its result on Monday. So far, it has rejected about 4500 maps and keep 20.

 

[Tzar Sasha]

I'm sure 25 rows will be OK on my 1600x1200 . I'll leave it to Tzar Sasha to shout if it doesn't fit in 1024x768.

Personally, I liked the 20. I had to remove some toolbars on my browser to see all 20 rows between each icon bar. However, 25 will work. It's only 5 more rows. I shouldn't get too lost...

 

[Moonsinger]

Personally, I liked the 20. I had to remove some toolbars on my browser to see all 20 rows between each icon bar. However, 25 will work. It's only 5 more rows. I shouldn't get too lost...

If you have more than 25 rows, more headers will appear and you can use them as footers.

 

[Tzar Sasha]

Okay, maybe it's a timer thing... I don't know.
I just checked on Mapfinder 2.4 and it has stopped.
Results were 7 maps acceptable and 536 maps rejected.
Settings were for 250 maps acceptable.
There is a message at the bottom of the Mapfinder window that says "Mapfinder stopped because the correct version of Civ is not running!" Yet you could plainly see C3C v1.22 in the background at year 4000BC of an unacceptable map.
Timer settings were/are:
16000 loading
10000 saving
15000 restart
500 keystrokes
5000 civ dialogs
These all seemed to be ample timings.... Any suggestions as to what went wrong?
I've got an 800mhz Celeron with 250MB ram.... Do I need to increase the timing and if so which?

 

[HellboundHeathcliff]

All this discussion about timing makes me wonder if Mapfinders defaults shouldn't be set perhaps twice as long as needed for a slow system and let the players with fast systems tweak it if they desire. A readme.txt would be a good place to put the instructions. I actually haven't tried it yet because the lack of a readme.txt file and this discussion makes it sound like a trial and error till it runs right utility.

 

[Moonsinger]

Okay, maybe it's a timer thing... I don't know.
I just checked on Mapfinder 2.4 and it has stopped.
Results were 7 maps acceptable and 536 maps rejected.
Settings were for 250 maps acceptable.
There is a message at the bottom of the Mapfinder window that says "Mapfinder stopped because the correct version of Civ is not running!" Yet you could plainly see C3C v1.22 in the background at year 4000BC of an unacceptable map.

Timer settings were/are:
16000 loading
10000 saving
15000 restart
500 keystrokes
5000 civ dialogs
These all seemed to be ample timings.... Any suggestions as to what went wrong?
I've got an 800mhz Celeron with 250MB ram.... Do I need to increase the timing and if so which?

Sorry, that message is a little bit mis-leading since sometimes it has nothing to do with the correct version of C3C. Here is exactly how it happened:

In the latest edition, I added a feature to track the timestamp of the autosave game file. If the autosave timestamp doesn't change when Civ3 save the game, that mean it has detected an incorrect autosave directory or something is wrong with the saving process. For example, if you point MapFinder to an autosave directory for PTW, MapFinder would knows that it's wrong when it sees C3C running in the background. Pretty smart, right? I thought so! Anyway, in your situation, it most likely happened as follows:

1. MapFinder was crusing just fine until around 543 iterations (536 rejected + 7 acceptable = 543 iterations) and C3C needed to be restarted. So far so good.

2. MapFinder gave order to your system to restart C3C and put itself to sleep or to wait for 15000ms.

3. After 15000ms, MapFinder waked up and attempt to order C3C to generate a new map. In the meantime, your system is a little bit busy and hadn't successfully launched C3C yet.

4. Because the timmer is out of synch in step #3, the time stamp of the autosave file remain the same; therefore, the reason for the wrong version message. Of course, I will re-word that message in the future. In any case, MapFinder was working correctly and this isn't a bug.

Anyway, to make the long story short and to solve this problem, all you have to do is to increase your C3C delay restart time to 20000ms or 25000ms. On a 800 MHz machine, it could easily take more than 15 seconds to launch C3C. You can time this by clicking on the C3C icon to launch it wait to see how long it actually take to come up. Therefore, 15 seconds (= 15000ms) is way too tight! Need to give it at least 5 more seconds to relax during the restart C3C process and it will be fine.