Hi Gerd If starting with unsigned deltas in polyline encoding, and attempting to reduce the length by testing reduced x/yBase values, there isn't any point in testing Base-1, as the normal number of bits will be the same (added sign bit) and there will be some overflows. Patch attached to change this - it gives me a very slight reduction in RGN size and might give a performance improvement (I didn't detect any). I've also fixed a couple of log. logic errors and changed minDx/y initial to 0, which is much less confusing because they are symmetric with maxDx/y which do use 0. Ticker
Hi Ticker, Funny, I would have used Integer.MinValue for maxDx/maxDy to make it less confusing. Do you have an explanation for the change in size? My understanding is that your patch just skips a test which will never find a better encoding? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 13:58 An: mkgmap development Betreff: [mkgmap-dev] Polyline base optimisation Hi Gerd If starting with unsigned deltas in polyline encoding, and attempting to reduce the length by testing reduced x/yBase values, there isn't any point in testing Base-1, as the normal number of bits will be the same (added sign bit) and there will be some overflows. Patch attached to change this - it gives me a very slight reduction in RGN size and might give a performance improvement (I didn't detect any). I've also fixed a couple of log. logic errors and changed minDx/y initial to 0, which is much less confusing because they are symmetric with maxDx/y which do use 0. Ticker
Hi Gerd I could have changed it to this, but it is looking for lowest val <= 0 and highest val >= 0 so 0 seemed the better choice. The slight improvement is because it gives up after 2 tries that don't show improvement, and before, for unsigned, one of these tries was wasted. Ticker On Thu, 2021-04-08 at 15:11 +0000, Gerd Petermann wrote:
Hi Ticker,
Funny, I would have used Integer.MinValue for maxDx/maxDy to make it less confusing. Do you have an explanation for the change in size? My understanding is that your patch just skips a test which will never find a better encoding?
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 13:58 An: mkgmap development Betreff: [mkgmap-dev] Polyline base optimisation
Hi Gerd
If starting with unsigned deltas in polyline encoding, and attempting to reduce the length by testing reduced x/yBase values, there isn't any point in testing Base-1, as the normal number of bits will be the same (added sign bit) and there will be some overflows.
Patch attached to change this - it gives me a very slight reduction in RGN size and might give a performance improvement (I didn't detect any).
I've also fixed a couple of log. logic errors and changed minDx/y initial to 0, which is much less confusing because they are symmetric with maxDx/y which do use 0.
Ticker _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Ticker, my understanding is that it is looking for the largest and the smallest delta value. It is possible to have a smallest value > 0. It probably doesn't matter because of the way how xBtits and yBits is calculated but I still think that your version is even more confusing. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 17:34 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Polyline base optimisation Hi Gerd I could have changed it to this, but it is looking for lowest val <= 0 and highest val >= 0 so 0 seemed the better choice. The slight improvement is because it gives up after 2 tries that don't show improvement, and before, for unsigned, one of these tries was wasted. Ticker On Thu, 2021-04-08 at 15:11 +0000, Gerd Petermann wrote:
Hi Ticker,
Funny, I would have used Integer.MinValue for maxDx/maxDy to make it less confusing. Do you have an explanation for the change in size? My understanding is that your patch just skips a test which will never find a better encoding?
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 13:58 An: mkgmap development Betreff: [mkgmap-dev] Polyline base optimisation
Hi Gerd
If starting with unsigned deltas in polyline encoding, and attempting to reduce the length by testing reduced x/yBase values, there isn't any point in testing Base-1, as the normal number of bits will be the same (added sign bit) and there will be some overflows.
Patch attached to change this - it gives me a very slight reduction in RGN size and might give a performance improvement (I didn't detect any).
I've also fixed a couple of log. logic errors and changed minDx/y initial to 0, which is much less confusing because they are symmetric with maxDx/y which do use 0.
Ticker _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Gerd It needs to know if there are deltas < 0, deltas > 0, and number of bits to represent the largest delta, regardless of sign. If all the deltas were, say between 10..20, max=20 is significant, but min=10 is of no relevance; only if it was < 0 would it have an effect. Ticker On Thu, 2021-04-08 at 15:45 +0000, Gerd Petermann wrote:
Hi Ticker,
my understanding is that it is looking for the largest and the smallest delta value. It is possible to have a smallest value > 0. It probably doesn't matter because of the way how xBtits and yBits is calculated but I still think that your version is even more confusing.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 17:34 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Polyline base optimisation
Hi Gerd
I could have changed it to this, but it is looking for lowest val <= 0 and highest val >= 0 so 0 seemed the better choice.
The slight improvement is because it gives up after 2 tries that don't show improvement, and before, for unsigned, one of these tries was wasted.
Ticker
On Thu, 2021-04-08 at 15:11 +0000, Gerd Petermann wrote:
Hi Ticker,
Funny, I would have used Integer.MinValue for maxDx/maxDy to make it less confusing. Do you have an explanation for the change in size? My understanding is that your patch just skips a test which will never find a better encoding?
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 13:58 An: mkgmap development Betreff: [mkgmap-dev] Polyline base optimisation
Hi Gerd
If starting with unsigned deltas in polyline encoding, and attempting to reduce the length by testing reduced x/yBase values, there isn't any point in testing Base-1, as the normal number of bits will be the same (added sign bit) and there will be some overflows.
Patch attached to change this - it gives me a very slight reduction in RGN size and might give a performance improvement (I didn't detect any).
I've also fixed a couple of log. logic errors and changed minDx/y initial to 0, which is much less confusing because they are symmetric with maxDx/y which do use 0.
Ticker _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Ticker, OK, I've committed your patch as is. The (my) code in makeShortestBitStream() was already hard to read, maybe you find a better way to handle the backup/restore tricks of xSameSign and ySameSign? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 18:03 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Polyline base optimisation Hi Gerd It needs to know if there are deltas < 0, deltas > 0, and number of bits to represent the largest delta, regardless of sign. If all the deltas were, say between 10..20, max=20 is significant, but min=10 is of no relevance; only if it was < 0 would it have an effect. Ticker On Thu, 2021-04-08 at 15:45 +0000, Gerd Petermann wrote:
Hi Ticker,
my understanding is that it is looking for the largest and the smallest delta value. It is possible to have a smallest value > 0. It probably doesn't matter because of the way how xBtits and yBits is calculated but I still think that your version is even more confusing.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 17:34 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Polyline base optimisation
Hi Gerd
I could have changed it to this, but it is looking for lowest val <= 0 and highest val >= 0 so 0 seemed the better choice.
The slight improvement is because it gives up after 2 tries that don't show improvement, and before, for unsigned, one of these tries was wasted.
Ticker
On Thu, 2021-04-08 at 15:11 +0000, Gerd Petermann wrote:
Hi Ticker,
Funny, I would have used Integer.MinValue for maxDx/maxDy to make it less confusing. Do you have an explanation for the change in size? My understanding is that your patch just skips a test which will never find a better encoding?
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 13:58 An: mkgmap development Betreff: [mkgmap-dev] Polyline base optimisation
Hi Gerd
If starting with unsigned deltas in polyline encoding, and attempting to reduce the length by testing reduced x/yBase values, there isn't any point in testing Base-1, as the normal number of bits will be the same (added sign bit) and there will be some overflows.
Patch attached to change this - it gives me a very slight reduction in RGN size and might give a performance improvement (I didn't detect any).
I've also fixed a couple of log. logic errors and changed minDx/y initial to 0, which is much less confusing because they are symmetric with maxDx/y which do use 0.
Ticker _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Gerd I didn't have any problem with *SameSignBak logic - compared with the rest of it - extraBit/routeNode & house number flag array, how these should behave for zero deltas and continuation road segments etc etc Ticker On Fri, 2021-04-09 at 08:01 +0000, Gerd Petermann wrote:
Hi Ticker,
OK, I've committed your patch as is. The (my) code in makeShortestBitStream() was already hard to read, maybe you find a better way to handle the backup/restore tricks of xSameSign and ySameSign?
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 18:03 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Polyline base optimisation
Hi Gerd
It needs to know if there are deltas < 0, deltas > 0, and number of bits to represent the largest delta, regardless of sign. If all the deltas were, say between 10..20, max=20 is significant, but min=10 is of no relevance; only if it was < 0 would it have an effect.
Ticker
On Thu, 2021-04-08 at 15:45 +0000, Gerd Petermann wrote:
Hi Ticker,
my understanding is that it is looking for the largest and the smallest delta value. It is possible to have a smallest value > 0. It probably doesn't matter because of the way how xBtits and yBits is calculated but I still think that your version is even more confusing.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 17:34 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Polyline base optimisation
Hi Gerd
I could have changed it to this, but it is looking for lowest val <= 0 and highest val >= 0 so 0 seemed the better choice.
The slight improvement is because it gives up after 2 tries that don't show improvement, and before, for unsigned, one of these tries was wasted.
Ticker
On Thu, 2021-04-08 at 15:11 +0000, Gerd Petermann wrote:
Hi Ticker,
Funny, I would have used Integer.MinValue for maxDx/maxDy to make it less confusing. Do you have an explanation for the change in size? My understanding is that your patch just skips a test which will never find a better encoding?
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 8. April 2021 13:58 An: mkgmap development Betreff: [mkgmap-dev] Polyline base optimisation
Hi Gerd
If starting with unsigned deltas in polyline encoding, and attempting to reduce the length by testing reduced x/yBase values, there isn't any point in testing Base-1, as the normal number of bits will be the same (added sign bit) and there will be some overflows.
Patch attached to change this - it gives me a very slight reduction in RGN size and might give a performance improvement (I didn't detect any).
I've also fixed a couple of log. logic errors and changed minDx/y initial to 0, which is much less confusing because they are symmetric with maxDx/y which do use 0.
Ticker _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
participants (2)
-
Gerd Petermann -
Ticker Berkin