Bug in splitter?

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Bug in splitter?

Thorsten Kukuk

Hi,

I try to split SRTM data with splitter r200, but splitter only
creates one big tile.

Some interesting statistics from osmconvert:
nodes: 2518069911
ways: 27074150
node id min: 1
node id max: 2679465052

Ok, I'm a little bit above a 32bit signed integer, but:
Splitting nodes into areas containing a maximum of 6.000.000 nodes each...
Area (30.8935546875,-125.0244140625) to (50.2294921875,-101.9970703125) contains
 -1.776.905.107 nodes. DONE!

This looks like an overrun somewhere. At first I thought it's
the 32bit int overflow, but for this the number is far too high.

Any ideas what's going wrong here?

  Thanks,
    Thorsten

--
Thorsten Kukuk, Project Manager/Release Manager SLES
SUSE LINUX Products GmbH, Maxfeldstr. 5, D-90409 Nuernberg
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 16746 (AG Nürnberg)
_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bug in splitter?

Felix Hartmann-2
Oh, so it's a splitter bug? I tried splitting SRTM data from phyghtmap
with the following command unsuccessfully:
phyghtmap --jobs=1 --osm-version=0.6 --step=25 --output-prefix=as2
--line-cat=1000,200 --source=view1,view3,srtm1,srtm3
--area=80:-11.01192:129.99999:83.20162 --max-nodes-per-tile=0
--max-nodes-per-way=250 --pbf


BTW -- how can I make sure with phyghtmap that multiple input files
don't overlap, nor miss data? 129.99999 is already interpreted as 130...

On 01.04.2012 21:48, Thorsten Kukuk wrote:

> Hi,
>
> I try to split SRTM data with splitter r200, but splitter only
> creates one big tile.
>
> Some interesting statistics from osmconvert:
> nodes: 2518069911
> ways: 27074150
> node id min: 1
> node id max: 2679465052
>
> Ok, I'm a little bit above a 32bit signed integer, but:
> Splitting nodes into areas containing a maximum of 6.000.000 nodes each...
> Area (30.8935546875,-125.0244140625) to (50.2294921875,-101.9970703125) contains
>   -1.776.905.107 nodes. DONE!
>
> This looks like an overrun somewhere. At first I thought it's
> the 32bit int overflow, but for this the number is far too high.
>
> Any ideas what's going wrong here?
>
>    Thanks,
>      Thorsten
>
_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bug in splitter?

Thorsten Kukuk

Hi,

On Sun, Apr 01, Felix Hartmann wrote:

> Oh, so it's a splitter bug?

Since osmconvert and osmosis have no problems with
the data, I assume it's a splitter bug. At least the
message about the number of nodes looks wrong to me.

> I tried splitting SRTM data from phyghtmap
> with the following command unsuccessfully:
> phyghtmap --jobs=1 --osm-version=0.6 --step=25 --output-prefix=as2
> --line-cat=1000,200 --source=view1,view3,srtm1,srtm3
> --area=80:-11.01192:129.99999:83.20162 --max-nodes-per-tile=0
> --max-nodes-per-way=250 --pbf
>
>
> BTW -- how can I make sure with phyghtmap that multiple input files
> don't overlap, nor miss data? 129.99999 is already interpreted as 130...

I don't know, luckily I didn't needed this.

  Thorsten

> On 01.04.2012 21:48, Thorsten Kukuk wrote:
> > Hi,
> >
> > I try to split SRTM data with splitter r200, but splitter only
> > creates one big tile.
> >
> > Some interesting statistics from osmconvert:
> > nodes: 2518069911
> > ways: 27074150
> > node id min: 1
> > node id max: 2679465052
> >
> > Ok, I'm a little bit above a 32bit signed integer, but:
> > Splitting nodes into areas containing a maximum of 6.000.000 nodes each...
> > Area (30.8935546875,-125.0244140625) to (50.2294921875,-101.9970703125) contains
> >   -1.776.905.107 nodes. DONE!
> >
> > This looks like an overrun somewhere. At first I thought it's
> > the 32bit int overflow, but for this the number is far too high.
> >
> > Any ideas what's going wrong here?
> >
> >    Thanks,
> >      Thorsten
> >
> _______________________________________________
> mkgmap-dev mailing list
> [hidden email]
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

--
Thorsten Kukuk, Project Manager/Release Manager SLES
SUSE LINUX Products GmbH, Maxfeldstr. 5, D-90409 Nuernberg
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 16746 (AG Nürnberg)
_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bug in splitter?

Felix Hartmann-2


On 01.04.2012 22:06, Thorsten Kukuk wrote:

> Hi,
>
> On Sun, Apr 01, Felix Hartmann wrote:
>
>> Oh, so it's a splitter bug?
> Since osmconvert and osmosis have no problems with
> the data, I assume it's a splitter bug. At least the
> message about the number of nodes looks wrong to me.
>
>> I tried splitting SRTM data from phyghtmap
>> with the following command unsuccessfully:
>> phyghtmap --jobs=1 --osm-version=0.6 --step=25 --output-prefix=as2
>> --line-cat=1000,200 --source=view1,view3,srtm1,srtm3
>> --area=80:-11.01192:129.99999:83.20162 --max-nodes-per-tile=0
>> --max-nodes-per-way=250 --pbf
>>
>>
>> BTW -- how can I make sure with phyghtmap that multiple input files
>> don't overlap, nor miss data? 129.99999 is already interpreted as 130...
> I don't know, luckily I didn't needed this.
Well for Asia there is no way around, at least without 16GB RAM. With
8GB one has to resort to at least 3 parts...

I'll ask on Panarchos site...

>
>    Thorsten
>
>> On 01.04.2012 21:48, Thorsten Kukuk wrote:
>>> Hi,
>>>
>>> I try to split SRTM data with splitter r200, but splitter only
>>> creates one big tile.
>>>
>>> Some interesting statistics from osmconvert:
>>> nodes: 2518069911
>>> ways: 27074150
>>> node id min: 1
>>> node id max: 2679465052
>>>
>>> Ok, I'm a little bit above a 32bit signed integer, but:
>>> Splitting nodes into areas containing a maximum of 6.000.000 nodes each...
>>> Area (30.8935546875,-125.0244140625) to (50.2294921875,-101.9970703125) contains
>>>    -1.776.905.107 nodes. DONE!
>>>
>>> This looks like an overrun somewhere. At first I thought it's
>>> the 32bit int overflow, but for this the number is far too high.
>>>
>>> Any ideas what's going wrong here?
>>>
>>>     Thanks,
>>>       Thorsten
>>>
>> _______________________________________________
>> mkgmap-dev mailing list
>> [hidden email]
>> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bug in splitter?

Gerd Petermann
In reply to this post by Thorsten Kukuk
Hi Thorsten,

a quick look shows some places where nodes are counted with a int value, so this will overflow. The negative value is reported with a formatting routine, I assume this converts the wrong value again.
I'll try to post a fix tomorrow.

Gerd