Sat2Chart

[phiggins]

skipperearly,

Thank you for your compliments.

What kind of charts are you you creating kaps or mbtiles?

[skipperearly]

At the moment I am testing with mbtiles and multizoom as they are so much easier to pass on for different systems, although I am aware of the reduced color depth to kaps.

[phiggins]

Ok thanks,

I may be able to give you an indication of the file size of the resulting chart(s) but the time it will take depends on too many variables including how long the download(s) will take.

[yachtvalhalla]

[QUOTE=martinworswick;3125091
There are some large libraries of kap charts on various users websites,are you aware of any libraries of mbtiles?
Regards,
Martin[/QUOTE]
Hi Martin,

Focusing on Asia, and mostly SE Asia, are my collection of charts (both Kap and Mbtiles), anchorages and routes (plus some other stuff) at this link:

Terry

Terry

[phiggins]

Skipperearly:

One feature which I would find useful is to get as early as possible some indication on how long this job will need and what file size to expect.

In the next version , Sat2Chart will estimate how large the resulting file size will be for mbtiles. As you can see in the attachment there are a couple of charts with small file sizes and one huge one around 5 gigabytes, that one will take a long time to process.

Hope this will satisfy your requirement. As I said, estimating the time is not possible.

[Viking Sailor]

Hey Paul,


Can you measure the initialization time, multiply that times a constant, and divide the results into the file size to get a time estimate? The constant can be a value based upon the type of file being generated and maybe selected for file size. You will probably need to do some testing to determine value(s) for the constant. If you can get within say 10% it would probably be good enough.


Paul

[hoolie]

Why don't you do that for yourself, then it's tailored to your own system

[phiggins]

Viking Sailor,

How do I calculate the Download time?

[Viking Sailor]

Originally Posted by phiggins

Viking Sailor,

How do I calculate the Download time?

Download either the beginning of the desired file, or another file of known size. Measure the time required to do this download. Use this measurement to estimate a nominal time to download the desired file. Display the estimated time to download and process the file. Download and process the file. Repeat the calculations for the next file.

[phiggins]

You cannot take one download and use it to estimate the rest. They all may be different. One may have zero download time and another may take hours. Can't be done.

[Viking Sailor]

Paul,

I have from time to time downloaded a number of files. Sometimes the download process provides a download rate (MB/s) and/or a time to download completion. I suspect that this info is provided by the download server via an API. When I offered the above suggestion I assumed that you would be aware of such an API and would have applied it to solving this problem. Thus, I could only conclude that there is no API available to provide the needed download statistics. As an alternative the suggested method is based upon my observations that over the short term the download rate is fairly consistent from a given server over an established link. However, I have also observed that the server load or the quality of the link can vary resulting in an erroneous time calculation. The thing is the time calculations only need to be your best guess to be helpful. And it doesn't even have to be time, it could be a percentage of completion which can be at a variable rate. (As in: Windows updates)


Paul

[phiggins]

I think providing an estimate of the size of the resulting mbtiles files is sufficient for a user to determine how long it will take.

BTW love your Netflix series (Vikings).

[hoolie]

Originally Posted by Viking Sailor

Paul,

I have from time to time downloaded a number of files. Sometimes the download process provides a download rate (MB/s) and/or a time to download completion. I suspect that this info is provided by the download server via an API. When I offered the above suggestion I assumed that you would be aware of such an API and would have applied it to solving this problem. Thus, I could only conclude that there is no API available to provide the needed download statistics. As an alternative the suggested method is based upon my observations that over the short term the download rate is fairly consistent from a given server over an established link. However, I have also observed that the server load or the quality of the link can vary resulting in an erroneous time calculation. The thing is the time calculations only need to be your best guess to be helpful. And it doesn't even have to be time, it could be a percentage of completion which can be at a variable rate. (As in: Windows updates)


Paul

As I suggested above - why don't you do your own tests if you're worried about it? Most people when they find the PC's busy just go and do something else for a while.

TBH I think Paul has a hell of a lot of other things to occupy himself without spending hours conjecturing about your download speeds.

[Viking Sailor]

Originally Posted by phiggins

I think providing an estimate of the size of the resulting mbtiles files is sufficient for a user to determine how long it will take.

BTW love your Netflix series (Vikings).

OK! Looking forward to seeing what you implement.





I've never watched that series.

[skipperearly]

Having started this idea of "estimating space an d time" I feel I should comment on some of the statements above.
When creating mbtiles with option sasplanet and e.g. 5 splits I can see in each download dialog various estimates including time. I appreciate that too frequent updating of the estimates will result in undesirable overheads but there should be something better than just leave it to the user to guess based on the expected file sizes.https://www.cruisersforum.com/forums...lies/peace.gif

Something else:
When testing and producing mbtiles in different parts of the geography I produced the attached error.

I was first trying to produce charts from polygons and in another run (rebooted in between to make sure!) following a path. Found it confusing that the "select zoom levels" dialog asks for polygons when only path is the source and vice versa. Also pre-selected zoomlevels can easily be overlooked (e.g x11 onwards) and then Sat2Chart tries to make tiles for the path and then also the tiles from the last polygon run.
Would like to suggest:
- offer a "clear all zoom selection" option
and
- some dialog logic based on path OR polygon vs path AND polygon as source for map creation.

just my2cent - and its a great program - thanks!