Notices


Reply
Thread Tools
Posts: 482 | Thanked: 550 times | Joined on Oct 2010
#601
Originally Posted by jj0 View Post
I think I've come up with the reason the font is messed up, sometimes supersmall in the directions.

You said, the size is determined automatically, so that the instructions fit in the box, right?

Well this is it, sometimes the directions are super long, and they seem to be fitted on only one line, and then when super small split...

I think this is a wrong approach, the font should always stay the same, the instructions should have line folding, and then after a couple of lines, if they don't fit in the box, well too bad, they should just be cut...
I agree. The most significant part of the instructions is usually the first few words.
 
Posts: 1,548 | Thanked: 7,510 times | Joined on Apr 2010 @ Czech Republic
#602
Originally Posted by petur View Post
I was preparing a trip, and just browsing some location, so no cached tiles, everything needed to be downloaded as I was dragging the map around. Some things I noticed:
- when dragging a lot, I was greeted with a screen full of blue squares, tiles waiting for a slot. Took quite some time for them to get a slot and draw a map.
- same for zooming in/out: lots of tiles waiting for a slot.

I thought the visible tiles had priority? Or maybe they are all busy downloading, are slow, and thus need so much time.
Yep - blue tiles -> maximum number of concurrent tile download threads reached. The tiles are actually downloaded LIFO, but because each download takes some time, downloads for an "invisible" area might be using the available slots.

Originally Posted by petur View Post
I guess aborting the download would be an option if the tile is no longer visible?
I'll have to check if the file download module/library supports this.
Provided I can get this to work efficiently (without starting and stopping downloads for the same tiles all the time ), there might be quite a few advantages:
  • less data transfered
  • quicker reaction time to a changing location
  • possibly also less concurrent download threads


Originally Posted by whats_up_skip View Post
If I turned off the route it gave me no warning or information on how to get back to the route. This is could a good thing some times, but a big problem other times.
Originally Posted by skykooler View Post
This is one of the big things I am waiting for, too: automatic route recalculation.
Automated rerouting will be added - I already have an idea how to do it (distance from route segment larger than X -> reroute), I'll just have to actually write it. There will also be an OFF switch for it, of course.
BTW, I'm also thinking about doing something like:
"rerouting in 5 seconds" + displaying a cancel buttons and preparing the route in the meantime.


Originally Posted by cheve View Post
@MartinK: I also like the 'Map Screen' in the power button menu; however, it would be more user-friendly(ihmo), that an on screen 'goto-map-screen' button is implemented at each level of the menu screen.
Just do a long press on the "escape" arrow - it should switch you directly to the map screen. I really have to finally document this on the wiki

Originally Posted by cheve View Post
I have just discovered that I am able to have two copy/instance of modRana running at the same time and these instances seem to be independent of each other. Is it by design?
Cheers,
Well, not really by design but certainly possible But both (or more ) instances share the same options storage file and I don't really test it with this in mind
__________________
modRana: a flexible GPS navigation system
Mieru: a flexible manga and comic book reader
Universal Components - a solution for native looking yet component set independent QML appliactions (QtQuick Controls 2 & Silica supported as backends)
 

The Following 2 Users Say Thank You to MartinK For This Useful Post:
Posts: 1,548 | Thanked: 7,510 times | Joined on Apr 2010 @ Czech Republic
#603
[QUOTE=jj0;975130]Well, I drove a looong way and first I want to say:

A SIGNIFICANT, SIGNIFICANT improvement. Changing tile loading to LIFO makes a great different, the nav is useable now at high speeds. Thanks a lot, Marting.

Now for the problems however. I've identified three major issues, one of which is critical:

Originally Posted by jj0 View Post
1. Map overlay results in at least doubling the CPU and mem load and makes it unuseable completely, the system slows down and just a big problem... You've already mentioned you're working on this.

2. Same with autorotation, to a lesser degree, it's useable but slows it down considerably. But that's how a nav should work, so this feature is needed. I understand you're also already addressing this issue.
Yep, I'm working on this - I have working visibility checking for rotation (previously an unnecessarily large are was loaded and drawn to cover the entire rotated viewport). I'll do the overlaid tile caching next - the tiles will be overlaid once and then cached, not drawn twice on every refresh like previously.
NOTE: Once clutter support is in, things like smooth zooming & rotation, fake 3D and mindbogling tile loading effects will become possible.

Originally Posted by jj0 View Post
3. This one I think is critical - the tile loading is still a SUPER slow process. It is slightly better when using the sql than regular files. The speed of loading each individual tile for display seems to vary, however in extended navigation it howers somewhere slightly under 2 secs per tile. This is terrible, since usually you're displaying a screen of 8 tiles or so, they go from green to display (this is for already downloaded tiles) one by one, each in under 2 seconds. The whole display thus takes 15 seconds, this on zoom in and zoom out, so this problem is a significant one.
The tiles really should not take so long to load - most of them are very small and should load instantenously.
The tile loading is done sequentially in a background thread (so that the GUI doesn't have to wait for the tiles to load) - this thread might be resource starved or blocked by something.

Originally Posted by jj0 View Post
It seems to me that tiles are also loaded around the visible area of the screen, since when dragged they are showed immediately for a while.
Once loaded from storage or network, tiles are cached in memory. There is a limit on the number of cached tiles - once this limit is reached, old tiles start to be discharged.

Originally Posted by jj0 View Post
However there is some intense processing when loading the tiles, which results in this incredibly slow loading process, which is bad.

Finally, I strongly recommend making the tile download also LIFO, as it's FIFO I think, and it's a big prob when on the move, zoomed in, downloading new tiles. It should be LIFO just like tile loading, with the ones no longer needed download paused and resumed after the visible ones are downloaded.
I'll have to check if it is possible to pause a download - it might be easier to just cancel ant then start again, concerning the generally small size of the tiles.

Originally Posted by jj0 View Post
Another quite big problem is the size of the font. Many times, as has been mentioned here before, it gets messed up and is SUPER tiny, meaning you cannot read it. This issue usually corrects itself with the move to next driving direction, however it's always on a specific direction, i.e. if it's tiny, you go to next which is big for instance, and then back to the tiny one, it stays tiny.

More often than not, the directions' font is tiny which is a big problem, since many times I wanted to check the km to next turn etc. but couldn't.
As there were already quite a few requests for a more readable text in the navigation info box, I'll make this my priority for the next release.

Originally Posted by jj0 View Post
Also, you definitely should change the setting for distance when speaking the directions to time-based interval setting, i.e. how many secs before a turn based on recent speed.

I also recommend adding, total remaining distance towards goal in the directions blue window display, not only next turn, but total remaining km.
That should fit to the right quite nicely, like this:
in 50 meters <-- blank space ---> 50/120 km
turn right to the Sesame street


Originally Posted by jj0 View Post
These are my major remarks in testing the latest version.
Thanks ! The feedback from your field-testing is very important and much appreciated!

Originally Posted by jj0 View Post
Thanks for the big improvement, and please try to focus the tile loading especially, as well as considering my other suggestions.
__________________
modRana: a flexible GPS navigation system
Mieru: a flexible manga and comic book reader
Universal Components - a solution for native looking yet component set independent QML appliactions (QtQuick Controls 2 & Silica supported as backends)
 

The Following User Says Thank You to MartinK For This Useful Post:
Posts: 1,548 | Thanked: 7,510 times | Joined on Apr 2010 @ Czech Republic
#604
Originally Posted by xerxer View Post
i'm going to donate to this , it's very very great
Thanks in advance !

Originally Posted by xerxer View Post
and i'm waiting for an update soon
I'm working on it.

Originally Posted by jj0 View Post
And I would recommend adding two buttons to the mainscreen when the zoom-in, zoom-out and menu buttons become available by clicking the screen:

- Toggle map overlay on/off, I do this quite often, sometimes it's useable to view the sat nav over the map, sometimes not.
- Sound on/off, this is also useful while navigating sometimes.
This might be a job for the (not yet properly implemented ) map screen widgets - the menu already has quite a few (5 + nav. box) elements and might become too crowded if I add more by default.

What about a stopgap - an on-screen button that instantly launches the app menu ?

Originally Posted by jj0 View Post
I have two additional recommendations, it'd be useful to show current travelling speed.

And then, alongside the remaining km to next turn, and total remaining km, also the estimated time to next turn, and total remaining estimated time.
Well, the screen real estate in the navigation box is already limited

I'll have to add something like "show time/show/distance/ show time and distance" and a "show speed" toggle to options, so that the amount of displayed information can be customized


Originally Posted by jj0 View Post
I think I've come up with the reason the font is messed up, sometimes supersmall in the directions.

You said, the size is determined automatically, so that the instructions fit in the box, right?

Well this is it, sometimes the directions are super long, and they seem to be fitted on only one line, and then when super small split...
Yep, its exactly like this - many words = small font. This is also how all text rendering works in modRana.

Originally Posted by jj0 View Post
I think this is a wrong approach, the font should always stay the same, the instructions should have line folding, and then after a couple of lines, if they don't fit in the box, well too bad, they should just be cut...
I'm going to implement something like this - use a big font (size proportional to screen size), if it doesn't fit on a single line, try to fold to more lines & cut lines that don't fit in the box. It would be also sensible to indicate the some text has been cut.
__________________
modRana: a flexible GPS navigation system
Mieru: a flexible manga and comic book reader
Universal Components - a solution for native looking yet component set independent QML appliactions (QtQuick Controls 2 & Silica supported as backends)
 

The Following User Says Thank You to MartinK For This Useful Post:
Posts: 482 | Thanked: 550 times | Joined on Oct 2010
#605
I have a weird problem. When approaching one exit, this is what Google says: "Take exit 340A on the left toward Lincoin Dr/Kelly Dr".
Modrana always says "Take exit three hundred and forty A on the left toward Lincoin Doctor slash Kelly Drive."
Apart from the misspelling of "Lincoln" (Google's fault), why does it say Doctor the first time and Drive the second?
 
Posts: 2,225 | Thanked: 3,822 times | Joined on Jun 2010 @ Florida
#606
Originally Posted by skykooler View Post
I have a weird problem. When approaching one exit, this is what Google says: "Take exit 340A on the left toward Lincoin Dr/Kelly Dr".
Modrana always says "Take exit three hundred and forty A on the left toward Lincoin Doctor slash Kelly Drive."
Apart from the misspelling of "Lincoln" (Google's fault), why does it say Doctor the first time and Drive the second?
My guess is that since the parser has no "human" way of understanding language, and it's meant to speak for more general/broad uses than JUST routing, "Dr" can be either Doctor or Drive - by extension, it uses heuristics to guess when it means what - in this case, it sees "Dr" at the very end of the string of text, and presumably assumes that that can't be Doctor because English never uses "doctor" like that, at the end of a sentence as an abbreviation.

However, the "/" doesn't make sense to it as the termination point in a clause, nor does it parse that "[something] Dr/[something else] Dr" means is two swap-able sub-clauses, both of which are thus to be seen as to be read as if they were at the end of the sentence.

Since this is the first time in a long while that I've looked at this thread, I have no idea what modRana uses as a text-to-speech backend.

Of course, you can't even take out "Doctor" as a reading of "Dr" even in Nav software, because you can't assume you won't run into a road called "Dr Stevenson Dr" or something.
 

The Following User Says Thank You to Mentalist Traceur For This Useful Post:
Posts: 650 | Thanked: 497 times | Joined on Oct 2008 @ Ghent, Belgium
#607
this weekend brought up two bugs related to GUI rotation:

1) when you have modRana configured to auto-rotate, and you start it while holding it in portrait, it will not rotate and stay landscape. After rotating the phone to landscape and back, it follows.

2) when configuring the GUI to force a certain rotation, it will only use that when you restart it.

The above two made me stop on my bike to get the rotation right (had to take it out of the holder to get the orientation right)


I also discovered that while in bike mode, it will not route nor find/accept any address given to it. It seems that for bike routing it gets that info elsewhere, and this is not working. Good thing the route is maintained when switching between modes - this was my workaround
__________________
Affordable mobile internet in Belgium: Try Mobile Vikings
2 GB, 1000 SMS and 15 euro of talk time for.... 15 euro
 

The Following User Says Thank You to petur For This Useful Post:
Posts: 482 | Thanked: 550 times | Joined on Oct 2010
#608
One other suggestion: When two turns are close together, concatenate the directions, for example to make it say "Turn left on Elm Street, then turn right on Main Street", or even "Turn left on Elm Street, then turn right". This would help immensely in the city, where the notification distance for one direction is often passed while another is being spoken.
 

The Following User Says Thank You to skykooler For This Useful Post:
Posts: 17 | Thanked: 2 times | Joined on Nov 2009 @ Pierrelatte, France
#609
hello, can some guru help me ? I had installed modrana with the aptget method but here is the problem :


[...]
le-1.3/free modrana 0.24-1 [1974kB]
Fetched 4872kB in 10s (446kB/s)
(Reading database ... 28938 files and directories currently installed.)
Preparing to replace espeak-data 1.44.05 (using .../espeak-data_1.45.14_all.deb) ...
Unpacking replacement espeak-data ...
Preparing to replace libespeak 1.44.05 (using .../libespeak_1.45.14_armel.deb) ...
Unpacking replacement libespeak ...
Selecting previously deselected package espeak.
Unpacking espeak (from .../espeak_1.45.14_armel.deb) ...
Selecting previously deselected package espeak-extra-data.
Unpacking espeak-extra-data (from .../espeak-extra-data_1.45.10_all.deb) ...
Selecting previously deselected package modrana.
Unpacking modrana (from .../modrana_0.24-1_all.deb) ...
Processing triggers for catorise ...
################################################## ##################################Setting up espeak-data (1.45.14) ...
Setting up libespeak (1.45.14) ...
Setting up espeak (1.45.14) ...
Setting up espeak-extra-data (1.45.10) ...
Setting up modrana (0.24-1) ...
postinst running
generating *.pyc files
Listing /opt/modrana ...
Listing /opt/modrana/cache ...
Listing /opt/modrana/cache/images ...
Listing /opt/modrana/cache/tracklogs ...
Listing /opt/modrana/data ...
Listing /opt/modrana/data/gui_source ...
Listing /opt/modrana/data/gui_source/backup ...
Listing /opt/modrana/data/gui_source/icons ...
Listing /opt/modrana/data/gui_source/svg ...
Listing /opt/modrana/data/poi ...
Listing /opt/modrana/data/tiledata ...
Listing /opt/modrana/modules ...
Listing /opt/modrana/modules/configobj ...
Listing /opt/modrana/modules/device_modules ...
Listing /opt/modrana/modules/googlemaps ...
Listing /opt/modrana/modules/googlemaps/local_simlejson ...
Listing /opt/modrana/modules/googlemaps/local_simlejson/simplejson ...
Listing /opt/modrana/modules/googlemaps/local_simplejson ...
Listing /opt/modrana/modules/googlemaps/local_simplejson/simplejson ...
Listing /opt/modrana/modules/googlemaps/local_simplejson/simplejson/tests ...
Listing /opt/modrana/modules/pycha ...
Listing /opt/modrana/modules/pyrender ...
Listing /opt/modrana/modules/pyrender/blank ...
Listing /opt/modrana/modules/pyroutelib2 ...
Listing /opt/modrana/modules/upoints ...
Listing /opt/modrana/modules/urllib3 ...
Listing /opt/modrana/nbproject ...
Listing /opt/modrana/themes ...
Listing /opt/modrana/themes/default ...
Listing /opt/modrana/themes/night ...
Listing /opt/modrana/tracklogs ...
Listing /opt/modrana/tracklogs/log ...
Listing /opt/modrana/tracklogs/misc ...
Listing /opt/modrana/tracklogs/online ...
seting modRana folder ownership and permissions
modRana upgrade: removing possible old cache files
W: Duplicate sources.list entry http://repository.maemo.org fremantle-1.3/free Packages (/var/lib/apt/lists/repository.maemo.org_extras_dists_fremantle-1.3_free_binary-armel_Packages)
W: Duplicate sources.list entry http://repository.maemo.org fremantle-1.3/non-free Packages (/var/lib/apt/lists/repository.maemo.org_extras_dists_fremantle-1.3_non-free_binary-armel_Packages)
W: Duplicate sources.list entry http://repository.maemo.org fremantle-1.3/free Packages (/var/lib/apt/lists/repository.maemo.org_extras_dists_fremantle-1.3_free_binary-armel_Packages)
W: Duplicate sources.list entry http://repository.maemo.org fremantle-1.3/non-free Packages (/var/lib/apt/lists/repository.maemo.org_extras_dists_fremantle-1.3_non-free_binary-armel_Packages)
W: You may want to run apt-get update to correct these problems

/ # apt-get update

Hit http://repository.maemo.org fremantle-1.3 Release.gpg
Ign http://repository.maemo.org fremantle-1.3/free Translation-fr
Ign http://repository.maemo.org fremantle-1.3/non-free Translation-fr
Get:1 http://repository.maemo.org fremantle-1.3 Release.gpg [489B]
Ign http://repository.maemo.org fremantle-1.3/free Translation-fr
Ign http://repository.maemo.org fremantle-1.3/non-free Translation-fr
Hit http://repository.maemo.org fremantle-1.3 Release
Get:2 http://repository.maemo.org fremantle-1.3 Release [7958B]
Ign http://repository.maemo.org fremantle-1.3/free Packages/DiffIndex
Ign http://repository.maemo.org fremantle-1.3/non-free Packages/DiffIndex
Ign http://repository.maemo.org fremantle-1.3/free Packages/DiffIndex
Ign http://repository.maemo.org fremantle-1.3/non-free Packages/DiffIndex
Hit http://repository.maemo.org fremantle-1.3/free Packages
Hit http://repository.maemo.org fremantle-1.3/non-free Packages
Get:3 http://repository.maemo.org fremantle-1.3/free Packages [7355kB]
Get:4 http://repository.maemo.org fremantle-1.3/non-free Packages [222kB]
Fetched 7585kB in 50s (151kB/s)
Reading package lists... Done
W: Conflicting distribution: http://repository.maemo.org fremantle-1.3 Release (expected fremantle-1.3 but got )
W: Duplicate sources.list entry http://repository.maemo.org fremantle-1.3/free Packages (/var/lib/apt/lists/repository.maemo.org_extras_dists_fremantle-1.3_free_binary-armel_Packages)
W: Duplicate sources.list entry http://repository.maemo.org fremantle-1.3/non-free Packages (/var/lib/apt/lists/repository.maemo.org_extras_dists_fremantle-1.3_non-free_binary-armel_Packages)
W: Duplicate sources.list entry http://repository.maemo.org fremantle-1.3/free Packages (/var/lib/apt/lists/repository.maemo.org_extras_dists_fremantle-1.3_free_binary-armel_Packages)
W: Duplicate sources.list entry http://repository.maemo.org fremantle-1.3/non-free Packages (/var/lib/apt/lists/repository.maemo.org_extras_dists_fremantle-1.3_non-free_binary-armel_Packages)
W: You may want to run apt-get update to correct these problems

/ # modrana

== modRana Starting ==
device string (first parameter): n900
process 7579: arguments to dbus_connection_get_data() were incorrect, assertion "connection != NULL" failed in file dbus-connection.c line 5757.
This is normally a bug in some application using the D-Bus library.
process 7579: arguments to dbus_connection_set_data() were incorrect, assertion "connection != NULL" failed in file dbus-connection.c line 5721.
This is normally a bug in some application using the D-Bus library.
Aborted


/ # modrana

== modRana Starting ==
device string (first parameter): n900
process 7592: arguments to dbus_connection_get_data() were incorrect, assertion "connection != NULL" failed in file dbus-connection.c line 5757.
This is normally a bug in some application using the D-Bus library.
process 7592: arguments to dbus_connection_set_data() were incorrect, assertion "connection != NULL" failed in file dbus-connection.c line 5721.
This is normally a bug in some application using the D-Bus library.
Aborted

how can i un-install now ?

i'm sorry but my english is so so so poor ... thanks for your help
 

The Following User Says Thank You to jose6a For This Useful Post:
Posts: 1,378 | Thanked: 1,604 times | Joined on Jun 2010 @ Göteborg, Sweden
#610
to uninstall and remove all settings etc as root:

Code:
apt-get purge modrana
Those errors should not be fatal, at least they are not when I get them on a desktop system

Last edited by handaxe; 2011-04-17 at 19:12.
 

The Following User Says Thank You to handaxe For This Useful Post:
Reply

Tags
bada rox, martin_rocks, modrana, navigation, openstreetmap, the best, wehasgps

Thread Tools

 
Forum Jump


All times are GMT. The time now is 14:34.