Samsung Galaxy S II (i9100)

Billi25
2012-01-20
2013-05-29
1 2 3 4 > >> (Page 1 of 4)
  • Billi25
    Billi25
    2012-01-20

    I tested GpsMid on my brandnew Samsung Galaxy S II (i9100). I installed "GpsMid 0.7.71-map69 for Android" from the nithly build page and loaded the latest selfbuild jar (was done for my now defect SE G705, "app = GpsMid-Generic-full"). And it works - kind of.
    At first it took my some time to get used to the "simple" design and to get in "touch" with GpsMid, so far I controlled it always by keys. First impression was that it looks poor, cheap besides other apps on the giant AMOLED screen… But hey its not an official release, still early beta. The map is there and most stuff works like I know it. GPS works, track recording works.
    I used the Wiki-page to learn/test all the touch areas/gestures. Wow! Works good, only sometimes my finger didn't find the
    right spot.

    I noticed following bugs/points for improovement:

    (1) The map screen looks blurry, like an badly upscaled/softened picture.
           See nullhttps://docs.google.com/open?id=0B86JYWvmfswENjRhMmI5NzktMzgyOC00ODlkLTkzMGYtODdhNzFhYjBhZGQz
           Picture taken with digital camera. Note that the zoom-touch-buttons border is thinner. The screen of SGS2 has the size of: 480x800 pixel.

    (2) Screen rotate to landscape is working. But "Map recenter: | " is off screen for advanced move, lower edge for eagle and of center for centered map.

    (3)  Search List: GpsMid freezes/hangs and crashes: Search works fine until double tap list entry OR executing any item from the menu (Route etc). I cant leave the search screen at all.

    Any Idea how to fix it? Do you need more information? I'll be happy to test.

    CU
    billi25

     
  • Billi25
    Billi25
    2012-01-22

    Hello developers,

    disappointing news after a bit more testing on my Samsung Galaxy S II. GpsMid for Android does not record the track:
    (a) after the screen is off,
    (b) if GpsMid is put to background.
    Without background recording GpsMid is not useful for me.

    I was not able to set my bluetooth receiver as GPS source (the list in discover GPS stays empty, but in Androids bluetooth setting I can see and connect to the receiver).

    The Radar view in GpsMid always says "Satelites n/a"

    (I installed also "GpsMid 0.7.57-map69 for Android, hires android 2.3", but could not recognize any difference)

    So, I am looking forward to test the next nightly builds.
    Billi25

     
  • sk750
    sk750
    2012-01-28

    Hello billi25,

    first, I need to tell I'm no Android user (did not ever compile an APK myself),
    but several other GpsMid developers are Android users,
    so I hope they will take care of the Android specific issues.

    What you might want to do is post a feature request for 1)
    and add bug reports for 1) to 3) and the background recording issue.

    Please also attach related debug logs to the tracker items.

    sk750

     
  • Billi25
    Billi25
    2012-01-28

    Hello sk750,

    Thanks for the answer.
    Before I open tickets I have some more questions/remarks.
    - Tested again with the nightly (About screnn: 0.7.71 (2012-01-07), but homepage says: (Thursday, 01-Sep-2011 20:12:25 UTC)
    - The log files are empty, except a message that logging is not compiled into this version. so no logs from Android :-(
      Is debug logging possible on Android, where is it?
    - Is issue (1) a bug report or a feature request? Is the display resolution predefined/fixed?
    - Issue (2) happens only if "GUI options / use icon menu" is not selected. Changing the option switches the behavior.

    regards
    billi25

     
  • sk750
    sk750
    2012-01-29

    Hello billi25,

    the nightlies are currently not updated, the date you see on the about screen is the map bundling date.

    - I don't know if it's an automatic process no more working or still has to be done manually. Maybe apmon
    or jkpj can comment on this.
    - Don't know about debug logging on Android, but expect it to be possible if it's compiled into the apk.
    - I think the resolution is somehow predefined would see (1) rather as a feature request.
    - You should add to the bug report of issue (2) that it happens only without icon menu. Mainly my touch screen experience with GpsMid is however that most of the time I do not need the menu at all. Maybe I'll also add long press of the * touch control for toggling track recording, then it's quite complete.

    Sorry I don't know much about Android and the nighty builds.

    sk750

     
  • Markus Bäurle
    Markus Bäurle
    2012-02-13

    Hi billi25,

    my detailed reply to your messages was 2/3 ready when my keyboard got stuck, which it does sporadically. It's a totally annoying problem in the PS/2 Linux driver. If the "h" key is repeating constantly, I can even be happy if the reboot works normal, but rescuing the text is out of question.
    So, sorry, I'm not going to retype it tonight, please be patient for a few more days.

    The short answer to the issue you described is this:
    One must know that we are using a framework, J2MePolish, which makes J2ME applications run on Android with the help of an adapter layer which implements some things (UI elements especially) itself and maps other stuff to the Android APIs.
    This framework has several bugs and problems and apps using it will probably never look as shiny as native Android apps.

    One of these problems is the incomplete support of the Bluetooth API, which is why you don't see your GPS mouse and have an empty satellite view.

    On the other hand, there would probably be no port of GpsMid to Android at all without this framework, so I'm not blaming these guys. I'm actually very grateful to them.

    Regarding Android, we are in a difficult situation.
    It would be much more promising to use the Android APIs directly, leaving back the aged J2ME with all its limitations.
    But this would mean rewriting large parts of GpsMid, also because its design does not strictly separate UI and underlying logic. The consequence would, in reality, be a branching of GpsMid if you ask me, as I don't think it would be feasible and sensible to strictly keep both ports, Android and J2ME, working at the same time. This is because improvements in favour of Android would most probably pose challenges for the J2ME part.
    But at the moment this is not really something to decide as I don't see that anybody could spend the time and energy needed for writing a native Android UI. We might gradually get this, screen by screen maybe. I don't know if this is feasible, I haven't investigated if both UIs could be mixed at all.

    So long,
    Markus

     
  • karel
    karel
    2012-02-21

    Hi billi25,
    can't help you with all your problems (especially the one regarding to screen resolution), but:
    - workarround for issue 3 is to disable row limit for search results. I haven't investigated the issue further.
    - problem with track recording after screen is off or app is paused is IMHO caused by calling trace.pause() in pauseApp() method in GpsMid.java. Yes, pausing the app is what the method pauseApp() should do. But in this case, trace.pause() should not close locationProducer. LocationProducer should be closed only if no track is recorded (gpx.isRecordingTrk() == false ?).

    Karel

     
  • Billi25
    Billi25
    2012-02-21

    hello all,

    thanks for the answers and hints.
    And indeed, the workarround for (3) works. I deleted the "500" in the field.
    If the screen resolution is fixed, I will adopt my geocaching-style for big high-res screens, as many icons are very tiny. That was good for the smaller SE-Screen, but makes me feel bad sighted on the SGSII.

    I tried a lot of other Map software on the phone. So far I like the OsmAnd (which is not always intuitive) and Locus (which is very good), but so far GpsMid offered the best offline routing of all. So, please keep up the good work.

    regards
    billi25

     
  • sk750
    sk750
    2012-02-22

    @karel: Thanks for investigating this. Can you provide an _Android-tested _ patch? I do not have an Android device but would test the patch then on Symbian with J2ME.

    @billi25: What makes GpsMid's offline routing superior in your opinion?
    Criteria could be e.g. efficiency of destination input, offered route modes, variety of route instructions, supported special cases for bicycle and foot routing (e.g. walk against one way direction) which IMHO are the main advantages compared to some other software. As main disadvantages I see calculation speed, no complete Germany / Europe map and sometimes not calculating the optimal route.

     
  • Billi25
    Billi25
    2012-02-22

    Compared to all other software the main advantage of GpsMids route calculation is the quality of the calculated route. In my area (Dresden, Germany) the OSM data is very good, so it knows nearly all one ways, speed limits etc. Most other android software relay on online services and the returned routes are not as efficent. Especially google maps got really crappy after the last big map update, as there is a lot of wrong information now. E.g. google deleted one-ways, added paths as roads and tries to route me straight through a graveyard… The offline routing alghorithm in OsmAnd find a much longer route and seems to miss turn restrictions. So GpsMid strength is really the route itself. Thanks.

    In most cases the calculation speed is not an issue. It was not an issue on my SE G705, and is not an issue with the Samsung Galaxy SII. but I noticed that the dual-core CPU can get really bussy on Android… I think this is a problem of code optimizing/translation in J2MEPolish, as is was fast on the SE G705.

     
  • sk750
    sk750
    2012-02-22

    Indeed, forgot the turn restriction support of GpsMid, it's an important feature.

    Regarding route calculation speed - actually it depends much on speed of reading the route data from the memory card and available RAM. Are you using a zipped map or a map directory? How much RAM is available on the SGSII for GpsMid (long press clock on map) and what's the max mem value in the settings?

     
  • sk750
    sk750
    2012-02-22

    Regarding issue 3: Is it an alternative workaround to activate in Setup/Gui the option "Suppress warning about exceeding search results"?

     
  • Billi25
    Billi25
    2012-02-23

    Hi

    a fresh started GpsMid shows values like "Freemem: 3356568" and "Totmem: 8789984". While after havy map zooming, and route calculation (>120 km) the values increase for totmem, e.g. "Freemem: 4152782" and "Totmem: 20744923".
    The max setting is 27207, I never changed it in Android.

    The option "Suppress warning about exceeding search results"? was activated. Changing it and the changing the "limit for search results" back and forth could not reproduce the error I had before. WTF? I'll try to reinstall GpsMid and test again.

     
  • karel
    karel
    2012-03-05

    Hi all.
    Regarding issue 3: "Suppress warning about exceeding search results" - yes, that's exactly what I did. Sorry for confusion.
    Regarging "pausing GpsMid" - I changed line 762 in Trace.java to: if (locationProducer != null && !gpx.isRecordingTrk()) {

    Karel

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-05

    Regarding going Android native - not sure what's the good way to go, and if it's possible to mix Android native UI code with J2MEPolish.

    A related question to ponder - how relevant are J2ME and Android in terms of future prospects for user counts, market share of each platform etc.? From around 2007, it seems by google searches that people have been asking "Is J2ME dead?"  The answer seems to be at least that Android is gaining lots of ground (1).

    Another question which has been asked on the J2MEPolish forum is "Is J2MEPolish dead?" (2) Well, there have been new releases after that, so it's not dead, but with the news about Blackberry not offering Java on tables (http://blog.enough.de/?p=621#more-621 ) it looks like things are not going too brightly for J2ME.

    (1) A posting (though in the context of browser stats) discusses the issue like this:

    "Most of the world’s phones are still feature phones, like 70% of them. And Android is only about 50% of the other 30%, so about 15% of total. And yet they compare the user base of the feature phones with that of Android smartphones, and they point out that they are about equal. How is that even possible?

    It’s possible because, again, we’re not talking real units here, we’re talking “browser usage”, and in total the time spent in a browser on a feature phone is also a lot less than the time spent on a smartphone’s browser. So even though there are still a lot more J2ME feature phones than Android smartphones, the time spent in total, is about equal for each right now. Of course, they make it look like Android is “losing” to J2ME phones, which is pretty ridiculous, because there are already several times more feature phones than Android phones."

    http://www.androidauthority.com/net-marketshare-shows-highly-misleading-stats-again-42137/

    (2) http://forum.enough.de/viewtopic.php?t=1822&sid=29489ba46af379d7cd65e7185d98fa42

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-05

    One comment about bad looks - I find that the appearance of GpsMid on a 800-ish by 480-ish or bigger display gets a lot better when switching road borders off.

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-05

    I've now updated the Android snapshot, and am building the updated Osm2GpsMid snapshot.

    There's no Android debug version build in the current build scripts, so now android debug nightly available at the moment.

     
  • sk750
    sk750
    2012-03-05

    @karel: Thanks for the change, just pushed the line into tthe git repository. Tried also to solve the issue that is triggered when "Suppress warning about exceeding search results" is off.

    @billi25: It'll be in the next nightly, hope jkpj can update it another time soon.

    @jkpj: Though I still do not own an Android device I would not mind if further development would mainly focus on Android devices and e.g. just the most important features would be put into the J2ME version, especially keeping GpsMid for J2ME able to read map formats of future Osm2GpsMids.

    Then ideally I think there would be separate source codes for Osm2GpsMid, GpsMid.jar, and GpsMid.apk with the main development being for GpsMid.apk and important changes being pulled into GpsMid.jar?

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-05

    I reupdated the Android snapshot (nightly), and the rebuild&update of the J2ME daily snapshot is ongoing.

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-05

    To use a bluetooth GPS receiver on an Android device:

    1) go to the device's Settings, go to Development, enable mock locations for GPS

    2) go to Android Market, install one of the bluetooth GPS apps, e.g. https://market.android.com/details?id=googoo.android.btgps

    3) Start the Bluetooth GPS program, connect to the Bluetooth receiver

    4) set GpsMid to JSR179 (or whatever the internal GPS receiver is called), start using GPS

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-05

    "- Issue (2) happens only if "GUI options / use icon menu" is not selected. Changing the option switches the behavior. "

    Hmm, this suggests that perhaps I (or someone else) has made an error of omission, and Android screen resizing support is active only for icon menu. Frankly, it never has occurred to me someone would use the Android version without icon menu.

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-05

    "The Radar view in GpsMid always says "Satelites n/a""

    Haven't looked very closely, but I think this is a shortcoming in the J2MEPolish framework - the GPS interface bridging Android GPS info into J2ME interface doesn't seem pass the detailed satellite info. This is one of the places where switching to use the native Android API would probably enable the missing functionality. Also the number of satellites (shown in the upper corner on J2ME / WinCE devices) is missing on Android.

     
  • sk750
    sk750
    2012-03-06

    @jkpj:
    - Screen rotation works here on J2ME with and without icon menus
    - I have suppressed the search warnings in GuiSearch. Just saw there's also code for giving search warnings in SearchNames. Why do both classes need to give search warnings when SEARCH_MAX_COUNT is reached?

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-06

    Map screen rotation works fine for me on Motorola Milestone (Android) cyanogenmod with and without icon menus.

    Can't remember if there's a reason for search warnings in two places, might be just a bug.

     
  • Jyrki Kuoppala
    Jyrki Kuoppala
    2012-03-06

    About (1), blurry map and (2) - I've also with some 800 x 480 devices (ZTE Blade I think)  seen screens which seem upscaled. On the Milestone the picture looks god. I don't know what's going on, but I suspect J2MEPolish.

    Also, I've seen there are some part of screen missing issues with the Android version. For example, on the Motorola Milestone (480 x 854 pixels), lower part of the virtual keyboard is not visible in portrait mode, there's a big black stripe as the lowest part of screen. In landscape mode all of the keyboard shows. Same goes for the "tacho" screens - lowest part of the screen is black. Also, occasionally I've seen that the direction texts on the map screen are not at the lowest part of the screen, but there's some map below the navigation directions.

    Problems with the virtual keyboard are not limited to the somewhat unusual Milestone resolution - also on the 7" Lanyu 910 tablet where I've finally got GPS working (resolution 800 x 480) has the issue of virtual keyboard not showing up, and even more of it is missing than on the Milestone. With the Milestone, I can see at least part of all keys, but on the Lanyu 910 there's one or two rows totally missing.

    I suspect the problems might be with J2MEPolish. Could be that if resolution is too low some way could be found to force high res, but I'm not familiar enough with J2MEPolish to know how. Hearing how the problems present themselves on other devices might help in finding out what's going on.

     
1 2 3 4 > >> (Page 1 of 4)