![]() |
Plans for Maemo Mapper v2.0
I'm on the way toward releasing Maemo Mapper v2.0. Maemo Mapper v2.0 is going to represent a major shift in the paradigm I've been using up to this point, mostly as a reaction to the release of the N800. The focus with the 770 was very much on performance, due to the limitations of the hardware. The focus in Maemo Mapper v2.0 will be on features and aesthetics, including drag-panning, heading-based view rotation, and maybe even some transparency or anti-aliasing. Heck, I'll even be testing the feasability of implementing a "perspective" view, allowing you to see more of the map in front of you.
By popular request, I will also be aiming for cross-platform compatibility in hopes of a Windows (or at least x86) release, so you can run Maemo Mapper on your laptop at any resolution you wish. (For that reason, I am even considering renaming or forking the project. Any ideas? Not GpsDrive. :)) One of the big changes will be to use a database cache instead of a file system cache, to address the internal fragmentation of thousands of small files eating away at the free space of memory cards. Before I go further in the development of this particular new feature, I wanted to get some opinions about how to introduce it. Those of you with "extensive" map caches may not appreciate having to re-download the maps that they've downloaded. I'm planning to provide a command-line utility to "convert" a file system cache to the new database cache format. Would you prefer that I also continue to allow Maemo Mapper to work with a file system cache? I can set it up so that each repository can be switched independently between a file system cache or a database cache. (It would complicate the code and reduce performance slightly.) I guess my real question is this: does anyone see any value in maintaining the old file-system-based map cache? The only benefit I see is being able to browse through the individual map files, deleting or adding files as you see fit. The file system hierarchy is too complex for this, though, and if you're savvy enough to sort it out, you're probably savvy enough to mess with a database cache directly. Are there other, more relevant reasons to maintain backward compatibility with file system map caches? Also, I'm open to other ideas to implement in v2.0, but keep in mind that I'm only one man, and my plans for v2.0 are already taking up a lot of my spare time. Hopefully, I'll find the time to document some of these plans in the Garage, but until then, feel free to post suggestions to this thread or the Feature Request section of the Maemo Mapper Garage Tracker. Note: I will still, always be releasing every new version of Maemo Mapper for the 770, as well as for the N800. Some of the newer features may not run as well under particular conditions, or may require the use of swap, but it will ALWAYS be runnable (and usable) on the 770. I still own my 770, and I want to be able to run Maemo Mapper on it, too. |
Re: Plans for Maemo Mapper v2.0
Gnuite, your posting carries considerable weight among Nokia tablet users, since maemo-mapper appears to me to be, along with FBReader, the slickest and most mature application running on either the N770 and N800. My first reaction is that you should also make the same posting to the maemo-users@maemo.org list.
My second reaction is that, in my experience, real-time navigation/mapping software developers experience a more substantial evolution of their package's paradigm than developers of more conventional software. In other words, don't fight change but rather welcome it as a natural process. Embarking on a cross-platform effort is not for the faint of heart. My impulse is to urge you to first find a like-minded confederate to "own" that porting job. You have to weigh the risk of burnout (from taking on too much yourself) against the risk of a crappy port due to a too-loosely-coupled collaboration. At any rate, I think you shouldn't worry about changes in the cache structure; I think it's too transitory for any user to seriously hack. |
Re: Plans for Maemo Mapper v2.0
Good to hear :)
Also, I don't see any reason to keep the old image caching mechanism. Perspective view (would require a gps I suppose) sounds like a great idea. |
Re: Plans for Maemo Mapper v2.0
Quote:
Howard |
Re: Plans for Maemo Mapper v2.0
My stupid suggestion for a name is MaeMapper - it does indeed look stupid when written down but, pronounced "mymapper" just might work. And, it would be clear that it's a close relative to the original... ouch. Nevermind. In any case, I admire your work on the software and also the active community here.
Sorry, my first post to this forum, am expecting to get my N800 tomorrow and intending to use it for navigation... still a bit uncertain whether I should just go ahead and buy the Navicore package from Nokia, as I need a GPS receiver in any case. Well. |
Re: Plans for Maemo Mapper v2.0
It's a great application..but with the inroads of navicore with connection independent functionality..is there a reason to continue on this path..maemo mapper was great..but with this availability of a 3rd party program..hmmm..I don't know. Not sure efforts shouldn't be directed somewhere else
|
Re: Plans for Maemo Mapper v2.0
Well, unless navicore is free and open source I see quite a few reasons to continue. Competition is good and any time you have a single choice it's not the best situation, whether it be in regards to features or bugs. Besides, MM is quite capable for a from-scratch program for the platform. I've used it many times, and if nothing else a map with a dot "You Are Here" and breadcrumb trail is invaluable.
As far as the RFC I would definitely move away from the file/directory-based structure. If nothing else it would make it easier to a copy a single DB file (i.e. if you went with SQLite) than a million tiny ones. I would also be curious to see which DB engine you might choose since obviously it can't be too big, but performance might be marginal for some of the simplest ones. Keep up the great work! Larry |
Re: Plans for Maemo Mapper v2.0
Great work!
Can you get the n800 to use n95 gps data? I would lov if nokia or somebody got that going, n series devices need tto sing together! |
Re: Plans for Maemo Mapper v2.0
Thanks for your work so far. I do not agree with those who think that you should stop your work because of Navicore.
We'll keep waiting for the 2.0 version, most of all (I'm talking about me now) because of the map rotating feature (tough perspective will be a very interesting one, too :D ) Bye, Omar |
Re: Plans for Maemo Mapper v2.0
HI Gnuite,
Your v2 sounds marvellous ;) It would also be nice to allow an integration with a contact database (via plugin I suppose) It would allow to store GPS coordinate for contacts and a "go to contact" functionnality ... What do you think ? Thanks for the great work ... Fred |
Re: Plans for Maemo Mapper v2.0
Quote:
And thanks, again, for the continuing 770 support ! |
Re: Plans for Maemo Mapper v2.0
I think the idea of a database cache is good - I personally (having downloaded large maps) wouldn't mind switching and downloading them again. Its a wifi device, its not as if I am paying for downloads.
Thanks for the great work. |
Re: Plans for Maemo Mapper v2.0
Great,
I am going to hold my breath and wait ... :-) MaemoMapper is my top application (with FBReader). Do not include File system cache - it is why you want to switch to 2.0 version is not it ? So I would purge file system cache and for people owning megabytes of maps would release file system -> database convertor. I love your work ... still not breathing :-) |
Re: Plans for Maemo Mapper v2.0
Really glad to hear, finally, there's some hope we are getting rid of that filesystem approach for maps. Go, go, do whatever you need to completely remove it and replace with something better: achieving the possibility to move a single file rather than gazillions of tiny files back and forth the card would be super-great! Really it would be a nice boost to the usability of the software, and do what you need to gain best performance, including totally getting rid of backward compatibility.
On 770, I think you should give the user the possibility of turning on/off those eyecandies (antialias, transparency, rotation, perspective) as I suspect they could reduce performance. And I must say the route calculation web service often doesn't respond, so it would be wise implementing that in the software (even if it takes hours to give a response). Finally, the name: what about "maemaps"? Try searching it with Google, you can see there are 0 occurrencies for it! But I understand you need a "platform independent" name, so what about "2ndPilot"? |
Re: Plans for Maemo Mapper v2.0
Quote:
|
Re: Plans for Maemo Mapper v2.0
Quote:
|
Re: Plans for Maemo Mapper v2.0
I have found MaemoMapper to be a little confusing, and I consider myself fairly a savvy user. I think that using some of the screen real estate to put some basic buttons/shortcuts on the left or right side would be excellent.
Some examples of these shortcuts might be: (Favorites) - Ala Google Maps, tap it and you get a list of common addresses you use. (Map to from GPS Location) - So it can map to an address from where you are with just a few touches. (Map from Address to Address) - With easy access to favorites. I realize these are built into Google Maps, but I think making these easy to access with a huge touchscreen would be way better than having to use my tiny Blackberry to navigate all the time. You might even be able to tap into their API so you don't even have to code the route calculations. With the map caching, superior screen and input I think your application can definitely outshine the competition. I will gladly donate to the 2.0 (buy gunite a ferarri fund) if the app moves in that direction. |
Re: Plans for Maemo Mapper v2.0
Hi Gnuite,
To me, the most important feature would be to be able to do routing and rerouting (especially rerouting) without an internet connection! The prices of memory cards are decreasing and their capacity is increasing. Thank you for version 1. |
Re: Plans for Maemo Mapper v2.0
As for the database vs filesystem issue, I would say go for database only.
If I want to look at individual map pictures I could just get/make a tool to extract from the database, assuming the format is known. |
Re: Plans for Maemo Mapper v2.0
The only advantage that I see in the filesystem method is that I can merge map sets together easily. E.g. I can take a low-detail basemap of the country and combine it with detailed maps of several cities by copying them all to the 770.
But if there was an easy way to merge and edit the databases then it could prove to be more versatile approach for that sort of thing too. |
Re: Plans for Maemo Mapper v2.0
Quote:
I would have no problem stuffing a 4gb card specifically for MaemoMapper if it provided those functions. I think ease of use, and routing should be your targets for 2.0! Concerning the database style... I wouldn't mind moving forward and just going with the one file. I've found moving thousands of files back and forth to be painful for my memory cards and I wouldn't miss any merging benefits the old style provides. |
Re: Plans for Maemo Mapper v2.0
Quote:
I will start with a Maemo-specific design, since that is my priority, and I'll just keep in mind that I'll want it to (someday) work with normal GTK as well. Then, I'll change the maemo-specific bits to build differently depending on the target environment. |
Re: Plans for Maemo Mapper v2.0
Quote:
As others have mentioned, the existence of an alternative doesn't refute the desire for a free, open source solution. If that were the case, Maemo Mapper never would have come about in the first place, as GpsDrive had already been ported to maemo. |
Re: Plans for Maemo Mapper v2.0
Quote:
I'm still investigating alternatives, though. Someone mentioned CDB, but that's less useful in a constantly-changing map cache scenario. Maemo Mapper already uses sqlite3, so if I can reuse it for the map cache (POI would still be in a separate database) with sufficient performance, I will prefer to do that. sqlite3 (or any relational database, for that matter) could also give me the opportunity to implement more complex features, such as "aging out" (based on either time or fixed-disk-space) of older, not-recently-accessed maps . I'm also open to suggestions. |
Re: Plans for Maemo Mapper v2.0
Quote:
|
Re: Plans for Maemo Mapper v2.0
Quote:
It sounds like most people don't care too much about that drawback, though, so I'll just release a FS-to-DB converter, and that will be that. Quote:
|
Re: Plans for Maemo Mapper v2.0
I also vote for the database style map storage. The "one file per tile" just is too much of a burden for the file system.
Feature request: Can you add selectable "smart" auto scale function to the Mapper? How this might work is that the zoom level depends on either: 1) Speed: Maybe at highway cruising speed the screen would cover approx. 5 minutes (selectable?) of travel. Once the speed increase/decrease the map zooms automatically. (However, stop and go traffic might become problematic...) 2) Approaching waypoint/turn: The map zooms in automatically so that the turn can be seen in the screen with maximum detail. For example once the waypoint is within the 5 minutes window (as in 1 above), this zoom function takes over so that the waypoint is always seen in the screen. Just one idea that could be automated, which I found doing manually when traveling with the MaemoMapper... |
Re: Plans for Maemo Mapper v2.0
Quote:
Quote:
|
Re: Plans for Maemo Mapper v2.0
John, I would like to know (either you ever want to do it or not*) how it would be possible to calculate routes offline, given that MM just loads images that could be highways and streets but could be also 2-century old maps, or satellite images (on which the streets are quite hard to identify on zooms less than 2 or 3).
Will we ever see something like this? *I say this because lots of people are asking for it, and because it is already into the garage feature request, despite I usually download my route instantly through the phone or the day before, through my internet connection at home. |
Re: Plans for Maemo Mapper v2.0
Quote:
Quote:
|
Re: Plans for Maemo Mapper v2.0
Quote:
For the same reason that Maemo Mapper can't just draw a map (instead relying on bitmaps from repositories like Google Maps), Maemo Mapper can't just calculate routes (instead relying on online mapping services like Google Maps). |
Re: Plans for Maemo Mapper v2.0
Quote:
I know that you can recalculate the route distance to end, so maybe there is a way to recalculate the distance_to_next_waypoint_that_has_a_description, and just add to the flite string a "In 500 metres", then repeat it once it is close. (I know, it is quite a stupid explanation, 'cause all the things, once explained in words, are simple. "Ok, please, give me a program which can drive me here and there. It's not that hard. Take the GPS data, draw them onscreen, and underneath you could place some tiles from Google maps!") |
Re: Plans for Maemo Mapper v2.0
First, thanks for a great program. I like the idea of a DB-based system, assuming it leads to performance (speed) increases, memory cards are cheap enough that I'm not too worried about using them up (currently have 1.3 GB of map data).
One feature I'd like to see is POI labels. Right now, as far as I can tell, you can see POIs below a certain zoom level, and they can be color coded, but unless you click-and-hold, you don't see any detail about the POI. I'd like to see an option to put a text label with the name and (optionally) the street address, GPS coordinates, phone numbers, or other custom fields next to them. This might be a real performance hit, especially if you've added tons of POIs from some of the online sources, but it would be nice to have it as an option. |
Re: Plans for Maemo Mapper v2.0
Hi,
I agree, keeping the map tiles in a db will be better. The current version of maemo-mapper works perfect for me and a new improved version will be most welcome. Here is what I wish I had in the current version (feature requests for the v2): * download the route from address to address build in (when you enter the address, on the same dialog, put a check box for "add this address to the POI") * download route from *tap on the map* to *tap on the map* (to be able to select the start and end of the route by tapping on the map. * labels for poi with the option to be visible always or just when you tap on them. * the option to have the route text on the screen all the time ("Turn left at Portland Rd NE") and a small button (semitransparent?) in a corner and if you press it you'll get just the turns (something like in the attached picture) http://www.gpsonsale.com/garmin/Images/i5-2.jpg the name... what about gpsm? gpsm.org is still available and thats about it! :-) Thanks gnuite for a great application! I almost forgot! The option to download the zoom levels 1 3 5 7.... If I'm not wrong, right now you can download only starting with the zoom level 0. -ioan |
Re: Plans for Maemo Mapper v2.0
I would like to add my vote for a feature to pre-populate the map cache. Here in Canada the cellphone data plan charges are ridiculous, so I would prefer to pre-populate the cache when I'm within range of a WiFi access point. Before going out driving, for example, it would be nice to be able to drag-select an area on the map, then have data for all zoom levels for that area downloaded and added to the cache.
Thanks for a very useful N800 application! |
Re: Plans for Maemo Mapper v2.0
Quote:
|
Re: Plans for Maemo Mapper v2.0
Quote:
|
Re: Plans for Maemo Mapper v2.0
Quote:
|
Re: Plans for Maemo Mapper v2.0
Quote:
-ioan |
Re: Plans for Maemo Mapper v2.0
Quote:
|
All times are GMT. The time now is 13:32. |
vBulletin® Version 3.8.8