I'm starting to build an app, a mobile device. Performance is quite important (as it is in anything else) and I'd like to maximize it.
The database will be used pretty often. I'm looking to access it through the QtSQL library of the C++ Qt library.
I have experience in MySQL, but don't think that's the right choice.
So I'm looking at:
PostgreSQL
sqlite?
CouchDB?
The operating system is a stripped-down build of Ubuntu. Need regular CRUD as well as opportunity for optimization.
Thanks,
Mark
Since Apple uses sqlite on the iPhone and iPod Touch, I would take a good look at that. I assume that they will have evaluated everything available to them at the time to get to this conclusion.
Since sqlite is FOSS and "endorsed" by a well-known vendor of mobile devices, give it a try.
Did you consider Sybase Ultralite?
I'm a new student developer, but it appears that SQLite might be the best bet to go. Aside from being what seems to be the preferred iOS online database... it seems to have a lot of support with Android and the upcoming [Windows Phone 7 Platform][2] (via C-sharp-sqllite). Which seem to be the most dominating (or in the case of Windows Phone 7 - upcoming) platforms for most web development. Blackberry and webOS doesn't seem to have an issue either.
Related
I'm planning web application and considering silverlight as development platform. Will it help to solve browser compatibility issues? The app intended to be used on desktops only (no mobile).
Yes, it will solve browser compatibility issues, and could work on both Mac OS and Windows with the very same code.
The only drawback is that, the first time your user connect to your application, he will need to download the Silverlight plugin.
Awesome you would say? Well, unfortunately some people that probably never try to do something like image processing or advanced line of business application in a browser decide that plugins are not so cool and that you would be able to do the same thing with the magic power of HTML5.
We are still waiting to have the same possibility in HTML5 that we have in Silverlight or Flash, but plugins are already dead. At least as long as no big compay want to push them again.
So, my advice would be: don't start a project in Silverlight. You will have problems, even if you do not target mobile. For example it becomes harder and harder to find compatible good tools (like ReSharper, NCrunch, or even just a decent unit testing library). And in further release of Windows and Mac OS, it will probably not be supported at all (IE for Windows RT already does not support Silverlight).
Sorry man, Silverlight is dead, you arrive after the battle.
If your developing your application for an Intranet, I would say Silverlight is an excellent choice.
If you are developing for the Internet, use an HTML based language
We were beginning to start on an enterprise application using silverlight.
However after reading this post we doubt whether it is the right choice going forward.
The post says that, according to
Microsoft's declaration in PDC 2010,
Microsoft has changed their strategy
regarding silverlight and they no
longer view it as their technology to
deliver cross platform applications.
Instead they are targeting silverlight
as their development platform for
Windows phone 7.
Is this correct? Should we still continue with silverlight or go back to ASP.NET WebForms\MVC?
The Scope of the application is basically intranet with Windows 2008 servers and Windows XP and Windows 7 clients. However a subset of functionality needs to be available to the external users over the internet. There we cannot have any restrictions on what OS users can use.
based on the info you gave, I can't conclude whether silverlight is the way to go. But what I do know is that a number of Microsofties wrote some blogposts about the things said about Silverlight on the pdc. For example John Papa, Bob Muglia and Scott Guthrie.
Update about the scope
I think you already gave the answer when you described the scope of the application. A part of the application will be available to external users and you cannot have any restriction about the OS they are running. With that requirement I think Silverlight is not the best way to go. Not because the rumours about its future but because of its platform indepency. What are the reasons not to go for a ASP.NET/web solution? Silverlight doesn't work on each OS whereas plain HTML will work everywhere. (ok you need a descent browser)
Although for a good advice I'd need more information about the application.
Basically the question you have to ask yourself is this: do you need your application to be used on every platform, i.e. Windows, Mac, Linux, misc. flavors of Unix, IPhone and other mobile platforms?
If that's the case, then a web based solution is the way to go.
If Windows, Mac and partially Linux is enough, then save yourself and your team a lot of pain and use Silverlight.
In my opinion support for mobile clients is the key factor in your decision.
For sure the right platform for Intranet, Enterprice applications Is Silverlight. It is
stable, performs extreamly well, the environment and the development time is huuge less than web application development, the end User Experience is much better and so on and so forth... Once you want to show part of the system out the the intranet - just create some specific target modules that will address the needed audince. You won't have the universal "Reachfull" solution, that will target everyhing, you'll always need mobile versions or other devices and so on. But once you've built your project the right way with Services (same services that the Silverlight app will consume), it'll be easy job to consume them with new UI.
Hope you will choose Silverlihgt.
Silverlight is a great technology, but the Microsoft does not develop it anymore. So as a technology is a great decision. But if you want to make a Silverlight app usable on a NOT supported platform (e.g. Android or iPhone) you have to use 3rd party services. For example http://sl2html.com
When developing with Mono for an app to be run on Windows and Mac OSX (and maybe Linux) which would you suggest, WinForms or GtkSharp for the GUI and why?
Specific examples and success/horror stories would be much appreciated.
Cross platform development is a nice idea, but to be completely honest I have never seen an application that looks really good outside of its native environment. That is why I think, that if you really want to offer good user experience you should use native toolkits on all platforms that you want to support.
Of course, if you just want a proof of concept, then WinForms on a Mac (or Linux) are alright but if you want an application that competes against other native solution then you really should consider writing a separate native frontend for each supported platform.
A good example how that was done is MindManager from Mindjet. Their Mac version is a full, from scratch rewrite of the interface. The result is an application that gives Mac users the experience that they learned to expect on their platform.
So, back to Mono. There are Cocoa bindings in Mono. Personally, I have never tried using them, but if you want to target Macs with your application, they are certainly worth a look. It would be really great if they could be regarded as a valid choice for UI development on a Mac.
When deciding whether or not to use a cross-platform architecture, gague your requirements of having a tight UI against agility toward change and development time. Using a platform-specific UI framework is (almost) always going to give you better looking and better performing results than you'll get from a cross-platform one.
GtkSharp requires X11 on Mac OS X. Mac users will find that off-putting. I haven't tried Winforms on OS X, but mono's Linux winforms support was pretty immature (buggy) and slow. I don't think there is a mature .net GUI toolkit for OS X at this point, at least nothing at the level of Swing or SWT.
What platform will most of the users of your app have? If most of the users will be Windows users, use Winforms, but if most of the users will be Linux users, use Gtk#. I wouldn't know which one works best on OSX.
Read this article also: http://www.mono-project.com/Gui_Toolkits
I am willing to develop a mobile application. I wish to have something working for android, windows mobile, symbian and blackberry.
Which is the best way to do that?
I had read here:
You could aim to wrap the sections of
the platform specific APIs (iPhone SDK
etc.) that you use with your own
interfaces. In doing so you are
effectively hiding the platform
specific libraries and making your
design and code easier to manage when
dealing with differences in the
platforms.
I was hoping there exists a framework that does this for me, but it doesn't exist or I didn't find any.
I feel that sort of things will make my code harder to maintain and perhaps it's better to have one version for each platform.
Anyone with experience in the field?
Another links of interest:
most-promising-mobile-platforms
long-term-potential-of-iphone-windows-mobile-development-platforms
Does Java not count (in various guises)
Java on Symbian
Java for Windows Mobile
Java on Blackberry
Android Java Virtual Machine
It should be simpler to manage API differences in a consistent language/runtime platform where capabilities can be assessed in-code ... and configurations of code made at build-time.
As much as I dislike Java, it is fairly ubiquitous. As for the iPhone ... apart from it being much hyped and locked down ... you can get Java to run on jail-broken phones ...
What happened to Apple's open and friendly appearance? The cynic-inside knows the answer ;)
You might want to look into PhoneGap (http://phonegap.com/). From their own description page:
PhoneGap is an open source development tool for building fast, easy mobile apps with JavaScript.
If you’re a web developer who wants to build mobile applications in HTML and JavaScript while still taking advantage of the core features in the iPhone, Android and Blackberry SDKs, PhoneGap is for you.
In addition to using JavaScript, it supports JavaScript acccess to native controls and features of the phones (GPS, accelerometers etc...).
There really isn't any magic bullet that I'm aware of. Even within just the Blackberry platform, there are tons of different devices with different capabilities, screen resolutions, etc. And that's just from one, single manufacturer; Symbian and Windows Mobile are likely even worse.
The answer is likely that you should focus on relatively new and consistent platforms (accordingly with very few and all pretty much similar devices), like Android and iPhone OS, if you really want to reduce your code forking and maximize your audience.
My advice will almost certainly change within a few years when there are nine different iPhone OS devices and two dozen Android platforms.
The first question to ask yourself is if you need a native application, if you do not then designing a mobile web site solution should give you the most cross compatibility, failing that I would make a iPhone and J2ME solution (the J2ME can then be ported for Android relatively easily) for the greatest coverage of users
Or investigate Movilizer. Supports iOS, Android, WinPhone, WinMobile, Desktop PCs, embedded devices, ... and many more. It uses a design once run anywhere approach.
http://www.movilizer.com
try out different cross platform dev tools,
Developing cross platform mobile application
I need to develop some programs for mobile devices but haven't decided the platform to build upon. I'm looking for Palm or Pocket PC devices that have Touch screen and Wi-Fi connection and are cheep because I'll need to buy several of them.
I don't really need camera, mp3 players, video players, pdf readers or anything else since the apps are going to be simple data collection to feed via wireless to a server database.
I'm proficient with C and C#. I could learn Java if I had to.
What devices do you recommend? Linux devices maybe?
PS: Changed the title because I don't want a flamewar between platforms. Please, don't answer with Windows Mobile sucks/rules. I'm looking for devices instead.
Thanks
Windows Mobile
It supports C#, and Visual Studio comes with the mobile SDK. So if you know C# you probably already have the tools you need. And in spite of the iPhone/iPodTouch buzz, the Windows Mobile deployment is still 10X greater.
In order of preference
Neo Freerunner
Maemo & the N800 (cheap)
Beagleboard
If you are comfortable with Visual Studio then programming for windows mobile is extremely easy. The SDK for mobile comes with emulators for all the latest and popular versions of windows mobile- and you can even debug on teh device itself using a USB cable.
On windows mobile you have a choice: Develop a .Net application or develop native (likely MFC based). Either one gives you a great development environment.
As far as iPhone development goes- you would need an apple computer to install and use iPhone SDK- and you can't run an iPhone app on your phone. You would have to go through the process of getting it registered with iTunes for you to install your own apps on your own phone!
When I first started playing with mobile development I had a few questions:
Can I develop using my favorite IDE- Visual Studio. Will it be as easy as developing a desktop app: yes.
Will I be able to access the internet from my application without 'unlocking' or in some other way enabling the phone that was not intended by the service provider? yes.
Will I be able to access device specific functionality such as GPS easily? Is there good support for doing so within the API? Yes.
You should probably target the Windows Mobile platform. The Palm platform is rather archaic and no longer widely used. The development environment is also rather spartan, while Microsoft has full IDEs available for Windows Mobile development. You might also consider the iPhone/iPod touch platform - I have a feeling the number of devices will multiply at an exponential rate and I've heard that developing applications is much easier due to the completeness of the system stack.
You should probably at least evaluate the Apple iPod Touch. It certainly meets your basic "touch screen + WiFi" spec, and your users presumably won't object to all the the other nice features that will come along for the ride.
I don't know what your cutoff for "cheap" is, but $299 for the base model seems pretty reasonable for a high-quality touch screen and WiFi in a pocketable device.
Windows Mobile and CE used to suck, really, really badly. These days however it's definitely passable and worth checking out, especially if you code C#.
Just remember that it is the baby brother of the full framework and has nowhere near enough toys and throws a lot of NotImplementedExceptions. :)
Blackberry publishes its SDK on its web site. Its apps run J2ME, so with some Java experience it shouldn't be too difficult to get started. They also give you an emulator. Disclaimer: I have no experience in writing Blackberry apps, but I looked into it once.
I would not recommend a PalmOS based handset. I have written code for PalmOS and it's about as painful as writing raw Win32 code in C. Since Palm has switched its high end handsets to Windows Mobile, PalmOS will just remain stagnant and only run on the slower, less capable hardware.
If I were to write a mobile app, I'd agree that Windows Mobile is worth checking out.
It all depends on the users who you are targeting at, If you are looking for a wide market then you should be fine with J2ME/Blackberry . However most of them lack the touchscreen and wifi features ( The HTC range of phones [WIFI/TouchScreen/Windows Mobile] have a JVM built with it),so it would work on most of the Windows devices also.
If you are making a more niche product, moving with the current buzz 'iphone' will be good . Windows Mobile is also worth checking out
The best option here would be the Neo Freerunner, with that device you can build a dedicated unit were every aspect is made especially for you're needs. The Freerunner is WiFi enabled, and has a touch interface. If you use the Qt SDK, a lot of the work is already done for you. It comes complete with emulator, as a Live linux cd. You can run in a WM, such as wmplayer. Everything is included.
I'm not gonna lie, it will take tweaking. But the final product would be really nice and intuitive.
Looking at Windows Mobile devices, your requirement of touchscreen pretty much sets your pricing at the higher end of the spectrum. You'll get those things you say you don't need just because of that.
Here's expansys's selection of touchscreens.
Mobdeal is a handy one too as that effectively filters all phones by features.
I've developed against the HTC TYTN 2, HTC Touch Diamond and randomly a PSION Teklogix Ikon
There's generally very little difference between these models, some manufacturers have SDKs that can help sometimes.
I think your cheapest option will probably be something like getting HTC TYTN 2s on ebay. They're pretty old now (hence cheap) but have Wifi, camera, touchscreen, qwerty keypad all the things you seem to be after.
you can target iPhone "touch" platform with Apple's iPhone SDK. the development environment requires a Mac, but you can get the entire IDE + tool chain + excellent debugging and profiling tools for free. And the free documentation is top notch.
As a registered iPhone developer, it is free (no cost) to target the simulator, which is sufficient for most learning and development you'll likely need to up front.
To target the actual hardware device (and up to and including release/selling your app on the Apple's AppStore) is only $99/yr. If you got an iPod Touch for your hardware target, most of the SDK applies and you are not tied into a service contract for an iPhone.
iPhone app development environment is in Objective-C, but it is a really productive, object-oriented environment so do not concerned that that may be a language you are unfamiliar with.
If you decide that your mobile app(s) would be better suited as webapps, the iPhone/iPod touch platform again is an industry leader in this space, and you have the additional benefit or being able to target other mobile platforms (and not necessarily be tied to one mobile SDK).