I have been asked to develop a new, small, custom-specific CRM (Customer Relationship Manager) that will be used mainly on Linux desktops (compatibility with Windows and Mac OS X would be appreciated but it is not required).
This seems to be a good opportunity to try the new Vala language and some of its libraries (most notably libgda and the rest of Gnome-DB) but, of course, I still have to deliver a working product to the customer in time so... I'm still scratching my head and wondering.
To develop this application I would need:
A "glue" language (Vala itself). This is OK.
A GUI Library (GKT+ 2.X or 3.X). This is OK.
A database abstraction layer (libgda). Here I have some doubt.
Maybe a MVC framework like Bakery (Bakery 2.6 seems to be working
with GTKmm 2.4 only. It does not work with the GObject-enabled GTKmm
3, as long as I can see.).
Maybe a ORM like Hiberlite (libgda supplies data-aware widgets
and other tools but it is not a full-blown ORM, as long as I know).
At the moment, I'm confident about the first two items, only. Even the real amount of Vala support for libgda is not very clear to me (The ValaDoc describes as supported the interface of a old version of LibGDA while the Gnome-DB website says that the new 4.2 and 5.X versions of the library are GObject- and Vala- enabled). Most likely, Bakery and Hiberlite would not be available any time soon for Vala.
The nearest alternative seems to be:
C++
GTKmm (2.X)
Maybe Bakery 2.6
libgda
Maybe Hiberlite
A more mature stack but... maybe so mature to be fated.
Hence: would you try Vala for a new database-centric project like this?
Or would you wait for a more mature and more rich Vala ecosystem?
Thanks
Vala just means native compilation without requiring a framework (and versions) to bother about. Connecting to database still looks premature and definitely undocumeted (that's how I came to this post). Besides, there is no IDE. Glade is not really and IDE, but an interface designer.
Try out Lazarus and you will be in for a surprise, how conveniently database front ends can be developed. Pretty mature, native compilation, ready to use third party components, database support right through the IDE, options of using Gtk or Qt.
And it gives native exe's on Windows, Linux and Mac. Nothing comes even remotely close if you are developing cross-platform database front ends. Development time would be a fraction and performance comparable to C, if not equal.
Related
I have a few questions that I hope clarity and facts can be fed all of us about this. First, the last release of ironpython had the feature of running on the mobile platform highlighting that on mono/android it runs best and not so much on the other platforms which are IOS and Windows phone. So does this mean one doesn't have to use the SL4A? Can i argue that monodroid hooks into the core of what android is capable of and so might give a more robust access/implementation/rendition of apps on android?
Next, I just want to be sure of this: As with python you can create full fledged desktop applications with ironpython right? Cos everywhere (almost) I see ironpython they refer to it as a scripting tool and how you can script aspects of excel etc which has prompted me ask such a question. some say that the speed of applications written using ironpython is not that great and I was arguing asking for why that opinion is held by those who were speaking and no one could say anything worthwhile.
Lastly, with the movement at novell and xamarin, I have not understood where mono stands in their release schemes. I have a mac and I installed a recent version of mono and when I typed ipy I saw something like mono 1.1.0 or something like that. Does anyone have a clue of what is going on with it.
I fell in love with python from learning ironpython and I think the ironpython concept is just sheer brilliance its a pity microsoft esteemed F# other it.
First off, for future reference, multiple questions should really be, well, multiple questions. :)
On IronPython for Android: it works slightly better than on other platforms (where it doesn't work at all right now) but it's still very, very early code. There are some limitations with what you'll be able to do (mainly, you cannot inherit from Java classes from dynamic code). It's very experimental.
On desktop apps: You can absolutely write full apps in IronPython. There are some samples that show how it can be done. In particular, PyWpfSample and PyGtkSample. Startup times are not great, but the actual runtime should be just fine for most GUI apps.
I'm a .NET developer with some questions about a small Mac project coming up.
We are going to be creating a small program for Mac OSX. The software will need to have a simple UI (1 screen) and will need to consume a WCF web service.
Should we code on our Windows boxes or on Mac machines? We have a couple Mac Minis, but we mostly do Windows development.
What IDE/dev environment should we use? (Eclipse, Xcode, etc.)
How does Mac software deployment generally work?
...
No one has mentioned Java/Eclipse. Isn't that an option?
Here are two options:
Write in Objective-C, using Xcode on Macs
Write in C#, using Mono on Windows or on a Mac with SharpDevelop
The advantage of #1 is the resulting app will be much easier to deploy to others, but you might have trouble consuming the WCF service. It depends on what exact interface you are exposing and what the types of the parameters are.
The advantage of #2 is that you can use your C# skills. I don't know the state of WCF in Mono, and the GUI may not look native.
If you go with #1, the main issue will be using Objective-C. Since you are targeting the Mac, make sure you enable the Garbage Collector, or else you will have to also learn memory-management. If you use Xcode 4, it's similar in spirit to Visual Studio, although the details are different. For a simple project, following the tutorials will probably be enough -- I suspect the hard part will be interacting with WCF. To make your life easier, make the interface to the service extremely simple (simple parameter types and returns).
Your question is overly broad. You have two options:
When in Rome...: download Xcode, look at example apps, read Apple's documentation, find some tutorials, and learn Mac app development the usual way.
The Devil You Know...: look into stuff like the Mono project that will let you write a Mac app in C#, so you can use your existing libraries and don't have to learn Objective-C.
Which is better depends on your needs. If you just want to get a small app done and don't have a lot of time/money, I'd go with #2. If you're concerned about quality (compatibility libraries always have their quirks), I'd go with #1. My gut says #2 is better for you.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 7 years ago.
Improve this question
I have questions on true cross platform development and if it is possible to use ONE platform to develop software.
I have a few applications that I write that work on OS X, Windows and Linux. They all use native platform toolkits with some shared source across platforms. I have to boot to each platform, checkout, compile, etc.
I have in the past tried GUI toolkits like QT (I bought a commercial license for Win, mac, Linus, but very expensive and paying it year after year became hard). I have tried WxWidgets, development moves slow.
So what I am thinking about is there a way to run a single platform and cross compile for other platforms so I can build my products from a single platform.
I favor OS X, is there a way to write apps for OS X, Windows, Linux from OS X where I just need to test on each platform respectively. I have found information on cross compilers, stub compiling, etc, etc.
Does anyone have any thoughts? Is this even possible? Would someone make a killing creating such a beast?
In my company, we use the Mozilla Framework to develop cross platform dictionary products. These applications are based on the Gecko/Necko Browser and most of the development is done via Javascript, HTML, CSS, XSL, XUL,... Of course, our homebrew Search engine had to be cross compiled to the three platforms (with some other code which was first done in JS and then ported to moz' C++ for speed reasons).
In the end, we have a reasonnably cross-platform solution: Our developers mostly work like power web-developers (think client and server side at the same time). Because the Gecko is the same on the three platform, we know it will predictably look and behave the same way (except when there are bugs, but the platform is now quite mature on this point of view).
Our R&D knows how to port slow javascript code to rather quick portable C++ code (you do Mozilla code using the NSPR library, a cross-platform lib). Testing has always to be done on the three platform however, although with time and experience, you get to know what will break where (Font support, Audio support, Flash Support)...
Today, you should probably have a look at XULRunner which is really the Mozilla without any real browser interface (in 2002, we had to dismantle the Mozilla Browser to repaint it in our first product colour). Of course, it works well for simple applications but if you wish to make OpenGL, 3D Audio and other nifty things, XULRunner will appear too limited.
I believe Blizzard has some cross-platform framework for that kind of thing... :)
Good luck!
Pierre.
Trying to develop applications on a single platform that will be used on many platforms isn't a good approach. At best you'll make applications which feel alien to users of the platforms that you didn't develop for.
More likely, you'll run into subtle compatibility problems in areas which you never forsaw. Java is probably the best way to go if you want to go down this route. Cross compiling will lead to autogenerated code that will be a nightmare to debug and maintain.
Certainly, you may be able to use tools for porting in some cases, but I don't think that this is a problem that you can just provide an automated solution for in general.
Well, to be honest, the only guaranteed way to build an app to be cross platform in one go is to use Java, but it requires a rather large runtime to be installed first.
However, if that's not an option, I would recommend keeping an eye on recent developments in Qt - it is now available (or should be soon) under the LGPL, which presumably means you don't need to pay for it anymore. Using GCC with Qt, I've found, works perfectly when cross-compiling on different platforms, as long as you only use Qt's classes/code/objects and make sure that any non-Qt code you create or use is capable of being compiled cross-platform.
You don't mention any specific programming language (but I'm guessing C or C++ because of the GUI toolkits), so it's kinda hard to give a good answer to what you are actually asking.
If you want "true" cross platform, I would first consider if it is possible to solve your problem using a language that is less platform bound. Python, Java and plenty of others allow you to write on one platform, and run on many.
If you still want to use C/C++, GCC gives you the option of cross-compiling, and if you combine that with QT (which will soon be available under the LGPL) you should be able to get something working.
Java has tried to do the write once, run anywhere. It works well in some situations, but there are too many "unique" things in an OS. GCC provided the ability to cross-compile applications, but you run into the same sort of problems. The code will just have better performance. The RIA approach seems to work, but it doesn't feel like a native application.
Even using a cross platform GUI toolkit will not remove all your cross platform problems. There's more to an app than GUI, after all. Cross compiling will never be able to catch all the issues that a native build will catch, either. If you're going to support multiple platforms, you're going to have to at least test on each platform. But hopefully you can perform native builds with all warnings turned on, in addition to testing.
In the past few months, I found a few great videos exploring this topic of cross platform development. I hope you find them useful as well.
QTCreator
http://www.youtube.com/watch?v=aYiPvM7ZRHA&feature=channel
FlexBuilder
http://www.youtube.com/watch?v=_O_xDXRsh3Y&feature=channel
Mono / MonoDevelop
http://www.youtube.com/watch?v=U6VG_Z0aRek
I've personally had success using Silverlight / MacOS X
http://screencast.com/t/if8xenkt
RealBasic
http://www.youtube.com/watch?v=GWipoBeKSRk
For Cross Platform Development you can use Phone Gap, Appcelerator (Titanium), Corona... This all provides a framework based on JavaScript and finally able to run on different platforms. What I am using is Titanium for Mobile Development which allows us to develop applications with Native Codes.... (which is very good advantage). Phone Gap is giving a web app which is not native in general... This softwares are used for development of cross platform supports... with support on Windows, Linux and Mac OS.
Based upon my personal experience, I believe you should adopt Java: you will avoid many headaches.
You develop in whatever platform you like and you deploy in all other platforms with no need of compiling for each target platform, as you mentioned.
For example, I develop under linux, I hit "compile" only once and the resulting file is ready to be run anywhere (windows, mac, solaris, z/OS, you name it). A double click will run it on any platform (Java runtime must me installed, but most users have it installed already and if not, it's a matter of downloading, installing, "Next>", "Next>", "OK").
If you choose the "Java Web Start" deployment method, it gets even easier: the user just clicks the launch button on a webpage and the application runs (if the proper JVM is installed according to what specified in the JNLP descriptor) or the user gets redirected to the Java download page (if no suitable JVM is found).
GUI development (with Swing toolkit) is easy and powerful, especially if you use the right tool (i.e. Netbeans IDE).
I'm looking at rewriting an eMbedded Visual Basic app I wrote years ago. I'm unsatisified with it because of various problems clients keep having with it now and then over the years, mostly along the lines of the app not loading anymore because a required dll/activex control has gone missing! This is so frustrating and naturally difficult to debug when a client is using it far away. In alot of cases reinstalling the app doesn't fix the problem.
My preference would be to rewrite it in C# since I'm comfortable with C# and DotNet, but I'm also open to other platforms like blackberry or iTouch/iPhone so long as the platform can support maps and GPS. I'd start rewriting it in C# now but I can't be sure that I won't have the same problems in .net.
Has anyone else had similar problems with eVB apps which have gone away/persisted when moving to CF DotNet? Or would you suggest a different platform again?
Edit: Note that I wish to move away from eVB anyway, but if I move to CF DotNet I want to make sure I won't have the same missing dll/control problems.
I recommend .NET CF strongly, especially if you already know C# and .NET. Mono has been ported to the iPhone, so it is possible to write apps that will run on Windows Mobile and the iPhone. No Mono for Blackberry (yet, if ever), so that's a definite limitation. I personally can't stand Blackberries (I have both a Blackberry and a WM smartphone and the Blackberry makes me want to hang myself), but they do have a huge user base.
You should have migrated away from eVB years ago, but that's water under the bridge. If you want to continue targeting Windows CE/ Windows Mobile I'd recommend going to the CF - language is irrelevant, use what you're comfortable with.
There's no way to guarantee that whatever your "missing DLL" problem is won't happen again, since we have no idea what DLL went missing. If it was a 3rd party control, then you're at the mercy of the market. If the provider survives, it's likely their control will.
If you want to target iPhone/Blackberry then Java is more likely to be your language of choice - the tools I'm not as familiar with. Eclipse for Blackberry - iPhone may have their own tool.
As for Silverlight, you might look at it, but so far it's just way too slow to be a viable platform, at least on any WinMo device I've ever seen. We've delevered many, many CF apps for all sorts of verticals and have never had any usability problems (though we've been doing it a long time and know every limitation and what we should and should not be trying).
I suggest you take it one step further and look at Silverlight. One of the premises is that it's a more long-term-stable, portable, lightweight download and install, and it hasn't gotten krufty yet.
I think it has the potential to be the next VB for embedded. One of the difficulties with CF is that I've found it to be an insufficient subset of the real thing.
Another option is NS Basic/CE. It's highly compatible with eVB, so you will be able to keep most if not all of your code. The product has been continually updated so it runs on current devices.
The installer that NS Basic/CE creates includes all the dll files your program requires, so they will be included on installation.
I want to update/upgrade the standard Leopard install of Sqlite3 to >3.5 to use the new sqlite_xxx_v2 methods from a Cocoa project.
I can't seem to find any information on how to do this. Does anyone have any tips or a site that outlines the update procedure.
Also is 3.5+ supported on the iPhone. I understand it's embedded so shouldn't be an issue...
What you want to do is grab the amalgamation sources from http://sqlite.org/download.html . Then just compile that into / add it to your project. You don't want to replace the system sqlite- that'll have unintended consequences in other applications. Plus, I'm pretty sure the system sqlite isn't a stock sqlite... Apple has probably made their own modifications to it that core data relies on.
You can read up on the amalgamation stuff here: http://sqlite.org/amalgamation.html , but in short: '''The amalgamation is a single C code file, named "sqlite3.c", that contains all C code for the core SQLite library and the FTS3 and RTREE extensions'''
I'd also suggest not using the sqlite calls directly, they weren't designed to be used that way (says the author of sqlite). Instead, there are a number of cocoa wrappers out there, including fmdb: http://code.google.com/p/flycode/source/browse/trunk/fmdb/ (which I wrote) :)
-gus
You don't really want to upgrade the system version of SQLite on Mac OS X. The reason is that all Mac OS X software is qualified against the versions of the packages that it includes, as built by Apple's build process. Installing a different version of a package, or even building the same version yourself but doing so slightly differently than Apple does, may result in a system that behaves unexpectedly.
Finally, if you embed a newer version of SQLite — or any Open Source library or framework included with Mac OS X — into your own application, you should be sure to integrate the Darwin changes for it from Apple's public source site. That way you can be sure you'll get as close to the same behavior as possible from the library you've built yourself as the version Apple ships, which is especially important when it comes to functionality like file locking in databases.
I don't believe i've updated my version, but it's currently at 3.4.2, and i'm able to use the new methods with the current version.
And i'm running 10.5.5 with the latest (public) iPhone SDK.
It would likely be easier to just drop the library into your project and link it in from there.