I am writing an iPad app that will be expandable with new items via in-app purchasing. For example, my current plan is to have a jpg pattern and a matching plist file with the parameters I need to expand that pattern into a full picture.
The user will select one jpg/png from a list of small thumbnails - the list is held in Core Data - and the app will find the matching plist for displaying the jpg/png correctly. I'll only have about 10 of these open at one time. But I could end up with storing 1000s of jpgs and plists.
Does storage of lots of small files cause app problems?
I'm going the plist way, rather than storing the parameters in Core Data, so that if I need to add parameters later, I don't have to migrate the database, just change the access in code. (And when I'm creating the patterns, it's easier to concentrate on a plist file rather than a Core Data row.)
The app seems to work really well at the moment, but I'm worried about futures...
My app does also use Core Data for other things, so I could change over if the app will get bogged down with number of files.
Thanks.
Saving a large number of small files is not a problem as long as you have a well thought out means of naming and tracking the files.
Remember that the user does not have the same flexibility and ease of file management on a mobile as they do on non-mobile platforms. Designs that work on non-mobiles are unworkable on a device used on the move with one finger.
However, when you say:
And when I'm creating the patterns,
it's easier to concentrate on a plist
file rather than a Core Data row.
... the use of "row" suggest that you haven't fully grasped Core Data's utility. Core Data doesn't use rows, columns, tables or joins. It's an object graph management system that sometimes uses SQL way behind the scenes.
Core Data is designed to handle data in a way that meshes seamlessly with the rest of the object oriented API for the UI and other services. When you use other data management systems like plist, you will most likely end up manually duplicating a lot of Core Data's functionality anyway.
Related
when I develop an iPhone App (Time Tracker, ToDoList etc) I never know whats the best way to deal with the data. Once I used a plist, next time sqlite or CoreData.
How do you decide whats the best for your project?
(Only talking about data management)
For Example if you want to develop:
Time Tracker App > Is PList your choice?
RSS Reader App > CoreData?
Photo App > sqlite?
EMail Client > ?
For a beginner can you point me roughly to proper directions?
(I know it depends a lot on the App and what you like to do with
it but any thought will help)
I'm far away from developing complicated apps, they are still pretty simple.
Thanks for help,
Marc
You can use these rules of thumb to decide what storage model will work for your app.
If the data fits in memory entirely and is relatively unstructured, use plist
If the data fits in memory entirely and has tree-like structure, use XML
If the data does not fit in memory and has a structure of a graph, and the app does not need extraordinary query capabilities, use Core Data
If the data does not fit in memory, has a complex structure, or the app benefits from powerful query capabilities provided by relational databases, use sqlite
If the data must be secret (e.g. a password), use keychain.
Note that these choices often overlap, because multiple storage models will fit the same app. Your final decision depends on your personal preferences - you pick a technology that you understand better.
There was a very good question about sqlite vs. Core Data on Stack Overflow, you may want to read through the answers to that question.
My rule of thumb for each of those would be:
Time Tracker App > Core Data
RSS Reader App > Core Data
Photo App > Core Data
EMail Client > Core Data
Though in each case there would be things you would store on the file system. For instance, the photo app would obviously put the actual photos on the file system. The text of the emails would be on the file system etc. The actual RSS messages might be text files too, but with meta data in Core Data objects.
At some point, you might find that the data you are storing is outgrowing the scalability of Core Data. At that point you would consider moving to SQLite.
The point is that Core Data is so easy to use and so superior to the alleged lighter weight alternatives, why wouldn't you use it?
As far as I know, in iOS we can keep data in JSON files, .plist or .xml files, NSUserDefaults (which is basically a .plist file), SQLite database, or Core Data (which can be set to have different data stores, but usually we use .sqlite files).
I am currently supposed to make my app be able to work with tens of thousands of records, and I am using Core Data. The thing is that my client told me Core Data is probably not the best solution for large datasets, and I am confused because as far as I know, if we have so many records we should obviously not keep them all in memory (in NSMutableArray or other collection objects), and Core Data seems to be the best and fastest way to implement all the functionality.
Please let me know if you have the same problem in the past and which solution you chose.
I have used core data for a 64 thousand object database, worked perfectly as expected. Not sure exactly what you are trying to do but core data will work fine.
I am making a simple iPhone app that will basically be an editor.
As such, I need some way to store the documents the user creates.
Since on iPhone, the concept of the filesystem is not present for the user, I searched around to see what I should use.
I found this question & answer that basically says to use Core Data, but I recently found out about NSFileManager.
My question simply is, for user-created documents, what is the best storage system to use? Traditional files by using NSFileManager? Core Data? Something else?
Personally, I would use CoreData because it will abstract away all of the file-management code for you. If you are making simple text documents then this isn't such a big deal, but if you are working with a complex document architecture (i.e., a collection a numerous objects) then it can save you a lot of effort.
If the user wants to export their document it is very easy to write a function to do so with your CoreData objects.
The only downside to CoreData is that if you are using non-standard attributes it can get a little bit tricky, but it is certainly not a deal breaker in most cases.
People create document formats without CoreData all of the time, so there are plenty of examples out there, and it will just come down to personal preference. There really isn't any generalized right answer to this - it a design decision that should be evaluated on a per-app basis.
If all of your data for displaying the file is contained in one long string (like HTML) then I would recommend that you use the file manager, since it will be easy to get a list of files in a certain directory to display to the user for opening. However, if they are not self contained (like NSAttributedString, which has many stored formatting regions along with the actual content) then you should use CoreData, as it will be easier to keep all the pieces together.
I am developing a relatively simple new free iOS app. As it's free I decided to go straight for iOS 5 as a target to allow me to try out some new features, particularly story boards.
Once I've got my head around them storyboards are wonderful, I've written hardly any boilerplate code at all up the point of a working prototype.
My model at the moment is is non persistent, after all it's just a prototype, but I've started to wonder about the best way to make it persistent. The model consists of essentially a list of custom objects which contain text, images and locations.
And so to the questions.
If I have a fully working model in memory is there a relatively simple to convert that into a persistent core data representation ?
If I want to have synchronisation to a backend web service at some point will this help ?
Should I start again with a core data application and throw away the prototype ?
All depends on your data set. If you have a large graph with relationships, Core Data might be a good idea. For simpler data sets, you could write it out to a .plist in the user's documents directory. For very small amounts of data, you could persist them into NSUserDefaults.
I have a basic question regarding populating Core Data with data. I am building an application, which will show ATMs on a map. I would like to ship the application with a preloaded database, but to give users the option to receive updates when they launch the app. I am thinking about using a property list for the update. Basically send a plist of all the ATMs, parse that plist and populate the sqlite. I will have around 7000 entries in the property list file, each entry containing 5-6 keys with short string values. But according to the Apple iOS Developer Library:
You can create a property list—or some other file-based
representation—of the data, and store it as an application resource.
When you want to use it, you must open the file and parse the
representation to create managed objects. You should not use this
technique on iOS, and only if absolutely necessary on Mac OS X.
Parsing a file to create a store incurs unnecessary overhead. It is
much better to create a Core Data store yourself offline and use it
directly in your application.
Should I still be sending a property list or rather think for an alternative solution to update the application's database?
P.S. I am thinking about using a Rails app for providing updates - basically sending a plist file.
I had nearly the same question a few months back, did quite a bit of searching to find a nice easy answer, failed to find it and eventually settled on a roll-your-own solution that took a bit more time than I would have hoped, but was at least very helpful in learning to understand Core Data.
Basically the solution was to write a little utility that parsed my source data (which for me is a comma-separated text file, parsed using the quite handy 'cCSVParse' library - http://michael.stapelberg.de/cCSVParse ) and inserted it into Core Data Managed Objects and then saved that off as a sqlite persistent store. Then the sqlite store(s) can be shipped with the app, and uploaded by the user when they buy more data.
You could write a conversion from plist (or whatever) into the core data representation within the app itself, but if the data is just going live out the rest of its days in some core data form, why not let your beefy dev box do the heavy lifting before you send the data to the user, instead of shipping the data to the phone and making it do the work?