brightstardb / brightstardb Goto Github PK
View Code? Open in Web Editor NEWThis is the core development repository for BrightstarDB.
Home Page: http://brightstardb.com/
License: MIT License
This is the core development repository for BrightstarDB.
Home Page: http://brightstardb.com/
License: MIT License
Does it support Microsoft Sync Framework / offline changes tracking ?
Now that the sphinx documentation sources are ahead of the old Help and Manual, remove H&M docs and move the sphinx docs up to the doc/src directory. The build settings on readthedocs.org will need updating too.
Go through the existing samples, check that they all compile and update code and connection strings where necessary.
From fiatluxph on the codeplex mailing list:
I know that the following code will delete an object:
var country = Data.Instance.Context.Countries.FirstOrDefault(d => d.Id == id);
Data.Instance.Context.DeleteObject(country);
Data.Instance.Context.SaveChanges();However, the next time I query the Countries collection, I find that the element
with that id is still in the Countries collection, all be it with its properties set to their
default values.
The main reason is that the deployment approach of MSTest is slowing down the running of individual unit tests / subsets of tests during development. There are also some nice features in NUnit that we could make use of.
Need to decide the best way to support graphs. Probably the simplest is to allow a collection of graph URIs to define the source dataset and a single graph URI that will be the target graph for all inserts. The downside of this simple approach would be that an update of a property could conceivably remove a statement from one graph and add its replacement to a different graph.
However, keeping track of graph sources for different properties introduces its own problems as it is possible for a single property to have values that come from multiple different source graphs.
User has reported that in Step 6: Creating a context class it is too early to hit f5 and get a full site up and running - as there are new views created yet. It looks like the two sub-sections of step 6 are out of order.
These timestamps are maintained internally but they never get copied into the output DTOs.
Because other commit point APIs use the commit point offset rather than the commit ID, this should be the ulong value we provide as commit id in the statistics structure.
Extend the current APIs used for processing updates so that a default graph URI can be specified.
In tidying up and simplifying the site I managed to remove the only link through to the BrighstarDB Users google group. This definitely needs to be reinstated - maybe as a link on the docs page, but also as a footer link. It might be nice if latest questions on the group could get displayed in the blog sidebar too...
When we were IL merging all the required libraries into BrightstarDB.dll it was fine to have nuspec files with no dependencies. From 1.3 we should explicitly call out our dependencies instead. This requires some updates to the nuspec files and some local testing to ensure installation into a project goes smoothly.
Look at if / how this can be mapped to SPARQL 1.1 grouping. Need to also consider impact on aggregation.
I think last time we looked the problem was that the default dotNetRDF virtual node uses a 32-bit node ID and we use a 64-bit node ID. Need to get back in to that part of the code-base to reacquaint myself with the issues.
Hi
We are considering using Brighstar for a project whereby we need to have an embedded database - that supports Entity Framework - for offline scenarios. However going through the documentation there seem to be various places that mention the purchasing of a licence in order to use Brightstar in this way. Their are a number of URL's mentioned for obtaining this licence, but do not seem to resolve:
http://brightstardb.com/buy/
http://brightstardb.com/customer/
...could you possibly confirm whether we do in fact need to purchase a licence and how much it is?
Thanks!
James
In particular now that docs are online we can just ship the API documentation and instead point to readthedocs.org
The build for the MSI installer and nuget packages has been thoroughly broken by the recent changes to the project.
Issue from the Google group (https://groups.google.com/forum/#!topic/brightstardb-users/e9NahY7bf3k).
I have an entity A containing a collection of entities B.
Entity A has an inverse property pointing to a property in entity B.I have tried two methods for removing an entity B object.
- Call the DeleteObject method on the context followed by SaveChanges.
This removes the desired entity from the B entity set in the DB, but does not
remove it from the A entity set to which it was linked - the fields are set to null
but the Id still exists.
This means when I test the count of objects has reduced, it fails.
- If I explicitly call Remove on the collection of B entities within the A object
and then call SaveChanges, the item is removed from the A object in the DB,
but still exists in the B entity set in the DB.
(2) is expected behaviour. (1) sounds like an issue with the deleted B entity object not getting removed from the collection (possibly because the collection is generated from an inverse property relationship).
Azure and cluster code are not being actively maintained at the moment.
Dotnetrdf patches are no longer necessary.
Server directory is obsolete.
Delete the azure and cluster code. They have not been modified since the last release so if they are ever needed it will be possible to retrieve them from the 1.3 or 1.4 release tags.
Delete the dotnetrdf patches - same reasoning, and they are no longer needed in any form.
Delete server directory entirely.
The SDShare service is not really a core part of BrightstarDB and can be better managed under its own repository.
Work on this is being done in the feature/QueryGraphSupport branch. Planned for inclusion in the 1.3 release.
This will enable the use of these APIs with other SPARQL 1.1 compliant endpoints.
I am currently a student and I have no experienced contributing in open sourced projects (I,ve never worked on a project like this).This is my first time in GitHub, sorry if this is not the right place to post this.
I want to try BrightstarDB because I am considering it for a windows phone app that I want to start next month. I follow this doc (http://brightstardb.readthedocs.org/en/latest/Developer_Quick_Start/) but there are problems in my solution that can't use it, I have installed BrightstarDB from Nuget in Visual Studio 2012 (and there is not Entity Context/Definition in the item list), and set the target framework to 4.0 and Any CPU setting. But errors are raised when I try it. Below is a image from VS with the errors displayed.
How I can solve this?
It would be nice to have a Brightstar store manage some basic statistics. Useful triple-level statistics might be:
But it might also be good to track B-Tree and database statistics e.g.
It may be harder to track the internal structure stats but they could be periodically updated in a background thread or at the end of a big update.
Since BrightStar is now open source, is there still a need for a license?
If so, I'm confused as to when and why a license is needed in some areas and not others.
The Quick Start foe Entity Framework needs a license, and I don't understand why.
If this is not the right place to ask this question, please direct me to an appropriate location.
Thanks!
Targetting Mono 3.0.
Include basic library and a service/service runner stack plus relevant unit tests at a minimum.
Question from the User's group:
I have this SPARQL Update query, which works in Polaris and in the nuget library:
INSERT DATA {
urn:a urn:is "c:\users\r\nbla"
}But when I use ExecuteTransaction (or the transaction feature in Polaris), and I enter > following to the "Triples to Add" field, I get a "Transaction failed" message
urn:a urn:is "c:\users\r\nbla"
The problem is with \u apparently. If I enter "c:users", it works.
It looks like the explanation is that the parser is not treating \ as an escaped backslash character and is instead treating the \u as the start of a unicode character code. This will be a problem in our NTriples/NQuads parser implementation (which is used by ExecuteTransaction) and not in the dotNetRDF parser implementation (which is used by SPARQL Update).
For append-only stores it is currently possible to use the consolidate option to shrink the database in-place. It would be good to have another option that works like consolidate but copies the data into a new store.
This would enable the concept of having a master store and replicas that are periodically created by snapshotting the master.
A long running query causes the GUI to hang - the query should be executed on a background thread, or if querying a server the request should be handled asynchronously
This is because the generated WCF client includes task-based async methods which .NET 4.0 doesn't know how to process (but .NET 4.5 does). As a result the app runs fine in IIS 8, but breaks in IIS 7 as it tries to interpret a Task return value as something that should be serializable.
Fix is to generate the WCF client without the async methods.
Create a PCL port of BrightstarDB. Goal is to allow targeting Windows Phone 7.5+ and Windows Store apps.
Tried to import a smallish GZipped NTriples file (91801_M_UK_NMM_oilpaintings_dc.nt.gz) using the local import option in Polaris. The user interface became unresponsive almost immediately and could only be killed through task manager.
Unzipping the file and importing the .nt file worked without a problem, so its probably not a data issue.
The "About" box displayed by Polaris still has the old commercial copyright on it.
An error exist in the interface IFilm.cs, (http://brightstardb.readthedocs.org/en/latest/Developer_Quick_Start/)
Add support for this syntax.
Solution Project - https://docs.google.com/file/d/0BxhZYhbuSf6IbktHclVzYjBVOTQ/edit?usp=sharing
Hi kal, here is another problem found, I made the change you tell me in the other post (Can't run transformation) and the fix worked! but now can't compile this new version, the problem can be because I am writing the code in the constructor (In a WPF application) instead of console app? I know that the example you show in the docs is in console.
Currently the results formats supported are only for SPARQL results sets, and CONSTRUCT queries always get serialized as RDF/XML. We should extend support to allow serialization of graphs using other formats.
The import process consumes increasing amounts of memory when importing a large file, leading eventually to an out of memory error.
Memory management needs to be improved so that it is possible to import large files without having to chunk them up into smaller pieces.
This is dependent on completion of issue #8
Provide a full implementation of the BrightstarDB service interface using RESTful principles.
Should have pluggable authentication on both client and server so that the application programmer is in control of how REST calls are authenticated and authorized for particular actions.
Currently only one SparqlResultsFormat instance can be passed through. It would make sense to support passing through an ordered list of formats, perhaps as mime type strings. This would allow applications to specify a format for SPARQL results sets and another format for graphs.
Create an MSI for x64 platforms that installs into ProgramFiles not ProgramFiles (x86)
Provide a way for the BrightstarDB service to be configured to preload BTrees and / or pre-cache certain SPARQL query results.
An RDF plain literal is a literal consisting of a string and an optional language tag (http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/#section-Literals). In entity framework to access the language tag on a plain literal property we will need a wrapper class with properties for the language tag and the value.
This should be optional - existing applications of entity framework should not be required to convert all string properties to plain literal properties and it should be possible to mix and match in an entity definition.
The plain literal type needs to also be supported in LINQ to SPARQL
Very basic case of creating a context and deleting an entity I just added.
Getting this exception:
No Entity Found for Subject http://www.brightstardb.com/.well-known/model/wildcard
Code to reproduce:
[Entity]
public interface IWeb
{
string Title { get; set; }
string AbsoluteUrl { get; set; }
}
public void Test()
{
var connectionString = @"Type=embedded;StoresDirectory=d:\temp\BrightstarTest;StoreName=test1";
var context = new MyEntityContext(connectionString, true);
var web = context.Webs.Create();
web.Title = "Some Test";
web.AbsoluteUrl = "http://www.google.com";
context.SaveChanges();
context.DeleteObject(web);
context.SaveChanges();
}
Getting the "No RDF type found mapped to Uri"
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.