Comments for JSEDLAK http://jsedlak.org Thu, 11 Feb 2010 14:14:04 +0000 http://wordpress.org/?v=2.9.2 hourly 1 Comment on That’s Twitterific! by XNA Connection http://jsedlak.org/2010/02/11/thats-twitterific/comment-page-1/#comment-1288 XNA Connection Thu, 11 Feb 2010 14:14:04 +0000 http://jsedlak.org/?p=581#comment-1288 I'm also maintaining a list of english XNA users, you can check it there: http://twitter.com/xnaconnection/xna-anglophone I’m also maintaining a list of english XNA users, you can check it there:
http://twitter.com/xnaconnection/xna-anglophone

]]> Comment on FGF: A Collection Class with Events by John Sedlak http://jsedlak.org/2010/01/31/fgf-a-collection-class-with-events/comment-page-1/#comment-1287 John Sedlak Mon, 01 Feb 2010 11:13:54 +0000 http://jsedlak.org/?p=573#comment-1287 I am assuming that you are speaking about the event portion of the code when you talk about avoiding generics? The reason for this is that a single event handler couldn't listen to multiple lists. It is a very small and insignificant reason, I know, but it also allows for complete coverage of use cases. As for deriving from System...Collection - I honestly didn't know it existed. I will have to double check to see if this object is available on the 360 and Zune HD because Reflector seems to be having issues reflecting on them. The reason for the index and the object is because originally this interface was modeled after the INotifyCollectionChanged interface, which supplies both the index and the object as well as another parameter telling how the collection was changed. I am hoping that with the release of .NET 4.0, this interface will be better placed in the framework's libraries such that I can use it. This will give me interoperability with other .NET software. Again, the fact that the index exists does not limit the use to index based collections but does supply complete coverage which is an overriding goal for FGF. The point here is that if it is needed or can be used, the index parameter is there. If it weren't there than the use would be limited to non-indexable collections. Having said all that, your implementation seems perfectly reasonable and is a good way of approaching the situation. I am assuming that you are speaking about the event portion of the code when you talk about avoiding generics? The reason for this is that a single event handler couldn’t listen to multiple lists. It is a very small and insignificant reason, I know, but it also allows for complete coverage of use cases.

As for deriving from System…Collection – I honestly didn’t know it existed. I will have to double check to see if this object is available on the 360 and Zune HD because Reflector seems to be having issues reflecting on them.

The reason for the index and the object is because originally this interface was modeled after the INotifyCollectionChanged interface, which supplies both the index and the object as well as another parameter telling how the collection was changed. I am hoping that with the release of .NET 4.0, this interface will be better placed in the framework’s libraries such that I can use it. This will give me interoperability with other .NET software.

Again, the fact that the index exists does not limit the use to index based collections but does supply complete coverage which is an overriding goal for FGF. The point here is that if it is needed or can be used, the index parameter is there. If it weren’t there than the use would be limited to non-indexable collections.

Having said all that, your implementation seems perfectly reasonable and is a good way of approaching the situation.

]]> Comment on FGF: A Collection Class with Events by Markus Ewald http://jsedlak.org/2010/01/31/fgf-a-collection-class-with-events/comment-page-1/#comment-1286 Markus Ewald Mon, 01 Feb 2010 07:01:59 +0000 http://jsedlak.org/?p=573#comment-1286 Having taken that route already, allow me to make some suggestions ;-) Is there any reason why you avoid generics in that code? I would hate to have to downcast in the event handler. Especially since with this design, if I later changed the collection's type, those downcasts would still happily compile and only fail a runtime. You also don't seem to derive from System.Collections.ObjectModel.Collection (which is made for this purpose and would implement IList, ICollection, IEnumerable, IList, ICollection and IEnumerable for you) Supplying an index *and* the object seems redundant. Such a talkative interface limits the concept to indexable collections. I have had a use for observable dictionaries in the past :) Here's my hammer for the same nail: :P https://devel.nuclex.org/framework/browser/Nuclex.Support/trunk/Source/Collections/ItemEventArgs.cs https://devel.nuclex.org/framework/browser/Nuclex.Support/trunk/Source/Collections/ObservableCollection.cs https://devel.nuclex.org/framework/browser/Nuclex.Support/trunk/Source/Collections/ObservableDictionary.cs Having taken that route already, allow me to make some suggestions ;-)

Is there any reason why you avoid generics in that code? I would hate to have to downcast in the event handler. Especially since with this design, if I later changed the collection’s type, those downcasts would still happily compile and only fail a runtime.

You also don’t seem to derive from System.Collections.ObjectModel.Collection (which is made for this purpose and would implement IList, ICollection, IEnumerable, IList, ICollection and IEnumerable for you)

Supplying an index *and* the object seems redundant. Such a talkative interface limits the concept to indexable collections. I have had a use for observable dictionaries in the past :)

Here’s my hammer for the same nail: :P
https://devel.nuclex.org/framework/browser/Nuclex.Support/trunk/Source/Collections/ItemEventArgs.cs
https://devel.nuclex.org/framework/browser/Nuclex.Support/trunk/Source/Collections/ObservableCollection.cs
https://devel.nuclex.org/framework/browser/Nuclex.Support/trunk/Source/Collections/ObservableDictionary.cs

]]> Comment on A Letter To Microsoft: On Framework Design by Björn http://jsedlak.org/2009/10/02/a-letter-to-microsoft-on-framework-design/comment-page-1/#comment-1253 Björn Sat, 03 Oct 2009 05:14:54 +0000 http://jsedlak.com/?p=220#comment-1253 INullable was added in the .NET 1.1 timeframe or so, hardly after Nullable was introduced in .NET 2.0. Additionally, INullable sole reason for existence is the fact that value types could not be null in the pre-2.0 era while DB values can be null (which is why it is in an "awkward namespace"). However, I am not sure whether the interface should be moved and Nullable(Of T) should implement it… P.S.: Please do not include the "(classic)" annotation on MSDN links. I like my "(lightweight)" :) INullable was added in the .NET 1.1 timeframe or so, hardly after Nullable was introduced in .NET 2.0. Additionally, INullable sole reason for existence is the fact that value types could not be null in the pre-2.0 era while DB values can be null (which is why it is in an “awkward namespace”).

However, I am not sure whether the interface should be moved and Nullable(Of T) should implement it…

P.S.: Please do not include the “(classic)” annotation on MSDN links. I like my “(lightweight)” :)

]]> Comment on Open User Information within Browsers by Björn http://jsedlak.org/2009/08/06/open-user-information-within-browsers/comment-page-1/#comment-1081 Björn Sat, 08 Aug 2009 07:04:44 +0000 http://jsedlak.com/?p=192#comment-1081 The Information Card thing can be considered a standard and it is pretty open for use; adoption is a different thing but there's the Informatin Card Foundation to change that. [1] http://informationcard.net/ The Information Card thing can be considered a standard and it is pretty open for use; adoption is a different thing but there’s the Informatin Card Foundation to change that.

[1] http://informationcard.net/

]]> Comment on Theming SharePoint (4 of Infinite) by naresh http://jsedlak.org/2008/07/10/theming-sharepoint-4/comment-page-1/#comment-890 naresh Tue, 31 Mar 2009 20:59:43 +0000 http://jsedlak.com/?p=57#comment-890 Hi, thanks for the post, i tried to run the chsrp file and i am getting the error "type 'Microsoft.Web.Management.Client.NavigationItem' has no constructors defined" and i can not see the properties Title,URL,Level. how do we fix this error? much apprciate for your response. Thanks. Hi, thanks for the post, i tried to run the chsrp file and i am getting the error
“type ‘Microsoft.Web.Management.Client.NavigationItem’ has no constructors defined”
and i can not see the properties Title,URL,Level. how do we fix this error? much apprciate for your response. Thanks.

]]> Comment on Preliminary XBLCG Data Compiled by malfunct http://jsedlak.org/2009/03/29/preliminary-xblcg-data-compiled/comment-page-1/#comment-888 malfunct Sun, 29 Mar 2009 23:25:43 +0000 http://jsedlak.com/?p=171#comment-888 Has MS given any indication what the total number of downloads for all CG was? Has MS given any indication what the total number of downloads for all CG was?

]]> Comment on Cross Browser Development by John Sedlak http://jsedlak.org/2009/03/21/cross-browser-development/comment-page-1/#comment-887 John Sedlak Sat, 21 Mar 2009 23:07:14 +0000 http://jsedlak.com/?p=169#comment-887 If I concede, it would mean this post was useless. I will try to update it with a sample, but now I have IE8 on all my machines... If I concede, it would mean this post was useless. I will try to update it with a sample, but now I have IE8 on all my machines…

]]> Comment on Cross Browser Development by Björn http://jsedlak.org/2009/03/21/cross-browser-development/comment-page-1/#comment-886 Björn Sat, 21 Mar 2009 21:15:26 +0000 http://jsedlak.com/?p=169#comment-886 i. My numbers were meant to be related to the numbers in the post itself :] ii. This page is supposed to be in Hypertext markup. Yet, there's no /hyper/text to support your claims... iii. Not really. Once you accept collapsing margins that feature is pretty nice to have. iv. I should comment through my blog only then - too bad me is as lazy as you... i. My numbers were meant to be related to the numbers in the post itself :]
ii. This page is supposed to be in Hypertext markup. Yet, there’s no /hyper/text to support your claims…
iii. Not really. Once you accept collapsing margins that feature is pretty nice to have.
iv. I should comment through my blog only then – too bad me is as lazy as you…

]]> Comment on Dear Guy In Front by mikeschuld http://jsedlak.org/2009/01/21/dear-guy-in-front/comment-page-1/#comment-877 mikeschuld Thu, 22 Jan 2009 05:33:23 +0000 http://jsedlak.com/?p=164#comment-877 John, you are without a doubt my hero. John, you are without a doubt my hero.

]]>