Loading ...
Sorry, an error occurred while loading the content.
1088
[AS] two free webinars from Net Objectives. Executive's View of Agil Apr 22, 9am Pacific. This Tuesday. An Executive's Guide to Agile
Al Shalloway
5:27 PM
#1088
 
1087
Re: interface naming conventions I’m teaching this week so have not had a chance to weigh in here, and even catching up looks daunting… but I’ll give my 2 cents. The problem, for me,
Scott Bain
Mar 11
#1087
 
1086
Re: interface naming conventions If you want to decouple the test invocation from a specific test binding, try the template method approach that Scott actually gave a code sample for, I think
Amir Kolsky
Mar 9
#1086
 
Fetching Sponsored Content...
1085
Re: interface naming conventions I've been trying to promote the idea that a 3 layer model of UI, business layer, and database maps nicely to the Test Framework (MSTest in our case), test code
rstone1998
Mar 9
#1085
 
1084
Re: interface naming conventions Frameworks are Adapters. In many cases frameworks are adapters to an outside entity (such as the user). This does not mean you do not need an adapter to
    Amir Kolsky
    Mar 9
    #1084
    This message has attachments
    1083
    Re: interface naming conventions Principles and design qualities. The principles on their own are useful for devs to follow. The reason is that they promote good code qualities. For example,
      Amir Kolsky
      Mar 9
      #1083
      This message has attachments
      1082
      Re: interface naming conventions Let’s start with the good, I agree with most of what you’re saying. Let’s discuss the partial class idea. We need to separate the Actor’s behavior from
        Amir Kolsky
        Mar 9
        #1082
        This message has attachments
        1081
        Re: interface naming conventions I’m coming to believe that “any” cohesion problems are, if not synonymous with SRP violations, different by so little that they could be treated
        rstone1998
        Mar 9
        #1081
         
        1080
        Re: interface naming conventions Writing an adapter should be very simple. It can by dynamic: Framework -> Adapter -> OurEntity. Or it can be static: Framework -> Adapter:OurEntity. (Your
        Amir Kolsky
        Mar 8
        #1080
         
        1079
        Re: interface naming conventions This is a fine point. ACTORS represents self-contained entities that have responsibilities which are inherent to the actors and define them. An ENCRYPTOR
        Amir Kolsky
        Mar 8
        #1079
         
        1078
        Re: interface naming conventions I didn't really address your comment on the Visitor pattern or some other type of adapter. While I like it, the problem with it may be my inability to write
        rstone1998
        Mar 8
        #1078
         
        1077
        Re: interface naming conventions So I followed your reasoning up to the point on actors. I get that you don't want to have two "actors" implemented in the same class as that would violate SRP.
        rstone1998
        Mar 8
        #1077
         
        1076
        Re: interface naming conventions Now to YOUR reasons: Reason 1: I agree with you that you have use the “I” convention. This is a mandate, you have to do it or (like Max) you will not be
        Amir Kolsky
        Mar 8
        #1076
         
        1075
        Re: interface naming conventions Part 2 – Extend vs. Implement In c# the inheritance mechanism involves the usage of the colon ‘:’. When you extend a class or when you implement an
        Amir Kolsky
        Mar 8
        #1075
         
        1074
        Re: interface naming conventions Your (and others’) basic mistake here, is that you think that putting an “I” in front of an interface follows the Hungarian Notation. The number of
        Amir Kolsky
        Mar 8
        #1074
         
        Fetching Sponsored Content...
        1073
        Re: interface naming conventions I'm in a similar boat. "I" is a "standard" and there's no way around it. I only write about one or two interfaces a year anyway. I use purely abstract
        Max Guernsey
        Mar 8
        #1073
         
        1072
        Re: interface naming conventions Simple reason using I as a convention is bad. It exposes an implementation detail that doesn’t need to be exposed. Not really bad, but it gets us focusing
        Al Shalloway
        Mar 8
        #1072
         
        1071
        interface naming conventions As few days ago Amir jumped me for using the dreaded Hungarian I in an interface name (Forgive me father for I have sinned!), and while I’ve heard his rant
        rstone1998
        Mar 8
        #1071
         
        1070
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi I did read the “correct” paper, specifically, the one you sent out most recently – that is what I was quoting the terminology from. You answer is what I
        Chris Hand
        Mar 6
        #1070
         
        1069
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi The tests do not test for LSP. The tests test the behavior of an entity. If the tests behave the same way for an object of a base type as well as the sub type
        Amir Kolsky
        Mar 6
        #1069
         
        1068
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi A behavior is something observable you can say about an entity. It has a context, a trigger and an outcome. There are two ways of using the word “behavior”
        Amir Kolsky
        Mar 6
        #1068
         
        1067
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi Yes, Scott said: "What I could do next is show the example of Square sub-classing Rectangle, and how the same test would not pass in that case, but perhaps
        chrishandone
        Mar 6
        #1067
         
        1066
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi Actually your answer is correct, in this context. The real point is that an important characteristic of LSP is how you handle when something bad happens - it
        chrishandone
        Mar 6
        #1066
         
        1065
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi Thanks for pointing out that paper - it was a good read, and very helpful Years ago we (Amir and I) were in a different long discussion, where I was trying to
        chrishandone
        Mar 6
        #1065
         
        1064
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi Am I missing something here? What is there to agree with? From: leanprogramming@yahoogroups.com [mailto:leanprogramming@yahoogroups.com] On Behalf Of Chris
        Amir Kolsky
        Mar 6
        #1064
         
        1063
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi Thanks Ray, I completely agree with that conclusion. -Chris
        Chris Hand
        Mar 6
        #1063
         
        1062
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi Of course, this is why Scott chose the Shape, Square, Rectangle example. From: leanprogramming@yahoogroups.com [mailto:leanprogramming@yahoogroups.com] On
        Amir Kolsky
        Mar 6
        #1062
         
        Fetching Sponsored Content...
        1061
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi although mathematically a square is a rectangle, Robert Martin makes a nice point that the behavior of a rectangle is fundamentally different from the behavior
        rstone1998
        Mar 6
        #1061
         
        1060
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi Anyone else desperately hope that Internet archaeologists never find this conversation thread? -- Max *Architecture | **Value-Stream-Oriented Architecture
        Max Guernsey
        Mar 5
        #1060
         
        1059
        Re: 7RE: [leanprogramming] RE: Will Overriding Leading to Compromisi Chris, I think I have identified the source of our disagreement. I assume that you went to the Wikipedia page for the Liskov Substitution
        Amir Kolsky
        Mar 5
        #1059
         
        View First Topic Go to View Last Topic
        Loading 1 - 30 of total 1,088 messages