@Straxus I don't know. No Android phone has tempted me so far, and I'm pretty sure I'll only resist the iPhone 5 until I first touch one
5
Lame reactions to the #iphone5 everywhere. I guess that means it's going to be hugely successful.
4
7
1
@janl But you asked whether "GET /post/blogname -d “”markdownbody” is RESTful design", which triggered my reflex
1
@janl My goal for hierarchical URIs is that the individual paths you get when chopping off segments should be meaningful
1
@janl @mwessendorf @abstractj Agreed: URI design vs. REST, which are orthogonal. Just don't say ugly URIs are not RESTful :-)
1
@janl @mwessendorf @abstractj Probably not, but because that'd be using GET wrongly, not because the URI contains certain characters
1
@janl It just makes it a bit easier to handle on the server side (less name clashes). Merely a matter of taste, IMO. I'm fine both ways.
2
@janl @mwessendorf @abstractj Strictly speaking, there is no way to build URIs that are unRESTful. URI characters don't matter.
1
@janl Surely cool, but only doable if :username is at the top of all or at least most hierarchies
1
@mwessendorf Pretty URI design has got nothing to do with REST :-) That said, I prefer /auth/users if that resolves to sty meaningful
1
1
@planetzuda You're right, and I shouldn't have written "amazed" but "annoyed"
+1 to almost all of the points made in @simonbrown's excellent presentation on the architect's role: codingthearchitecture.com/pr…
1
2
11
+1 to @gruber's view: “Apple’s skeuomorphic designs are a fad, much like the pinstripes and brushed metal of a decade ago”
@pavlobaron I see; I was lumping select/poll/epoll/kqueue and all the others together. The real anti pattern is one thread per connection
@pavlobaron I was just puzzled by your negative attitude to select/poll :-)
@b_erb These days, I'd consider the JVM to be a very reasonable choice for something like that
1