Talk:Development:Reviewing - Revision history

Sardisson: /* Code style questions */ more

2007-06-21T21:09:59Z

Code style questions: more

Sardisson: Code style questions

2007-06-21T15:32:41Z

Code style questions

Sardisson: /* Thoughts on review model */ oops

2006-05-19T22:57:16Z

Thoughts on review model: oops

Sardisson: discussion

2006-05-19T22:56:44Z

discussion

New page

==Thoughts on review model==

[6:49pm] ss|work2: Everyone should be targeting the "newbie" reviewers first and letting the sr and/or smorgan look at it after that.
[6:49pm] ardissone: i agree
[6:50pm] ss|work2: If it's small code, let a newbie reviewer look at it, then an sr.
[6:50pm] ss|work2: If it's big code, newbie reviewer than smorgan, who might find more.
[6:50pm] ardissone: or complex
[6:50pm] ss|work2: That will help the newbie reviewer see what they missed as well without burdening the sr queue
[6:50pm] ss|work2: Right.
[6:50pm] ss|work2: s/big/complex
[6:50pm] ardissone: small code can still have complex interactions

@@ Line 45: / Line 45: @@
 [11:26am] peeja: ok<br>
 [11:28am] kreeger: #import isn't technically recommended for anything other than Apple's stuff<br>
-[11:28am] kreeger: I try and only use it for Obj-C stuff, and use #include for everything else (c, c++)
+[11:28am] kreeger: I try and only use it for Obj-C stuff, and use #include for everything else (c, c++)<br>

← Older revision		Revision as of 15:32, 21 June 2007
Line 10:		Line 10:
	[6:50pm] ss\|work2: s/big/complex<br>		[6:50pm] ss\|work2: s/big/complex<br>
	[6:50pm] ardissone: small code can still have complex interactions<br>		[6:50pm] ardissone: small code can still have complex interactions<br>
		+
		+	== Code style questions ==
		+
		+	[11:15am] peeja: do we line-break long method calls?<br>
		+	[11:17am] ardissone\|away: i believe people have been scolded for breaking 80, yes, but i'm not the right person to answer<br>
		+	[11:17am] peeja: ok<br>
		+	[11:17am] ardissone\|away: that should be part of our code style doc if it's not...<br>
		+	[...]<br>
		+	[11:22am] peeja: do we try to use @class over #import?<br>
		+	[11:22am] ardissone\|away: i believe there's a rule for that for w/s/cocoa, but i don't know if we have one<br>
		+	[11:23am] kreeger: peeja: you can pre-declare a class<br>
		+	[11:23am] kreeger: rather than having to import that everytime a header is included<br>
		+	[11:23am] kreeger: removes a shit load of dependencies<br>
		+	[11:23am] peeja: ok<br>
		+	[11:24am] kreeger: same thing in c++:<br>
		+	[11:24am] peeja: when do you need to #import?<br>
		+	[11:24am] kreeger: instead of:<br>
		+	[11:24am] kreeger: #include "nsISupports.h"<br>
		+	[11:24am] kreeger: class nsISupports;<br>
		+	[11:24am] kreeger: although nsISupports wasn't the best example<br>
		+	[11:24am] kreeger: ok<br>
		+	[11:24am] kreeger: so the #import actually pulls the header into the file before compilation<br>
		+	[11:25am] peeja: rught<br>
		+	[11:25am] kreeger: so symbols (like classes and functions) can be defined at compiler time<br>
		+	[11:25am] peeja: er, right<br>
		+	[11:25am] kreeger: and #import imports the file only once<br>
		+	[11:25am] kreeger: and #include imports everytime<br>
		+	[11:25am] peeja: but in obj-c, where do we draw the line and use @class<br>
		+	[11:25am] peeja: yeah, I know<br>
		+	[11:25am] peeja: do we use it when we use methods?<br>
		+	[11:25am] kreeger: the trick is to only use @class when you just need to define a class<br>
		+	[11:26am] peeja: but don't need any of the methods in it?<br>
		+	[11:26am] kreeger: but when you need to start using some of the classes meethods, you will need to #include/#import the header<br>
		+	[11:26am] peeja: ok<br>
		+	[11:28am] kreeger: #import isn't technically recommended for anything other than Apple's stuff<br>
		+	[11:28am] kreeger: I try and only use it for Obj-C stuff, and use #include for everything else (c, c++)

@@ Line 1: / Line 1: @@
 ==Thoughts on review model==
-[6:49pm] ss|work2: Everyone should be targeting the "newbie" reviewers first and letting the sr and/or smorgan look at it after that.
+[6:49pm] ss|work2: Everyone should be targeting the "newbie" reviewers first and letting the sr and/or smorgan look at it after that.<br>
-[6:49pm] ardissone: i agree
+[6:49pm] ardissone: i agree<br>
-[6:50pm] ss|work2: If it's small code, let a newbie reviewer look at it, then an sr.
+[6:50pm] ss|work2: If it's small code, let a newbie reviewer look at it, then an sr.<br>
-[6:50pm] ss|work2: If it's big code, newbie reviewer than smorgan, who might find more.
+[6:50pm] ss|work2: If it's big code, newbie reviewer than smorgan, who might find more.<br>
-[6:50pm] ardissone: or complex
+[6:50pm] ardissone: or complex<br>
-[6:50pm] ss|work2: That will help the newbie reviewer see what they missed as well without burdening the sr queue
+[6:50pm] ss|work2: That will help the newbie reviewer see what they missed as well without burdening the sr queue<br>
-[6:50pm] ss|work2: Right.
+[6:50pm] ss|work2: Right.<br>
-[6:50pm] ss|work2: s/big/complex
+[6:50pm] ss|work2: s/big/complex<br>
-[6:50pm] ardissone: small code can still have complex interactions
+[6:50pm] ardissone: small code can still have complex interactions<br>