On Mon, Sep 10, 2012 at 7:23 PM, Vincent Massol <[email protected]> wrote:
> Hi Jerome, > > Hey I didn't really want to awaken this topic again ;) I already gave my > opinion on the other thread (I still need to reply to Sergiu's last mail > BTW). I just wanted to explain how checkstyle computes the class fan out > and share that since it was useful for me to fix one class fan out issue I > had. > > See below. > > On Sep 10, 2012, at 7:06 PM, Jerome Velociter <[email protected]> wrote: > > > On 09/10/2012 06:17 PM, Vincent Massol wrote: > >> With this finding writing a macro requires between 5 to 6 Classes: > >> > >> * DefaultContentDescriptor (only if there's content) > >> * XXXMacroParameters > >> * List > >> * Block > >> * MacroTransformationContext > >> * MacroExecutionException > >> > >> Our checkstyle config fails above 20 which means Macro classes have > about 14 additional deps on external classes allowed before they break. > Which should normally be more than enough. > >> > >> My understanding is that reading the code is hard when there are too > many references to external classes (incidentally it also makes the code > more brittle as there are more chances it'll break due to an issue with the > used classes). Apparently 7 (+/- 2) is the magic number (see > http://en.wikipedia.org/wiki/The_Magical_Number_Seven,_Plus_or_Minus_Two) > > > > So this shows "20" was magic to begin with. How do you go from 7 +/-2 to > 20 ? > > Would be interesting to ask the checkstyle guy why they picked this value. > My guess is that they used something like 7 + standard java classes. > > > I find it a bit specious to apply the theories exposed in the article to > a program's class external classes without a pinch of salt. It would be > saying their is a good match between the notion of "object" as described in > the experiments (so in the sense of an item of input, like a word in a list > of words or a sound in a list of distinct sounds) and the notion of a class > of objects in programming. > > Not my choice. All references to Class Fan Out say that the default value > should be around 7 and quote this article. > > > What I'm trying to say is that maybe we should rather accept the fact 20 > is magic to begin with and try and figure out if this works for us and > tweak it if necessary > > I've already answered to this. I'm -1 to change what we have in a random > manner with another random value. What I said is to show me some code and > we can discuss on this basis. Sergiu has pointed out some code but I didn't > get the time to look at it yet. I need to do that and if the outcome is > that we can't make it pass with a good design then we can discuss either > make an exclude for this special case or decide that the case is generic. > > > - rather than trying to rationalize it with bridges to psychology thesis > that are far from obvious (not that the thesis don't apply to programming ; > just that the impendance of an "input" or "unit of thought" is IMHO not so > easy to match with programming concepts without making blind guesses). > Maybe I'm wrong, it's just I find this kind of rationalization somehow > naïve. > > > > Hope that made sense. > > Well, there's not much point in saying that something is wrong if you > don't provide an alternative… If you're going against the established > standard then you better be an expert in the domain. I know I'm not. But > even if we agree that the study has no meaning, it doesn't change the fact > that to change what we have we need to have a good reason to do so and we > need to define the new value while still ensuring that the check still > makes sense. > > > Also, a note about ignoring classes that "don't require effort to > understand" (or to look at their source code), they would probably span > more than "standard Java classes" like Collections, no ? > > Yep, I mentioned that in my mail (I said "Collection" classes). > > > For example, what about common-langs (StringUtils, etc.) ? > > I wouldn't put those, not everyone know them and again static method calls > **don't count** for the class fan out … > > Anyway ATM I don't agree about starting to ignore classes without at least > one good example where we have to do so to make the fan out check pass. > > I write new code almost everyday and it's seldom that I have issues with > class fan out and most of the time when it happens it's because I've been > lazy and I can fix the issue in a few minutes. This makes me think that the > current value is good. As I've already said in the other thread for me it > would be much better to agree to do a checkstyle exclude of *only* the > ClassFanOut check for a given class rather than change the default for > everyone. > I completely agree about that last point. Specific exclusion of specific rules (not only ClassFanOut) would be a really better practice than excluding the whole file. About ClassFanOut in particular, there is something that is not properly taken into account when measuring this metrics. There is well know or simple classes, that could be used and depends upon, and other no so well know one, that will really increase the readability of the code, and its independence. Since the computation of ClassFanOut metric does not take that into account, it sometimes consider some code as badly design while it is simply using many small simple classes, which is usually not a bad design. Excluding some custom well-know classes from the count could helps mitigating that. IMO, classes representing References, Context, Collections (as you suggest), and ... well this is where we need an expert. So this is really not an easy topic and it will be very difficult to find a correct number, and correct exclusions. One last remark, when you leave a class with a ClassFanOut equal to the limit, there is a lot of chance that a simple bug fix could exceed the limit and cause lot of trouble while having not necessarily cause a worse situation than before. IMO, the priority is to fix the issue first, and to improve the code quality later if time is available. Of course, it depends on the details, and it's not always true. This is generally why I am always skeptical using these metrics as constraints. Could not the dev team be more responsible without constraint (but keeping an eyes on metrics of course !) ? This is just my own ideas and I am not expert ;) > > Thanks > -Vincent > > > Jerome > > > >> > >> So IMO we have 2 options that could make sense (ie we can rationalize > them): > >> > >> * Keep this default value of 20 total which includes some standard JDK > classes. 14 deps for Macro should be enough, even if we count, say 2-4 more > for standard Java classes like Collection classes.That still gives us about > 10 deps for additional XWiki classes. > >> * Decide to exclude some classes that are basically part of the Java > language (such as Collection classes) since they don't increase the reading > complexity of the code since everyone knows what they do and we don't need > to look their source code or doc to understand them. However, if we do so > then we should reduce the allowed Fan out to 7 (+/- 2) so let's say to 9. > So that gives us only 9 deps for XWiki classes, compared to about 15 ATM... > >> > >> IMO the second option is much harder for us than the first one and > that's why I'd keep the first option... > >> > >> BTW the reason I wrote about all this is because I made a change to the > ChartMacro and the fan out became 21… In the end I refactored the code and > got a slightly better design (and fixed a bug at the same time…). > >> > >> Thanks > >> -Vincent > >> > >> > >> On Sep 10, 2012, at 4:59 PM, Vincent Massol <[email protected]> wrote: > >> > >>> Hi devs, > >>> > >>> I wanted to understand how Checkstyle computes the Class Fan out so I > debugged it. > >>> > >>> Here are my findings: > >>> > >>> * Some classes are excluded by default: > >>> > >>> mIgnoredClassNames.add("boolean"); > >>> mIgnoredClassNames.add("byte"); > >>> mIgnoredClassNames.add("char"); > >>> mIgnoredClassNames.add("double"); > >>> mIgnoredClassNames.add("float"); > >>> mIgnoredClassNames.add("int"); > >>> mIgnoredClassNames.add("long"); > >>> mIgnoredClassNames.add("short"); > >>> mIgnoredClassNames.add("void"); > >>> mIgnoredClassNames.add("Boolean"); > >>> mIgnoredClassNames.add("Byte"); > >>> mIgnoredClassNames.add("Character"); > >>> mIgnoredClassNames.add("Double"); > >>> mIgnoredClassNames.add("Float"); > >>> mIgnoredClassNames.add("Integer"); > >>> mIgnoredClassNames.add("Long"); > >>> mIgnoredClassNames.add("Object"); > >>> mIgnoredClassNames.add("Short"); > >>> mIgnoredClassNames.add("String"); > >>> mIgnoredClassNames.add("StringBuffer"); > >>> mIgnoredClassNames.add("Void"); > >>> mIgnoredClassNames.add("ArrayIndexOutOfBoundsException"); > >>> mIgnoredClassNames.add("Exception"); > >>> mIgnoredClassNames.add("RuntimeException"); > >>> mIgnoredClassNames.add("IllegalArgumentException"); > >>> mIgnoredClassNames.add("IllegalStateException"); > >>> mIgnoredClassNames.add("IndexOutOfBoundsException"); > >>> mIgnoredClassNames.add("NullPointerException"); > >>> mIgnoredClassNames.add("Throwable"); > >>> mIgnoredClassNames.add("SecurityException"); > >>> mIgnoredClassNames.add("UnsupportedOperationException"); > >>> > >>> * All classes in java.lang.* are excluded too > >>> * Annotation classes are not counted > >>> * Classes in the same package are counted (they won't appear in import > since it's in the same package so don't count imports to get class fan out) > >>> * Static method calls are not counted. So for example StringUtils from > Commons Lang never counts for class Fan out > >>> * Enums are not counted (no new XXX() done. That's why static method > calls are not counted too BTW) > >>> * Classes used in class extend or implement are not counted too. > >>> > >>> Hope it helps > >>> -Vincent > _______________________________________________ > devs mailing list > [email protected] > http://lists.xwiki.org/mailman/listinfo/devs > -- Denis Gervalle SOFTEC sa - CEO eGuilde sarl - CTO _______________________________________________ devs mailing list [email protected] http://lists.xwiki.org/mailman/listinfo/devs
