Commons-primitives is vile and filthy
Just when it isn’t possible for the Jakarta inbred turdburglars to embarrass themselves any further, they release commons-primitives.
I am truly stunned by this. I mean, it’s one thing to be plain old incompetent, it’s quite another to elevate it to such breathtaking heights. What on earth were they thinking? Were they thinking? Did the entire group just go through a mass lobotomy as a team building exercise?
One of the main ‘justifications’ for this is performance, allegedly. Now I realise that commons twats are some combination of unemployed wonders and serial mental masturbators, but how on earth could they have missed that age old adage, ‘premature optimisation is the root of all evil’? There are very few people who can get away with it, Jakarta is blessed with not a single one of those people. What’s worse is that legions of halfwits will stupidly use this library with that Jakarta shit-eating grin we’ve all come to recognise so well. Nevermind the fact that for most applications, the performance gain is so trivial and insignificant that it really isn’t worth the extra jar and complexity of using non-standard collection classes.
The next ludicrous claim is that these misbegotten affronts to nature are easier to work with. Now maybe I’m old fashioned, but in my crazy world Iterator is a fuck of a lot easier to work with that DoubleListIteratorListIterator. Optic nerve burnt out yet? Well, if not, then I give you ListIteratorDoubleListIterator. If your brain hasn’t automatically shut down by now to protect itself from these vile names, then contemplate RandomAccessDoubleList.RandomAccessDoubleListIterator if you will.
November 13th, 2003 at 12:21 pm
The biggest problem seems to be that the jakara committers who might not be braindead or fond of their own reflections are also unwilling to spend the time to vote ‘no’ and justify themselves against the huge masses of their compatriots who’ll be offended by a negative vote.
November 13th, 2003 at 12:32 pm
I don’t think we should classify Jakarta as a whole as a bunch of idiots. I use some of the basic utility classes the commons group provides quite often (although I never venture into the area of using their more basic OO-looking proprietary objects”.
I submit to you:
jetspeed – useful
ant – useful
log4j – useful
lucene – useful
tomcat – useful (on some levels)
struts – useful, if for no other reason than like it or not, it’s become the MVC defacto standard, and development on it isn’t likely to die soon.
I mean… do I really want to write a search API from scratch? Do I want to use a less powerful logging system like the built-in one just because it exists? Do I feel like writing a portal architecture. No way. Are they perfect solutions? Of course not.
November 13th, 2003 at 12:43 pm
Anonymous said…
> I submit to you:
> jetspeed – useful
> ant – useful
> log4j – useful
> lucene – useful
> tomcat – useful (on some levels)
> struts – useful, if for no other reason than
> like it or not, it’s become the MVC defacto
> standard, and development on it isn’t likely
> to die soon.
Yes there are some usful apps in there, but correct me if I’m wrong but Struts (evil) was the only one created at Jakarta, all the others were donated to them.
November 13th, 2003 at 12:50 pm
anonymous you’re damning your own point with faint praise.
“Everybody’s doing it” is no excuse for stupidity.
Struts is a steaming pantload. The fact that
people use it a lot is NOT an indicator of its
inherent goodness but more an indictment of the crappy state of the status quo for open source server side frameworks in general.
November 13th, 2003 at 1:03 pm
For once, i have to agree with bileblogger on this particular one (tho not all his anti-jakarta rants). Commons-Primitives really seems to a hell of a lot of simple, redundant (TONS of redundencies), easily made by a home-grown code-generator pile of fluff…and its sole purpose is to save the developer from having to do casts and java.lang.Number boxing (like that’s the HARD part about Java).
and to make matters worse, the whole damn thing will be 100% obsolete in less than 8 months time when JDK 1.5 comes out, with autoboxing and generics.
so what really was the damned point of it all?
November 13th, 2003 at 1:38 pm
Blah blah blah… Apache is dumb… Blah blah blah… Somehow I know Marc Fluery must be involved with all of this… Blah blah blah…. I know I didn’t tell you anything about commons-primitives or why it is bad, but Apache is bad. Blah blah blah… Oh yeah, I hate JBoss… Blah blah blah…
November 13th, 2003 at 2:39 pm
Blah blah blah,
Stop Reading If You Dont Like The Content!
November 13th, 2003 at 3:05 pm
has Cocoon ever been spewed on?
November 13th, 2003 at 3:26 pm
Anonymous,
Right. Just like Hani can Stop Using Jakarta if he Doesn’t Like the Content.
November 13th, 2003 at 3:31 pm
Anonymous,
Don’t listen to Anonymous, she doesn’t know what the fuck she’s talking about.
November 13th, 2003 at 4:55 pm
Neither autoboxing nor generics make commons-primitives obsolete.
Dig a little deeper.
November 13th, 2003 at 6:24 pm
> Neither autoboxing nor generics make commons-primitives obsolete.
You are absolutely right! Commons-primitives are obsolete even without Autoboxing and Generics. ;)
November 13th, 2003 at 10:12 pm
Jarkarta commons-* is all ridiculous. If it were truly “common”, it would be a set of interfaces, not a “BagTreeSetStack ;-)” of utility methods. Wait, was that null?
November 13th, 2003 at 11:26 pm
I heard an interesting quote on the radio this morning: “The most dangerous ideas are not those that people are divided on, but those that people all agree on.”
November 14th, 2003 at 2:36 am
Feoh:
> Struts is a steaming pantload. The fact that
> people use it a lot is NOT an indicator of its
> inherent goodness but more an indictment of the
> crappy state of the status quo for open source
> server side frameworks in general.
Right, so the fact that everyone uses Struts is because there is no/has been no better opnesource alternative? I think we should remove the word opensource for a start. I can’t think of a single closed source web application framework in wide use.
jakarta hater:
Projects aren’t really ‘donated’ as such – that kind of implies that they are given to a new set of developers or something. When projects move to Jakarta its becasue the ASF is willing to donate resources etc. to aid their development. For me a Jakarta project can almost be considered a standard – there are things you can rely on the project having, like online javadocs, a relatively useful website, nightly builds etc. I have never downloaded a jakrata project and not be able to build it, which is more than I can say for a lot of other stuff out there.
Do I like all the Jakarta projects? Hell no, I just don’t use them unless I have to. I *like* the idea of the jakrat commons project and use some of them myself. Whats wrong with needing code to do something, finding it in a handy jar file, and then sticking it on your classpath?
Again, I don’t think that all Jakarta projects are great, and perhaps commons-primitives is vile and filthy (I haven’t looked at it in any detail), but who else out there aids the development of so many useful opensource java projects (with the possible exception of IBM)?
November 14th, 2003 at 2:38 am
I wrote:
> Hell no, I just don’t use them unless I have to
Opps, meant to say that I don;t use them unless there is a real need to. Early morning posts are a bad idea…
November 14th, 2003 at 3:22 am
When you say “an entire group of individuals”, you need to realise that we’re talking 3 asenine individuals. Maybe that makes life more worth living knowing that it was only possible for 3 f*ckwits to do it. If it had been 10 or 20, or the entire JBoss collective then ok thats something else :-)
November 14th, 2003 at 6:44 am
Hani, I’m puzzled. What is a “serious mental masturbator” ? How does one do it ? if indeed there is an “it” to do :-)
November 14th, 2003 at 7:27 am
“serious mental masturbator” – I think he said “serial”, as in “serial murderer”, meaning someone who commits the offence more than once (and not in parallel – for that, you need a government sponsor or two, some precision guided missiles, and a small country in need of regime change).
A “masturbator” – one who partakes in unproductive activity, purely for the pleasure of it. By extrapolation, I presume that “mental masturbator” refers to unproductive mental activity, which covers 90% of what programmers do (if you ignore the pleasure part).
November 14th, 2003 at 1:45 pm
>A “masturbator” – one who partakes in unproductive
> activity, purely for the pleasure of it.
unproductive activity??? Man, go see a doctor.. you may have a problem.
November 14th, 2003 at 3:20 pm
Rocco,
Stop being a wanker. I have kids, so I know what’s productive and what’s not.
November 14th, 2003 at 4:08 pm
You’re all wankers. Now stfu and slap mah jimmy fro!
November 15th, 2003 at 1:00 pm
1) A library to use when optimising is premature optimisation?
Publish an entry in which you damn the performance claims of the new library with some proof.
2) When did ‘easier to work with’ come down to the name of the class?
Isn’t that why all the corporate Borland lovers keep telling us to use IDEs? [and all the Intellij-ensia blindly follow their droning].
Anyways, can you start paying more attention to java-channel and get a broader range of projects to talk about? This all gets quite boring when it’s stuff we’ve seen already.
November 15th, 2003 at 1:06 pm
“The biggest problem seems to be that the jakara committers who might not be braindead or fond of their own reflections are also unwilling to spend the time to vote ‘no’ and justify themselves against the huge masses of their compatriots who’ll be offended by a negative vote.”
You don’t have to be a committer to be on the list and vote a ‘no’. True, it’s not binding, but your vote of ‘no’ shows that the ‘community’ does not want the package and your opinion and reasoning will have an influence on the decisions.
The fact you say ‘vote no’ suggests that you’ve never looked at an apache list before.
November 17th, 2003 at 7:02 am
Henri, umm …
Classnames aren’t important to you? What are you? A Perl haxor? ListIteratorDoubleListIterator? Wtf?
And as far as optimization goes, if you optimize, why do it half assed? If you do notice a performance problem handling primtive values with collection semantics: then STOP DOING IT! Don’t invent some half assed dumbass primitive collection! Use primitive arrays or something else! Why go through the trouble of adding a dependency to a piece of commons-* crapola?
November 17th, 2003 at 11:17 am
jwzrd,
“half assed dumbass primitive collection” – did you process your adjectives with a ListIteratorDoubleListIterator?
November 17th, 2003 at 12:18 pm
vagina dentia: http://radio.weblogs.com/0122027/2003/11/17.html#a100
November 17th, 2003 at 2:25 pm
blah blah boo hoo. Hani deleted my translations.
Me so sad. Would it help if I said something bad about Marc Fleury?
MARC FLUERY IS A BOOHEAD!
See, I’m just like you guys, a true blue bile clown. I mean clone. Umm. I’m cool too! Down with JBoss and Apache!
November 19th, 2003 at 5:48 pm
jwzrd wrote:
“If you do notice a performance problem handling primtive values with collection semantics: then STOP DOING IT!”
Um, yeah. What if you actually need a growable array of primtive values? You, know, like what if you have requirements that call for it?
and
“Don’t invent some half assed dumbass primitive collection! Use primitive arrays or something else!”
What do you suppose an ArrayIntList is? Why roll your own “growable list of ints stored in an array” when there’s one readily available?
November 19th, 2003 at 5:56 pm
Hani, go ahead and store 1m Integers in HashSet (it’s only 50+ MB)
After you’ve done that try the same with Fastutil’s IntOpenHashSet (6,5MB)..
Some of us need to manage shitloads of data, store it and handle it. Choosing to use faster and 8x less memory consuming alternatives hardly classifies as stupidity. The fastutil library is so much more efficient than the Collections ‘junk’ that one must truly wonder why Sun is not going to support generics on primitives.
December 3rd, 2003 at 5:56 pm
I have to agree on some… actually most of this. First off, I can up with a set of tools like this a year and a half ago (missed the boat again). The reason had nothing to do with performance. It was because I wanted to ensure that only certain types of objects were allowed in my collection. This was so I didn’t have to constantly cast, sometimes incorrectly, and it significantly helped my maintenance.
I mean, WTF? DoubleListIteratorListIterator? You dont make a double reference in a class/method name. Thats usually a sign you screwed up the structure.
Well the bottom line is most of their stuff is useable, if not lacking. I still like em overall.
December 3rd, 2003 at 6:18 pm
Ah – I didn’t read far enough… yea, I agree that struts blows. Why? Look at ASP.NET and struts. Like I said, Sturts blows.
Oddly enough, Salmon LLC. has a framework called SOFIA that is like ASP.NET on steroids. It blows struts away, and I cannot believe it NEVER gets press.
By the way people, do most of you know that a LARGE part of these open source projects are developed by employees of IBM and SUN? Especially true of Jakarta and Eclipse. Ever wonder why? Think about it…
As for the 50+MB HashSet, I have to agree too that SUN’s head is stuck far far up it’s own ass. Their collection stuff sucks, SUN supports open source and never uses it, made their own freak’in Logging classes, and many more.
YOu know what though? The best technology rarely wins, but that seems to be changing…
December 10th, 2003 at 3:24 pm
Doesn’t anyone prefer the Trove4J project over the Commons Collections?
March 13th, 2005 at 12:07 am
I love Jakarta Commons.
They’ve saved me A LOT of time, and Commons Primitives is nice to have in some cases.
Soooo… laaaaaaaaaaate.