.NET and Microsoft statistics

Can we dial back the insults and the trash talk? Daniel Fowler9 has a point—the document Piper cited suggests .NET has a faster, more efficient architecture (though maybe not quite as convincingly as she’d like to admit). Still, the critiques regarding the comparison between Sun's pet Store 1.0 and Microsoft's version hold water; you can clearly see the massive speed jump once the 2.0 version was optimized. Pet Store 1.0 wasn't meant to be some flashy commercial; it was just a baseline for using J2EE architecture. It’s obvious that .NET is more than just a solid alternative to the now much more widespread J2EE. As for what happens next, we'll just have to wait and see. Just less fanaticism and fewer jabs, please—let's stick to objective takes.

Maybe we could dial back the insults and the hostility a bit? I guess Daniel Fowler9 actually has a point here; the documentation mentioned by Piper seems to suggest that .NET is a faster, more efficient architecture.

Bryan Fisher7, I almost entirely agree with you, but I feel like the gentlemen in this thread aren't really accounting for the fact that J2EE is multiplatform...

One day, maybe, when .NET can run on Linux, Solaris, Windows, and so on... then we can truly start comparing them...

I suppose the person who absolutely needs raw speed will just stick to C/C++, while someone who prioritizes multiplatform capabilities and other features will just use Java...

So now, I'm honestly just confused as to why anyone would even bother using .NET, since it doesn't seem to offer the speed of C or the advantages of Java.????

Patrick Morris3 said:I mean, I seriously don't get why anyone would even bother using .NET at all, since you aren't getting the raw speed of C or any of the actual perks of Java, right?

Honestly, I’m convinced .NET only exists just to drive you absolutely insane! 😛
If the one single redeeming quality of Java—and looking at everything we've seen so far, that seems to be the only thing it has going for it—is that it's multiplatform while .NET still isn't quite there yet, then that's just plain pathetic. Because what you're basically saying is: I don't care if it's actually good or high quality, I just want it to be everywhere, no matter how much it sucks.
So why don't you apply that exact same logic to your supposedly "terrible" Windows builds and just say: I don't care about quality (looking at you, Linux)... I just need it to be ubiquitous (so, Windows)! 😁
It's pretty obvious that your logic is totally double-sided: you use one yardstick to measure every single flaw Microsoft has, but then you switch to this much more "flexible" ruler when it comes to things like Sun's pet Store 1.0 or whatever flaws come with open-source stuff. Now I gotta ask: why do you do that, why do you get almost orgasmic over the mere mention of Java, OpenOffice, Linux, Samba, Sendmail, or OpenXXXXX (and hey, feel free to swap out XXXXX for any product you think the "open-source crowd" stole from someone)? At the same time, you harbor this intense, fanatical hatred toward Microsoft, but you don't seem to feel that same heat toward other massive corporations like IBM or HP, right???

Daniel Fowler9🙄

Daniel Fowler9, let's just pretend for a second that you're ten years old and it’s perfectly fine if some things don't quite click for you yet...

Since Java is designed to be multiplatform, I guess it makes it harder to optimize...
The same goes for Linux being multiplatform; it makes writing it more difficult (though since it's open source, someone actually manages to build a better product than Microsoft does sometimes)..

Windows isn't multiplatform, so it should be easier to write (but even then, it's still riddled with security holes and ends up costing you way more than Linux does)....
DotNet isn't multiplatform either, which is why it's easier to optimize and maintain—which is honestly what you'd expect, especially considering Microsoft has poured more into DotNet than the entirety of Sun Microsystems is worth)....

And that point you made about "ubiquity"... it really doesn't hold up... I was talking about ubiquity in terms of computer architectures, and as far as you know, Windows isn't even ubiquitous across all computer architectures....

Patrick Morris3 said:Daniel Fowler9, look, let's just pretend you're a ten-year-old kid here and it's totally normal that you don't quite wrap your head around certain things yet...

The whole thing is that because Java is multiplatform, it’s way harder to optimize...
And since Linux is multiplatform too, it's tougher to write, though since it's OSS, some folks actually manage to build something better than what Microsoft puts out..

Windows isn't multiplatform, so it's easier to write (but honestly, it's still riddled with security holes and costs you a lot more than Linux does)...
DotNet isn't multiplatform either, which is why it's so much easier to optimize and maintain—which is exactly what you'd expect, especially since Microsoft poured way more into DotNet than the entirety of Sun Microsystems was worth....

And that whole point you made about "ubiquity"... it really doesn't hold water... I was talking about ubiquity in terms of computer architectures, and if you actually knew anything, you'd realize Windows isn't even ubiquitous across all architectures....

Well okay, if multiplatform solutions are just inferior when it comes to raw speed and how efficiently they use hardware resources, then why are we even wasting our breath talking about them? If that's the case, they aren't even worth our attention, and everyone would be better off just grabbing a non-multiplatform solution that actually runs fast, high quality, and just works better than any multiplatform junk.
Basically: just grab DotNet and a Windows server and toss those multiplatform behemoths from the past into the trash—you know, the ones everyone in the open-source world obsesses over (and most of the time, the only reason people love them is pure jealousy and just wanting to spite Microsoft).

Daniel Fowler9🙄

Well, I mean, if multiplatform solutions are truly inferior—both when it comes to raw speed and how efficiently they utilize hardware resources—then why even bother discussing them at all? In that case, they just aren't worth our attention. It seems like everyone would be better off simply choosing a non-multiplatform solution that actually works fast and provides higher quality than those multiplatform options.

I suppose if I were looking for some sort of primitive reasoning regarding all of this, then your answer would have completely satisfied me...

However, if one were searching for some actual intelligent thought on the subject, it would be difficult to succeed. Because, I guess, you would also have to consider the fact that most companies already rely on multiplatform solutions. It wouldn't make any sense for them to switch over to Windows because they would likely face bankruptcy... since, as far as anyone knows, those massive Microsoft licenses are certainly not cheap...
The second thing is that DotNet is a relatively new technology. Therefore, it becomes incredibly expensive if a company has to pay for retraining their staff who currently work in Java, not to mention the cost of the tools, which adds a bit more wealth to the equation, I suppose...
The third thing is that a single platform doesn't really exist. If a company ties itself to just one platform, they might end up going broke... and by platform, I don't just mean software SKU... because if you aren't careful, things move into 64-bit mode, and suddenly a company buys DotNet licenses for 32-bit, only to find that in three years they have to buy all those licenses again for a 64-bit solution... whereas, I think, you can just use Java on all kinds of solutions...

I mean, security is definitely a major factor here. Any company that actually takes its business seriously allocates a massive amount of resources to secure platforms—for instance, even Microsoft keeps its own DNS running on Linux...
So, I guess you really need multiplatform support. You’d want Windows for desktops, since that’s currently the most cost-effective option, while keeping those critical systems running on Linux or Unix environments...

Daniel Fowler9 said:By the way, I think I mentioned the Petco example before, but Jose Stewart7 criticized me for it. He claimed it was a deceptive Microsoft example designed specifically to push .NET over Java EE, even though reality suggests otherwise. The timeline for this Petco test went roughly like this: First, Sun Microsystems released that Petco application as a way to market Java EE. Then, following the launch of .NET, Microsoft rewrote that exact same application using .NET. At the time, Microsoft argued that their version was superior, faster, and easier to develop. After that, we saw those pathetic Java critics (kind of like Jose Stewart7) arguing that it wasn't a fair comparison because Microsoft utilized stored procedures and such, claiming the whole test was biased. Subsequently, Piper’s comparison emerged, which seemed to validate all those arguments from the Java detractors by creating an entirely new application and set of tests. In my view, those tests conclusively proved that, all things being equal, the .NET solution was vastly superior to the equivalent Java EE (semi)reindered mess.

You should look at http://www.middleware-company.com/j2...ench/faq.shtml
specifically the answers to questions like
Was Microsoft involved in this, did they fund this, where were the tests done?
The report states that a Microsoft employee was allowed to tune the .NET app. Were vendors of appserver A or appserver B involved, to tune their own appservers?
If you were to run the app on appservers A and B without using EJBs as many TheServerSide.com members have suggested, would it not run as fast as or a little faster than the .NET app?

And just to reiterate, PetSmart isn't an application intended for performance benchmarking; rather, it serves as an example of preferred architectural choices. My criticism is directed at the overall way Microsoft conceptualizes multi-tier architecture, which I find somewhat absurd. I suppose it's inevitable that this approach will generate followers like yourself—people who perhaps struggle to understand why non-trivial applications don't resemble a disorganized mess, or why such "elegant" solutions often result in widespread disapproval and turnover. 😁
That touches upon what I previously noted regarding the relatively low level of development culture found within the Microsoft-centric community.

Jose Stewart7 said:Look, I’ll say it again, PetSmart isn't some app designed specifically to be used for performance testing, it’s just a prime example of what people consider to be preferred architectural choices. When I’m critiquing things, I'm looking at the whole way Microsoft envisions multi-tier architecture, which is honestly just laughable, and of course, it creates followers like you who just can't wrap their heads around why non-trivial applications don't look like a total disaster zone, or why their supposedly "cool" solutions actually cause nothing but frustration and turnover. 😁
And that right there is exactly what I was talking about earlier regarding the low level of average developer culture within the Microsoft crowd alone

Honestly, I don't see any reason why those well-written versions—you know, the revised ones that the mid-sized firm rewrote for both Java EE and DotNet—couldn't totally work as a solid foundation for testing.

As for how Microsoft "thinks" about multi-tier architecture, that's really beside the point: they offered their way of doing things, and as always, it gets the job done. In this case, independent results are proving what we've known all along: DotNet beats Java EE hands down, and Piper's document covering the latest testing proves it, especially since they tried so hard to stay fair and not play favorites!

Anyway, now that we've all (hopefully) seen the light, there's no sense in beating a dead horse: so I'm calling on all of you to head over to the programming sub-forum and check out my new thread on building web services, where I walk through step-by-step how we can easily write a simple application using nothing but Notepad to search for books on Amazon.com via the command line!

Daniel Fowler9🙄

While you’re busy opening Notepad, I could be browsing through every book on Amazon.com. Where's the actual utility in that? It’s just like Microsoft—if they can find a way to make something unnecessarily complicated or slow, they'll do it.

As for their testing methods, I haven't seen them rely on anything remotely objective. If Microsoft pays for the study, you can bet the outcome is already decided. Using a fake application for benchmarks shouldn't surprise anyone.

Just look back at Intel and those EIDE controllers they claimed were faster than SCSI counterparts in their tests...

Daniel Fowler9 said:I suppose I don't see why the properly written revised versions—the ones the mid-sized firm rewrote for both Java EE and DotNet—couldn't potentially serve as a foundation for testing.

I suppose it isn't really comparable to the architecture used in .NET solutions. There are several shortcuts available there that might offer some short-term gains, but I suspect they eventually lead to a lot of unnecessary pain. 😁

As for how Microsoft envisions multi-tier architecture, I suppose that isn't really the point. They have presented their own specific approach which, as is usually the case, delivers results:

I suppose once you realize how critical this actually is, everything I’ve been saying will probably start to make more sense. It really comes down to that fundamental distinction between enterprise-grade solutions and those meant for home use.
By the way, I just read that Ford is making the move to Linux. http://linuxtoday.com/it_management/2003091500626NWBZDP How did this happen? I mean, seriously, what is going to become of this world? 😁

slycrane94 said:Look, I mean, I can just hop on Amazon and hunt down whatever books I need while you're still busy opening up Notepad, so where's the actual point? It’s honestly just like Microsoft—they'll find any way possible to overcomplicate things and slow everything down just for the sake of it.

Seriously, Axmi, what you just said is complete nonsense! I literally just laid out a whole detailed breakdown of why this is actually useful in the other thread! You really should have at least given it a quick read before jumping in like that! 😠

Daniel Fowler9🙄

Jose Stewart7 said:It just isn't comparable to how .NET solutions are architected, you know? There are all these little shortcuts in there that give you some quick wins early on, but man, they end up causing you a massive amount of grief down the road. 😁

[/B]

Look, once you finally wrap your head around how crucial that actually is, everything I’ve been rambling about is going to click for you. It’s really just a matter of understanding the difference between enterprise-grade stuff and something built just for home use.
Oh, and hey, I was just reading that Ford is switching over to Linux http://linuxtoday.com/it_management/2003091500626NWBZDP—I mean, seriously, how does that even happen? Honestly, what is this world coming to? 😁 [/B]

Man, I honestly don't have the energy for this anymore... Didn't you say earlier that you were totally done with arguing with me??? So why are you breaking your own rule now??? 😕

Daniel Fowler9🙄

Daniel Fowler9 said:Look, I honestly don't have the energy for this anymore... Didn't you say earlier that you were giving up on having any further discussions with me??? Why are you going back on your word now??? 😕

Daniel Fowler9🙄

I suppose it is because you are addressing me directly.
By the way,
did you happen to look at the link 🙂
?

Daniel Fowler9 said:Axmi, what you just said is complete nonsense! I explained the benefits in great detail on the other thread! You should have at least read it first! 😠

Daniel Fowler9🙄

I haven't seen a single intelligent explanation yet. I spend most of my time on Amazon—that's where I buy all my college textbooks—and since I always have a browser open, why would I bother using a command prompt that offers way fewer options than the actual Amazon site? I find what I need, toss it in the cart, and I'm done. But here, I’d have to go back to their website anyway just to actually buy the book. Is there really any time being saved here?

slycrane94, man, seriously, stop embarrassing yourself.

Bryan Fisher7 said:slycrane94, man, just stop embarrassing yourself.

Don't be an idiot. I might not be a developer, but that doesn't mean I don't know how to use software.

Why bother making a program that’s completely useless except as a demo? If you're going to build an example, why not make something actually useful, like a desktop widget for the weather or a TV guide? Instead, we're fighting over Amazon. Honestly, Amazon's site is perfectly functional as it is. You always have your browser open anyway, and it's much easier to order books there—something you can't even do through this garbage.

Bottom line: the utility of this program is zero. If you think that's some kind of win, then enjoy it.

Man, whoever appointed you as a moderator seriously has no clue what they’re getting into. Either put a paper bag over your head or just stop showing up to these threads, because honestly, you're just making yourself look bad.

I'm not even going to touch Daniel Fowler9's "fanboy" threads—I hate those just as much as the usual OSS cheerleading posts—but the example in her thread is just a basic web service demo; it wasn't meant to be a standalone tool. Anyone actually reading the post for the right reasons is going to get plenty of value out of the code itself. If you don't care about how web services work, there's no reason to be in that thread, let alone comment on the post or the code. It's actually better that the example isn't a finished product; its whole purpose is to demonstrate the concept. That's what helps someone actually learn enough to build something useful for themselves later. If you don't get that, then this IMF is in serious trouble if they think you're fit to moderate software discussions.

As for her little bias ("IE, definitely not Mozilla" accompanied by all the emojis), I'll pass. I don't want to trigger another one of those endless MS vs. OSS debates that seem to be taking over everything lately.

From what I've seen, people learn way more effectively when they're working with practical, real-world examples. It actually gives them a reason to care, and maybe even sparks enough curiosity to start tinkering with new features. That’s just not going to happen with her example.

If she had done the exact same thing for Google, you might actually be able to argue there was some utility to it—though it probably would have been better as a browser extension or something that her "brilliant" .NET technique could actually improve.

Just forget about being a programmer for a second and ask yourself: how much would you actually want to learn from examples that are completely useless?

It seems like you're struggling to grasp the point, so I won't waste my breath. My argument is simple: the actual purpose of the app doesn't matter. What matters is understanding the core logic, which is why the clean, straightforward example provided by Daniel Fowler9 is so useful. It’s not about what the code does; it’s about grasping the underlying principle so you can apply it elsewhere.

Your bashing isn't even based on facts; it's just a reflection of your biased, anti-Microsoft stance. If the example had been anything else, you would have found a way to tear it down anyway. As a moderator, you really ought to try being a bit more objective.