All Things Microsoft > Microsoft Software

How to make your Windows machine more stable and secure

<< < (32/36) > >>

jtpenrod:
What comes to window managers, some applications would just totally die if you were to change the default windowproc or the default widgets.

This, right here, is a sign of a technically inferior design. There is no way that changing a window manager or a widget set should kill an app. This doesn't happen on a *NIX system, where you can run your GNOME or KDE apps from any other window manager. As a desktop, I'm not too damn fond of GNOME. However, I do like a lot of GNOME apps. It's NBD to run said apps from Enlightenment. The window manager should not be a factor.

I happen to like the windows look, and the desktop itself is configurable enough.

Do you really? Or is it that you have settled for it because you know that there is precious little that you can do about it? "Configurable enough" is good enough for some, but not for me.  :p

The higher level language you take, the less programmers have to care about lowlevel issues, and the less they will care about them. The solution is not to teach developers about lowlevel issues, but to move to even higher level languages.

I'm glad you agree with me: dumbed-down development environments lead to dumbed-down, lazy, don't care developers. Furthermore, just who is going to write the interpreters for those "even higher level languages" if noone knows, or cares to know, about "lowlevel issues"?  :p


--- Quote ---
Opening that information would be same as revealing trade secrets to your competitors. It's just a bad idea, so people won't be doing it unless everyone is forced to do it. The issue is especially touchy with the winmodems and winprinters, because a lot of the functionality gets implemented in software. If this software was free (as in speech), it would enable competitors to merely copy the hardware and use the same software for their own hardware. This would save a significant amounts of money, and companies are responsible for making profit.

--- End quote ---


This is nonsense. Lexmark doesn't make a damn thing from selling their crappy printers. These things go for about $90, or so. All the profit comes from selling the replacement ink cartridges which go for $30 a pop. Go through three or four and you've invested more in ink than you did in the printer. This is where they make all their profit: get the user hooked on a cheap printer, and you've got a customer for life. Besides, just what deep, dark secrets do you think will be given away from disclosing the prorocols for a modem? There is more hardware in a real, hardware, modem or have you forgotten this little fact?

BTW: nice attempt to dodge the issue  :D

muzzy:
The widget issue is because applications subclass widgets and expect some kind of default behaviour. You can change it and the apps will run, but some applications will not function nice with it. In the very least, ownerdrawn windows will have nasty look as they'll stand out and won't theme. I'm not aware of how X11 window managers do the theming, but I'd say the widget library has to cooperate there. The applications written against a widget library that doesn't support such things will stick out. How can such a system be any better, if theme support only comes for you if you use a specific widget set? The reason why gnome/kde apps work with other window managers, too, is that they likely fall back to some default rendering mechanisms on the widget library level.

You can have all the same theming functionality in windows. The issue is that a lot of the time people use their own custom widgets which are subclassed from the existing ones. If the app only uses standard widgets, there aren't going to be issues with skinning the windows, and there are many skinning systems for windows as a proof. They just don't work with every application out there, for a good reason.

Although I'm not personally going for skinned systems, this looks quite configurable to me: http://box.crackmonkey.us/screenies/2789.jpg
And I really do like the default look of windows apps. It's good and clean.

I don't agree that dumbed down dev environments lead to dumb developers. These developers already exist. Powerful dev environments enable these people to develop software, too, which is a good thing. The problem is that these clowns don't take any responsibility for their shoddy products, which is a bad thing. However, if they manage to sell their software to someone, all is fine. You don't have to buy crap, and if a crappy products does its job then it's not a crappy product, no matter what a pile of spaghetti code it might be.

Also, you don't seem to have a grasp of the reality if you're serious about your "just who is going to write the interpreters" view. There have always been and there will always be people who research the lowlevel stuff. Most of the really good highlevel people know about lowlevel issues, and know when to care about them and when to not. Also, interpreters and compilers can damn well be written in high level languages as well. If you're going to take the chicken and egg scenario out, I dare you research about how C compilers can possibly be written in C.

Regarding the drivers and hardware issues, I don't consider it dodging an issue. Definitely it still applies to graphics hardware and printers and so on. As for the winmodems, if someone makes hardware compatible modem which will work with same drivers, they can do it cheaper if they dont have to develop drivers. Winmodem makers have to make their profit, too. Regarding the ink selling, that's a business plan that sucks and they're realizing it when they have to sue people who make compatible cartridges. Such stupidity can't last forever.

jtpenrod:

--- Quote ---
The widget issue is because applications subclass widgets and expect some kind of default behaviour. You can change it and the apps will run, but some applications will not function nice with it. In the very least, ownerdrawn windows will have nasty look as they'll stand out and won't theme. I'm not aware of how X11 window managers do the theming, but I'd say the widget library has to cooperate there. The applications written against a widget library that doesn't support such things will stick out. How can such a system be any better, if theme support only comes for you if you use a specific widget set? The reason why gnome/kde apps work with other window managers, too, is that they likely fall back to some default rendering mechanisms on the widget library level.

--- End quote ---


I'm not aware of how X11 window managers do the theming...

That much is quite obvious. All along, you have been telling us that if only we understood Winderz, that we'd all stop the criticism, and come to love it. Well, here you are, criticizing Linux and you admit that you don't understand Linux. Oh sweet irony! :D  You see, the reason it works in Linux is that nothing is tied together  that tightly. The X Server handles low-level pixmaps and primitives. The X Server doesn't give a damn about widget sets and themes. This is strictly up to the window manager. Therefore, it doesn't matter that you have GNOME apps or KDE apps running on something other than GNOME or KDE, or that you're running Enlightenment apps (ETerm, GkrellM, etc.) on something other than Enlightenment. Win-d'ohs, by design, lacks that flexibility. Once again: I consider this to be a technical limitation.

Although I'm not personally going for skinned systems, this looks quite configurable to me: http://box.crackmonkey.us/screenies/2789.jpg

It's a Linux window manager port to XP. Funny how that works: Linux apps get ported to Win*, but when has any Win* app ever been ported to Linux?

However, if they manage to sell their software to someone, all is fine. You don't have to buy crap, and if a crappy products does its job then it's not a crappy product, no matter what a pile of spaghetti code it might be.

Still short-sighted stupidity. That crap spaghetti code is way more difficult, time-consuming, and, hence, expensive, to maintain. You can either pay knowledgeable coders who know what they're doing, what they're worth to write good code from the get-go, or you can pay them a whole lot more to fix the dumbass's code when that app needs maintenence.

Also, you don't seem to have a grasp of the reality if you're serious about your "just who is going to write the interpreters" view. There have always been and there will always be people who research the lowlevel stuff. Most of the really good highlevel people know about lowlevel issues, and know when to care about them and when to not.

Here, you try to insult me, and only manage to prove my original point.  :rolleyes:

muzzy:
Well, there are plenty of window managers for X11 and I was under the impression that they only managed window placement and borders. I tried quickly reading about it when I was replying, and found the some window managers (icewm) supported themes only this way (themed borders), while some seemed to support more. However, it's not a window manager thing anymore, it's a widget library which does the theming support. If you want widget library with theming support, you can do it under windows as well. Not a problem. I'm under the impression that you only get the theming to work properly under X windows if all applications you run are using the same widget library (or happen to support same theming mechanism). I'll research this later.

Also, that bb4win is not a PORT, it's a reimplementation and somewhat sucky at that, happened to be the only one I remembered quickly. As additional bonus, I think it handles skinning through external application, for which there are alternatives as well. This isn't an issue of application portability, this is just some kids reimplementing a toy for windows.

About application development, I'm sorry but in real world software is written in a goal oriented fashion, and what goes into actual code doesn't matter. Good code is always great thing, but it doesn't matter how good the code if something important doesn't work. And it doesn't matter how good the code if the damn thing works. For server applications, "small" used to mean "takes less than half the system resources" in some cases, and if your business needs a software that performs some task, you're going to pay even for crap if it's really what you need.

Also, I'm not trying to insult you. You're out of touch of reality if you think that all people would suddenly be bad programmers if there are some bad programmers out there. Obviously the guys with one year of visual basic experience won't be writing compilers and interpreters. Big deal!

Also, regarding to understanding a system, a point you made up early in the post. You seemed to enjoy just punching it in my face, as if we're having an ego fight of some sort here. I pointed out my lack of knowledge on the X11 WM theming details because it sucks to pretend to know something when you don't. I have an idea how I think it goes, but having never worked with the theming crap I don't have any concrete experience. Now you're somehow extrapolating that this means I don't understand linux.

From what I understand of the theming issue here, windows approach is indeed a lot better. The issue with defwindowproc replacement is applications that have been written to expect something of it although the system makes no such guarantees. This includes hacks like adding new buttons to captionbars through evil trickery and other such stuff, even done by big players. Windows XP changed how things render, and apparently had some attempt at theming. I don't know details of how it works, but I recall some applications didn't like it.

I cannot possibly see how a looser approach to doing things is better. The whole idea of pixmapped themes just doesn't fit the X11 design. You can no longer run the applications remotely unless you're ok with them eating shitloads of bandwidth. Advantages of X11 just don't work together with the modern crap.

muzzy:
Plus, it looks like you totally dodged the issue of configurability in windows by tackling some irrelevancy which happened to also be incorrect. I should put some more effort into writing these replies, you're happily redefining the subject here :)

Navigation

[0] Message Index

[#] Next page

[*] Previous page