Yeah, too bad HTML almost completely lacks usuable GUI components. Trying to make desktop-like applications using ajax/HTML is a waste of time. All UI components need to be built from the ground up using Javascript. And that is just slow. Worse than Java's AWT. The document model simply isn't designed for desktop quality applicaitons. They way to really leverage ajax is to use something like Mozilla's XUL. Now that is a real web application environment. What we need is for all browsers to pick up XUL (or something like it)
PErhaps you are confusing "not being able to" with "not wanting to" or not caring. I am a good sysadmin. I know networks. I know just about every OS. I can program in several different languages. I am a wiz with a CLI. I know technology pretty damn well. But as I get older (30 now), I notice that I am falling behind the bleeding edge.... and I don't care. I don't even own a cell phone. If someone thought I was stupid for not knowning how to download a new ringtone to a cell phone, I would just smile and ponder the irony. I think they are stupid for WANTING to download a new ringtone to a cellphone. Why can't I download a new ring tone to a cell phone? I don't want to know how to do it. It is beneath me. Eventually it won't be simple things like ring tones. It will be big things. Like i won't know how to use the latest "Neuropathic Computer Interface." Won't make me stupid. I know that I could do it if a really want to. I will have nothing to prove to anyone. I'll let my kids or grandkids do it for me.
You just wait. One day you'll be just as "clueless" as the people you scoff at now. The person you are responding to seems pretty certain that these are tech geniuses he was talking about. I think you should give them some credit rather than assume they are stupid rather than just aging.
Like I said, plenty of reasons not to run Exchange.;=)
The calendar's integration into Exchange's proprietary protocols is a major reason for its success and continued purchases of upgrades - but a Calendar program could just as easily have been built around HTTP/CGI, which would allow most email clients that support clicking on URLs to access it, and allow a much wider variety of client programs and clientless browser interfaces, so you wouldn't have to go to the overhead of firing up Outlook just to check your calendar - a major issue in a laptop environment.
I'm not convinced that people WANT to use a web calendar system. For the most part, I consider web interfaces to to be a lowest common demononator solution. A "catch all," if you will. Great for when you are out and about or if you absolutely need to support a lot of platforms, but for day to day operations, desktop applications are almost always superior. Like it or not, no "Exchage replacement" will ever take off without either its own Outlook-like client or smooth integration with Outlook.
I choose a *nix based on the package management, generally. Ports just doesn't cut it for me on a workstation. Ports is OK for servers where you install and let it run for years (I'd prefer Debian though).
Interestingly enough, this same pattern can be seen in some large IRC channels. Particulary tech channels. I remember chatting in #linux. The regulars/ops would get extremely intolerant of newbies and try to censor a lot of things. Eventually, a group would break of to start a new #linux* channel to be free from the "op-pression". And within months, that channel would be just as bad as the original. And the cycle would continue...
Last time I checked, OpenBSD does not perform well under heavy loads and can actually be unstable too (under heavy load). I can't find the study off hand though.
Not bad. We're actually looking to replace Groupwise. One thing groupwise doesn't do is group calendering (kinda takes the "Group" out of "Groupwise," doesn't it?). Looks like atmail does. it.
I tried the demo long before it was slashdotted, and it was slow then. I decided that it could have been a variety of issues so I installed it on my own server (test server dedicated to Zimbra). It was just as slow running on the LAN.
Dual Athlon 2400+ with 1GB of RAM as I was the ONLY user. It wasn't Zimbra itself that was slow. It was the AJAX heavy web interface that was slow. You know, like Java GUI's are slow... but worse than that.
Browser makers are still trying to standardize on the same javascript/DHTML implementation. I'm not holding my breath for the next "generation" of browsers. What is on the horizon? Mozilla's XUL is pretty neat and makes descent applications, but Microsoft is never going to support it.
...Exchange's hardware requirements are 10-100x more demanding than an equally-functional setup using, for example, sendmail and dovecot.
You have got to be kidding me! Sendmail/Dovecot doesn't even approach the functionality of Exchange. Not even close. Dont' get me wrong, there are plenty of reasons to not run Exchange, but lack of features is not one of them. There is a reason why Exchange uses so much resources. Microsoft programmers are not THAT incompetent. The bloat comes from feature creep, not so much bad programming. The question is, are you using all the features of Exchange? If not, one might consider something simpler like sendmail/IMP, but a lot of people like the group calendaring and all that.
Unfortunately I don't see this taking off. I installed Zimbra and tried it out myself and it is just too slow. The interface looks really good for a web application, but it is dog slow and very unresponsive to user actions. I can't imagine anyone using the web interface as the primary way of using Zimbra. If Zimbra ever takes off, it is going to need smooth Outlook/Entourage/Evolution integration.
Furthermore, I think this is a good as web applications are going to get. Lets face it people, HTML and web browsers are just not made to run desktop style applications. AJAX is really cool, but the simple fact is that HTML lacks the most basic tools to build a good GUI. The document model just doesn't work for sophisticated applications.
Honestly, I don't really care what businessmen think. And I am not really in a position to say what is the best way to managed patches to an operating system kernel. But I guess that is just me.
First of all, we're just talking about the kernel here. For most users it would make absolutely no noticable difference if the kernel patch maintainer dropped dead and it took a month to replace him or her. What you really need to be concerned about is your distribution and how it is maintained, which has very little to do with the kernel itself. Even without a central kernel patch maintainer, your distribution maintainers are still able to patch the kernel they ship with whatever urgent patches you require. It just won't be in the "official" Linux kernel right away. I'm sure this happens even now. I know Gentoo includes many custom patches to the kernels they ship.
Um, you know you don't have to compile a kernel on the machine it is going to be installed on, right? In Debian, it is really easy to generate a package for a custom kernel (and modules) to be distributed.
And even if there were a lot of gamer Linux users, most are willing to dual boot into Windows anyway. I know I did that with GTA: San Andreas. I mean, if you're going to spend several hours playing a game in one sitting, what's a two minute reboot? On the other hand, if were actually PAYING for WIndows, I might mind.:-)
I've heard a lot about how BeOS has great features that make it way ahead of its time. I ran BeOS for a short time back in the day. It was neat and seemed to multitask well, but I saw nothing special that couldn't be emulated on another system.
Name one thing that BeOS can do that no OS could effectively emulate.
I wouldn't expect the "general Windows desktop population" to use BeOS anyway. If any Windows user, it would be the more tech savvy and curious. Users who might (or already have) try Linux, but are scared of the commandline.
Slashdot Mirror
Yeah, too bad HTML almost completely lacks usuable GUI components. Trying to make desktop-like applications using ajax/HTML is a waste of time. All UI components need to be built from the ground up using Javascript. And that is just slow. Worse than Java's AWT. The document model simply isn't designed for desktop quality applicaitons. They way to really leverage ajax is to use something like Mozilla's XUL. Now that is a real web application environment. What we need is for all browsers to pick up XUL (or something like it)
-matthew
Yeah, and ajax applications are even slower than Java.
PErhaps you are confusing "not being able to" with "not wanting to" or not caring. I am a good sysadmin. I know networks. I know just about every OS. I can program in several different languages. I am a wiz with a CLI. I know technology pretty damn well. But as I get older (30 now), I notice that I am falling behind the bleeding edge.... and I don't care. I don't even own a cell phone. If someone thought I was stupid for not knowning how to download a new ringtone to a cell phone, I would just smile and ponder the irony. I think they are stupid for WANTING to download a new ringtone to a cellphone. Why can't I download a new ring tone to a cell phone? I don't want to know how to do it. It is beneath me. Eventually it won't be simple things like ring tones. It will be big things. Like i won't know how to use the latest "Neuropathic Computer Interface." Won't make me stupid. I know that I could do it if a really want to. I will have nothing to prove to anyone. I'll let my kids or grandkids do it for me.
You just wait. One day you'll be just as "clueless" as the people you scoff at now. The person you are responding to seems pretty certain that these are tech geniuses he was talking about. I think you should give them some credit rather than assume they are stupid rather than just aging.
-matthew
It isn't necessarily about overclocking. It is about doing it quietly. ;-)
I'm not convinced that people WANT to use a web calendar system. For the most part, I consider web interfaces to to be a lowest common demononator solution. A "catch all," if you will. Great for when you are out and about or if you absolutely need to support a lot of platforms, but for day to day operations, desktop applications are almost always superior. Like it or not, no "Exchage replacement" will ever take off without either its own Outlook-like client or smooth integration with Outlook.
-matthew
I choose a *nix based on the package management, generally. Ports just doesn't cut it for me on a workstation. Ports is OK for servers where you install and let it run for years (I'd prefer Debian though).
-matthew
Interestingly enough, this same pattern can be seen in some large IRC channels. Particulary tech channels. I remember chatting in #linux. The regulars/ops would get extremely intolerant of newbies and try to censor a lot of things. Eventually, a group would break of to start a new #linux* channel to be free from the "op-pression". And within months, that channel would be just as bad as the original. And the cycle would continue...
-matthew
Last time I checked, OpenBSD does not perform well under heavy loads and can actually be unstable too (under heavy load). I can't find the study off hand though.
-matthew
Not bad. We're actually looking to replace Groupwise. One thing groupwise doesn't do is group calendering (kinda takes the "Group" out of "Groupwise," doesn't it?). Looks like atmail does. it.
-matthew
Well, it is more complex. It does things like calendering which Gmail doesn't do.
I tried the demo long before it was slashdotted, and it was slow then. I decided that it could have been a variety of issues so I installed it on my own server (test server dedicated to Zimbra). It was just as slow running on the LAN.
Dual Athlon 2400+ with 1GB of RAM as I was the ONLY user. It wasn't Zimbra itself that was slow. It was the AJAX heavy web interface that was slow. You know, like Java GUI's are slow... but worse than that.
Browser makers are still trying to standardize on the same javascript/DHTML implementation. I'm not holding my breath for the next "generation" of browsers. What is on the horizon? Mozilla's XUL is pretty neat and makes descent applications, but Microsoft is never going to support it.
-matthew
Have you asked your users if they are willing to use the dog slow web interface? What were they using before?
It works, it is just slow.
You have got to be kidding me! Sendmail/Dovecot doesn't even approach the functionality of Exchange. Not even close. Dont' get me wrong, there are plenty of reasons to not run Exchange, but lack of features is not one of them. There is a reason why Exchange uses so much resources. Microsoft programmers are not THAT incompetent. The bloat comes from feature creep, not so much bad programming. The question is, are you using all the features of Exchange? If not, one might consider something simpler like sendmail/IMP, but a lot of people like the group calendaring and all that.
-matthew
Unfortunately I don't see this taking off. I installed Zimbra and tried it out myself and it is just too slow. The interface looks really good for a web application, but it is dog slow and very unresponsive to user actions. I can't imagine anyone using the web interface as the primary way of using Zimbra. If Zimbra ever takes off, it is going to need smooth Outlook/Entourage/Evolution integration.
Furthermore, I think this is a good as web applications are going to get. Lets face it people, HTML and web browsers are just not made to run desktop style applications. AJAX is really cool, but the simple fact is that HTML lacks the most basic tools to build a good GUI. The document model just doesn't work for sophisticated applications.
-matthew
Honestly, I don't really care what businessmen think. And I am not really in a position to say what is the best way to managed patches to an operating system kernel. But I guess that is just me.
First of all, we're just talking about the kernel here. For most users it would make absolutely no noticable difference if the kernel patch maintainer dropped dead and it took a month to replace him or her. What you really need to be concerned about is your distribution and how it is maintained, which has very little to do with the kernel itself. Even without a central kernel patch maintainer, your distribution maintainers are still able to patch the kernel they ship with whatever urgent patches you require. It just won't be in the "official" Linux kernel right away. I'm sure this happens even now. I know Gentoo includes many custom patches to the kernels they ship.
Um, you know you don't have to compile a kernel on the machine it is going to be installed on, right? In Debian, it is really easy to generate a package for a custom kernel (and modules) to be distributed.
And even if there were a lot of gamer Linux users, most are willing to dual boot into Windows anyway. I know I did that with GTA: San Andreas. I mean, if you're going to spend several hours playing a game in one sitting, what's a two minute reboot? On the other hand, if were actually PAYING for WIndows, I might mind. :-)
-matthew
That interview was an aweful lot like a press release. Is this guy making money on Samba somehow?
-matthew
I've heard a lot about how BeOS has great features that make it way ahead of its time. I ran BeOS for a short time back in the day. It was neat and seemed to multitask well, but I saw nothing special that couldn't be emulated on another system.
Name one thing that BeOS can do that no OS could effectively emulate.
-matthew
I wouldn't expect the "general Windows desktop population" to use BeOS anyway. If any Windows user, it would be the more tech savvy and curious. Users who might (or already have) try Linux, but are scared of the commandline.
AFAIK, this isn't about desktop users. This is about servers. ASP.NET is for web servers.
-matthew