This is absolutely wonderful. Slashdot can offer up some real gems sometimes; this is one of them. Where do people like you hang out? I need smarter friends.
WELL SAID. The greatest thing about the dot com implosion was the pruning of the self-taught hordes.
Re:I was hoping they would wait.
on
New Red Hat Beta
·
· Score: 2
Huh? I've done 3 RH8 installs here at work as developer boxes, and screensavers do not work. Try setting a homepage in Konq (slashdot.org, for instance) - nope. And lots of people have complained about the panel; remove the clipboard and whatever else you have, then try getting rid of the blank area. You can't.
These aren't just bugs. They are not present in the default KDE install (from source or whatever). Some things were changed by Red Hat - basic things.
See dot.kde.org for many more problems.
As for removed features, I'd call using Konq as a full browser, with a homepage, a feature. Screensavers, too. The panel thing is annoying, but hardly deadly, I admit.
Re:I was hoping they would wait.
on
New Red Hat Beta
·
· Score: 1
1. Screensavers broken 2. Home pages in Konq no longer work; it comes up as a file manager automatically 3. The panel is gibbled; you can't remove blank areas 4....etc.
This was the greatest of all Commodore-related magazines; it had THE most technical articles and code (most of it assembly, or the insane read-data equivalent). Those guys totally maxed out the C64, the C128 and the Amigas, doing unbelievable stuff - I still remember writing my own beginning assembly, and sticking the code into that strange "protected" memory area, and then typing 'sys 49152' to execute...
It's been released as part of 3.0, as planned. So it's implemented, though I don't know how exactly (whether it's a KPart or what). Try typing webdav:// and see what happens.
You are 100% correct. Konqueror is, in fact, not much other than a container for various KParts. It's really, really easy to write software in KDE to view images, play music, etc. etc. simply by writing a few lines of code and including the correct KPart. C++ is crucial for this; GNOME will forever be hampered by their choice of C and the subsequent lack of first-class objects. When will Nautilus be able to copy files via SSH from a remote server and upload them via FTP to another server, all graphically? (In Konq, just open a couple of panes, type fish://servername and ftp://servername2, and then drag and drop. It's amazing.) Type audiocd:// to manipulate music. And so forth.
Sure, Cobol succeeded in its claim. Java is succeeding, too. They are both successful workhorse languages that get the job done. The "hype" you claim is so vile - write once, run anywhere, and so forth - may have been made before. But it's still nice to have; these are virtues. Java (the platform) does not deserve your scorn, as it is an excellent server-side solution.
Regarding containment: have you ever worked with J2EE? Object containment, like EJB servers. Or in MS world, MTS/COM+.
My comment about the lesser importance of specific languages is simply this: design is what counts. Bad design = bad implementation; no language can save you. Many people like to work with objects; J2EE is an object-oriented, heavy duty framework. Its attendant language is Java.
So by "multiple paradigms" (argh) you mean Python allows you to break object orientation. That's not an advantage for me, but maybe for you it is. I wouldn't want to be the one who documents that sort of code.
Cobol wasn't hypeware, either, so I don't get your point here. Cobol did its job well for many years, and now it's legacy code, mainly. It followed a natural lifecycle. Java will too. In the here and now, Java does a good job.
Python is nice. But Java is way more than just the language. Its got containment and infrastructure in the form of app servers, a vast number of APIs, a huge community, and a ton of third party support. From a pure language perspective, maybe Python is better, who knows. But really, it's design and infrastructure that count; languages are very often not as important.
By the way, what "multiple paradigms" (this from the guy that hates buzzwords and PHB speak) does Python support?
"Bosh! Many interpreters don't have such problems."
I wasn't speaking only about Java. In general, interpreted code is nice to have on servers.
"IMO, Java is hypeware and bloatware and PHB-friendly mediocrity at work."
It's pretty enduring for hypeware, don't you think? Anyway, aren't you the guy with the huge problem with object-oriented anything? Maybe a bit of bias showing through here?
Ease-of-management also comes into it. Migrating servers is a lot easier when you don't have to worry about recompilation; just copy over some.class files. That's the advantage of the virtual machine: no worries about weird big-endian/little-endian bugs, 32 bit vs. 64 bit, etc. Couple that with the fact that Java is the most mature JVM environment with great tools and APIs, and you can see why it's widely used.
I won't deny.Net will become a contender. But Java also benefits from competition in the app server space. Bea vs. IBM vs. iPlanet and so forth means higher performance and more features. Where is the competition for.Net's app server environment? In the MS Transaction Server/COM+ days, it was all just MS - no choices. Will the same be true for.Net? I think so, and that scares a lot of people.
Finally, Java runs well on *nix, and *nix is THE choice for big, serious sites. Sun makes Solaris, and Sun makes Java; that's all that needs to be said about that.
Also, by giving away StarOffice, Sun is promoting an open document standard. The KWord team, for instance, has hinted they are moving to adopt the OpenOffice document format. When MS gives away something, it's to advocate a closed format that only their tools can operate on. On the other hand, anyone can write a word processor (or whatever) to open StarOffice files.
I don't think you understand the level of integration and object reuse in KDE. If you are technically inclined, you should look at the code sometime. The programs are different, it's true, but they are to become one client with the functionality of the (currently) separate programs.
Before you spout bile, you should see what the current software has to offer, and then use your imagination a bit.
To be more precise: 1/3 of MS revenues are from MS Office. Other than Office and Windows, MS does not have a single market leader on the desktop, out of all their many products. So anything that affects the revenues of these two products is big news.
My point: the parent to my post was implying that the open source development method is somehow prone to tremendous bugginess, in this case with data storage/filesystems. The implication was that the often chaotic development of open source would lead to loss of data. I gave a refutory example from a prominent(!) closed source company.
I agree with your points about voting systems, in the main. Basically, having an open source system can't hurt and, as Michael pointed out, can lead towards true verification.
That is totally amazing. And for what it's worth, I think it's a worthwhile use of his time (not that my opinion on how someone uses their time matters, but whatever). I don't know, to me for some reason non-trivial acts of creation like this seem to touch whatever it is to be human -- our creative endowment is one of our signature traits, I think. What impulse would drive this sort of creative urge, to create beautiful mathematical shapes out of Lego, of all things? Whatever it is, it's mysterious, and it wasn't a waste of time. That figure eight knot is incredible.
Slashdot Mirror
This is absolutely wonderful. Slashdot can offer up some real gems sometimes; this is one of them. Where do people like you hang out? I need smarter friends.
We weren't "designed" at all.
Just the customers. You are only obligated to provide code to people who paid for the binaries.
WELL SAID. The greatest thing about the dot com implosion was the pruning of the self-taught hordes.
Huh? I've done 3 RH8 installs here at work as developer boxes, and screensavers do not work. Try setting a homepage in Konq (slashdot.org, for instance) - nope. And lots of people have complained about the panel; remove the clipboard and whatever else you have, then try getting rid of the blank area. You can't.
These aren't just bugs. They are not present in the default KDE install (from source or whatever). Some things were changed by Red Hat - basic things.
See dot.kde.org for many more problems.
As for removed features, I'd call using Konq as a full browser, with a homepage, a feature. Screensavers, too. The panel thing is annoying, but hardly deadly, I admit.
1. Screensavers broken ...etc.
2. Home pages in Konq no longer work; it comes up as a file manager automatically
3. The panel is gibbled; you can't remove blank areas
4.
http://uptime.netcraft.com/up/graph/?mode_u=off&mo de_w=on&site=travelocity.com
The site travelocity.com is running Netscape-Enterprise/4.1 on Solaris 8.
You mean Pacific southwest. Whistler and Blackcomb mountains are in the southwest of British Columbia, not the northwest of the United States.
YES! I used to love this magazine. Total classic gaming.
This was the greatest of all Commodore-related magazines; it had THE most technical articles and code (most of it assembly, or the insane read-data equivalent). Those guys totally maxed out the C64, the C128 and the Amigas, doing unbelievable stuff - I still remember writing my own beginning assembly, and sticking the code into that strange "protected" memory area, and then typing 'sys 49152' to execute...
Ah, memories...
I had no idea about the VMS emulation plans in the NT kernel. Excellent post, thanks.
From the KDE 3.0 release plan:
WebDAV support, Hamish Rodda...
It's been released as part of 3.0, as planned. So it's implemented, though I don't know how exactly (whether it's a KPart or what). Try typing webdav:// and see what happens.
You are 100% correct. Konqueror is, in fact, not much other than a container for various KParts. It's really, really easy to write software in KDE to view images, play music, etc. etc. simply by writing a few lines of code and including the correct KPart. C++ is crucial for this; GNOME will forever be hampered by their choice of C and the subsequent lack of first-class objects. When will Nautilus be able to copy files via SSH from a remote server and upload them via FTP to another server, all graphically? (In Konq, just open a couple of panes, type fish://servername and ftp://servername2, and then drag and drop. It's amazing.) Type audiocd:// to manipulate music. And so forth.
Sure, Cobol succeeded in its claim. Java is succeeding, too. They are both successful workhorse languages that get the job done. The "hype" you claim is so vile - write once, run anywhere, and so forth - may have been made before. But it's still nice to have; these are virtues. Java (the platform) does not deserve your scorn, as it is an excellent server-side solution.
Regarding containment: have you ever worked with J2EE? Object containment, like EJB servers. Or in MS world, MTS/COM+.
My comment about the lesser importance of specific languages is simply this: design is what counts. Bad design = bad implementation; no language can save you. Many people like to work with objects; J2EE is an object-oriented, heavy duty framework. Its attendant language is Java.
So by "multiple paradigms" (argh) you mean Python allows you to break object orientation. That's not an advantage for me, but maybe for you it is. I wouldn't want to be the one who documents that sort of code.
Cobol wasn't hypeware, either, so I don't get your point here. Cobol did its job well for many years, and now it's legacy code, mainly. It followed a natural lifecycle. Java will too. In the here and now, Java does a good job.
Python is nice. But Java is way more than just the language. Its got containment and infrastructure in the form of app servers, a vast number of APIs, a huge community, and a ton of third party support. From a pure language perspective, maybe Python is better, who knows. But really, it's design and infrastructure that count; languages are very often not as important.
By the way, what "multiple paradigms" (this from the guy that hates buzzwords and PHB speak) does Python support?
"Bosh! Many interpreters don't have such problems."
I wasn't speaking only about Java. In general, interpreted code is nice to have on servers.
"IMO, Java is hypeware and bloatware and PHB-friendly mediocrity at work."
It's pretty enduring for hypeware, don't you think? Anyway, aren't you the guy with the huge problem with object-oriented anything? Maybe a bit of bias showing through here?
Ease-of-management also comes into it. Migrating servers is a lot easier when you don't have to worry about recompilation; just copy over some .class files. That's the advantage of the virtual machine: no worries about weird big-endian/little-endian bugs, 32 bit vs. 64 bit, etc. Couple that with the fact that Java is the most mature JVM environment with great tools and APIs, and you can see why it's widely used.
.Net will become a contender. But Java also benefits from competition in the app server space. Bea vs. IBM vs. iPlanet and so forth means higher performance and more features. Where is the competition for .Net's app server environment? In the MS Transaction Server/COM+ days, it was all just MS - no choices. Will the same be true for .Net? I think so, and that scares a lot of people.
I won't deny
Finally, Java runs well on *nix, and *nix is THE choice for big, serious sites. Sun makes Solaris, and Sun makes Java; that's all that needs to be said about that.
The fact that it was reviewed by a guy named Gonzo John certainly lends it an air of credibility.
Also, by giving away StarOffice, Sun is promoting an open document standard. The KWord team, for instance, has hinted they are moving to adopt the OpenOffice document format. When MS gives away something, it's to advocate a closed format that only their tools can operate on. On the other hand, anyone can write a word processor (or whatever) to open StarOffice files.
I don't think you understand the level of integration and object reuse in KDE. If you are technically inclined, you should look at the code sometime. The programs are different, it's true, but they are to become one client with the functionality of the (currently) separate programs.
Before you spout bile, you should see what the current software has to offer, and then use your imagination a bit.
To be more precise: 1/3 of MS revenues are from MS Office. Other than Office and Windows, MS does not have a single market leader on the desktop, out of all their many products. So anything that affects the revenues of these two products is big news.
My point: the parent to my post was implying that the open source development method is somehow prone to tremendous bugginess, in this case with data storage/filesystems. The implication was that the often chaotic development of open source would lead to loss of data. I gave a refutory example from a prominent(!) closed source company.
I agree with your points about voting systems, in the main. Basically, having an open source system can't hurt and, as Michael pointed out, can lead towards true verification.
Kneejerk anti-OSS - I love it.
Michael is saying that open code and open hardware lead to a system that is fully verifiable, not that it's any more bug free.
Nice marketspeak, by the way - "vertically integrated", indeed.
Does the word DoubleSpace mean anything to you?
That is totally amazing. And for what it's worth, I think it's a worthwhile use of his time (not that my opinion on how someone uses their time matters, but whatever). I don't know, to me for some reason non-trivial acts of creation like this seem to touch whatever it is to be human -- our creative endowment is one of our signature traits, I think. What impulse would drive this sort of creative urge, to create beautiful mathematical shapes out of Lego, of all things? Whatever it is, it's mysterious, and it wasn't a waste of time. That figure eight knot is incredible.