MozillaQuest is running a nice little story on the upcoming Tabs feature in Konqueror. "The K Desktop Environment (KDE) certainly has done lots to narrow the gap between the Linux desktop and the Microsoft Windows desktop. And the addition of tabbed-browsing to KDE's Konqueror browser is one more large step in closing that gap. In our opinion, the K Desktop Environment already is just as good as, if not better than, the MS Windows desktop." Stay tuned for the next alpha!
Dot Categories:
Comments
It will be nice to have tabbed browsing but hows about fixing all the bugs in konqueror and getting java to work correctly etc before adding new features. For example connect to any webmin enabled box using https, in V3.0 it gave you broken pipe in 3.0.1 it just sat there after entering the login and pass and in 3.0.2 it does the same, one of my old boxes still has konq 2 on it and it works fine. Yes I have logged these as bugs but keep getting asked how to reproduce it can I send a site? No just install webmin and the perl modules to enable https and hey presto every time.
I love Konq easy to use light and in the later 2 versions was fairly quick but 3 is totally broke for me in most things I use it for, please lets have all the bugs fixed before adding new features.
I can't comment on your specific bugs (except that Java should really work if you use kwin in 3.0.2, and that in HEAD it supports Netscape's LiveConnect), but please, before complaining and suggestting "KDE should do this", try to get enough knowledge about what you're requesting so it at least make sense. Tabbed browsing is being developed primarily by a single developer, who is not really involved in the the HTTP/HTTPS transport and the HTML rendering; thus one being developed is no harm to the other... And, no, it's not a simple matter of having any developer switch to the area that needs some bug fixed; sometimes it is possible, but for things as complex as HTML rendering, it would take someone quite a lot of time to get acquainted with the code well; and that alone doesn't guarantee that the changes they'd make would not produce any regressions.
There were some fixes for Webmin committed recently which will be perhaps backported to KDE 3.0.3.
How about fighting worldwide hunger and wars first before posting comments on newsboards?
Perhaps you are forgetting that posting comments on newsboards do not increase worldwide hunger and wars... adding new features does add new bugs and burries old ones deeper.
You mean, you'd rather don't have tab browsing because it might have bugs? Or do you fear
that tabbrowsing mighty break https even worse?
yes, I think you will be surprised to know that there are *lot's* of users who would rather use a software with less features, but reasonable "bugless".
I'm by no means being overly critic, what the guys have done with KDE for free is fantastic, I just happen to agree with the previous post.
Jose
The first option is unavoidable, there is no bug free code and new features bring new bugs.
But although tabbrowsing doesn't have anything to do with https, is somewhat that. I like new features and KDE is getting better every day but maybe some more investment in the "new features -> fix bugs -> new features" cicle would be the best at this time instead of "new features -> fix some bugs -> new features -> new features"
I seemed to have recently had this problem with webmin, but it's working fine for me now with KDE 3.0.1 on SUSE 8.0 so ...
I agree that there is a need to shake out bugs before adding new features. But it seems that developers care more about adding some useless eye-candy than fixing bugs.
I think it's better to have a rock-solid and bug-free KDE with the current set of features than adding new ones forgetting about things that need to be fixed.
Sounds great!
Why don't you pull out your favorite text editor and start hacking the code to get all those pesky bugs killed?
Seriously, remember that KDE is free software. If someone want to add a new feature to Konqueror, that is his/her right under the GPL. Tabbed browsing is a feature that a lot of users want, and you can be rest assured that any existing bugs are being investigated. It's not always easy to kill a bug in software because killing one bug might disable something else or create new bugs. Give the developers time to do it right and don't complain unless you are paying that developer's salary.
There is a big difference between complaining and submitting a bug report.
I do not have much time to spend fixing KDE's bugs but I have done it in some occasions.
Of course anyone can add new features to KDE but there is someone in charge of the KDE release cycle isn't there?
Yes, bugs are being investigated and they aren't always easy to kill but some persist across KDE releases, some of those being just a matter of 2 - 3 lines of code (see first paragraph).
People should complain, that is the only way to get it even better, and should submit bug reports (I surely do it every time I find one), in fact people should do both.
And talking about bug reports... KDE's bug report system is awful and messy maybe they should consider bugzilla.
MozillaQuest is no good news site: This is no real news anymore, they just took the screenshot from http://konqueror.org and failed to mention Galeon as tabbed-browsing Linux browser.
Yes it's news to many people. Don't read it if you don't want to.
What I think he/she was referring to is that MozillaQuest is a "troll site". For the past two years or so, it's been cut-and-pasting the same articles over and over again and changing the dates. Basically, the guy behind it (Mike Angelo) is a Slashdot troll pretending to run a news site.
If Konqueror can get positive press, that's great. But I'd rather it got positive press from a reputable source. KDE shouldn't have to scrape so low into the journalistic sewage bucket to get a positive review.
There's nothing wrong with this article that I can see, sorry but your argument is vague and doesn't hold.
Actually, I didn't mention the article at all! I said that for the past two years, he's been cutting-and-pasting articles and changing the dates. There's the specifics. You have the link to the site. Read through the trash it's printed in the past. It's adolescent chest-thumping, not journalism.
If you happen to be fortunate enough to have access to plagiarism-detection software, run the text on these links through it:
http://www.mozillaquest.com/Mozilla_News_01/Mozilla_RoadMap_21jun01_Stor...
http://www.mozillaquest.com/Mozilla_News_01/Mozilla_0-9-3_branched-01_St...
http://www.mozillaquest.com/Mozilla_News_01/Mozilla_0-9-4_delayed_Story0...
http://www.mozillaquest.com/Mozilla_News_01/Mozilla_0-9-5_released_Story...
Of course, it's not illegal to plagiarize yourself. Just lazy.
I looked through these and they mostly seem different articles. Mozilla releases a snapshot every other day, so who can blame him for reusing text though when even the KDE announcements seem to do that. :-)
KDE announcements, although frequent, are actually rather informative and factual. They may be a bit boring, admittedly, but they're hardly MozillaQuest quality.
MozillaQuest generates articles by playing mad-libs with the template:
"Mozilla was scheduled to release their [milestone] milestone on [totally made up date], but it has been delayed due to bugs and bad management. It will likely be the buggiest milestone ever. Is Mozilla sweeping bugs under the carpet? As of milestone [milestone], Mozilla had [totally made up number] open bugs. We recommend you [something stupid] instead of downloading anything more current than [really old milestone]."
Have you read the articles all the way through? After a while, it's all one article. If there's nothing new to say, you usually don't print anything. This guy prints the same freaking article time after time after time. And he never prints a "corrections" section, which is rare in web media, but rarely as warranted as on MozillaQuest.
I like KDE. I think KDE is a good product with good people behind it. They should not simply lap up any positive attention they get. MQ is way below KDE.
Ok ok. Look I didn't know all this. Why don't you relax a bit? When we don't post articles people complain. When we post articles people complain. We just can't win. If you don't like the article quality how about you submit some good ones?
MozillaQuest is a well known Mozilla bashing site, so no wonder it praises Konqueror. Right.
Yep. Check out http://www.mozillaquestquest.com/ for a great parody of the MozillaQuest site.
So your unhappy that MozillaQuest gave Konqueror a good review? Your unhappy that KDE.NEWS published a none KDE sites positive feed back of Konqueror? Your unhappy that MozillaQuest cut-and-pastes articals from other websites? Is there a web site that doesn't cut-and-past other website articals? Please explain???
I wish people would remember that Opera had tabbed browsing well before Mozilla, or Galeon. Opera, OK, OPERA!!
I have exactly the same problem with Konqueror. When I mailled kde about it I was told they "thought" it was a bug in a module of code and they "think" they have fixed it and I needed to install the latest CVS to see if they had.
Bah. I have better things to do with my time than muck about with CVS trees. I thought the idea of KDE was to entice new users and ex windows users? Right so I report a bug and I am told to try the latest CVS release. Scenario: I know nothing about Linux (I actually do, been using it on and off since RH 5.1). Given this, er, what is CVS? How do I install from it? Sod that, don't have to do that in Windows, I just download an exe and presto. What is wrong with releasing a patch rpm for the problem? Or a tar.gz with a shell script? And don't say it is because you don't get paid open source blah blah I know a lot of freeware people who write stuff for Windows who manage to release beta upgrades and patches etc without messing around with a CVS tree.
At the moment my browser of choice is Opera 6 and my, how nice and fast is that, STILL the best tabbed browser imho. However, if KDE had fixed the bugs in Konq, IT would STILL be my browser of choice just like it was in my KDE 2 days. Not anymore. Sort it out KDE or I am off back to Gnome, ver 2 looks quite nice.
A shame really, KDE was looking the most promising of the two.
Now theres a point I am a poor sod who is stuck with a single ISDN connection due to BT's inability to convert my exchange to ADSL but that another story. Why not release just updates or service packs for we poor mortals on modems etc instead of a full download each time someting is updated to a new version?
Well binary service packs do make some sense, but a simple single compile option change, could make those service packs VERY large.
The service packs are different for every distribution (so I guess that makes it the job of the distribiter).
You could compile every stable release from source. That way, you would only have to download the differences between the source codes.
Johan Veenstra
Hi
Fixing the problem is all the KDE developers can do. They can't install the patch on your system. You need to fetch it from CVS / wait for the next release yourself.
> What is wrong with releasing a patch rpm for the problem?
Nothing - except that KDE doesn't release any RPMs at all. All RPMs provided on the KDE site are from packagers FOR the project, but not BY it.
> Or a tar.gz with a shell script?
This is called cvs. No develper can provide you with a binary-patch for ALL distributions around.
> Sort it out KDE or I am off back to Gnome, ver 2 looks quite nice.
You wanna threat? Well - just switch! We won't miss you. Why didn't you provide the RPM yourself instead of whining? tststs... ;)
CU
Andi
Excuse me, but I thought the aim of the likes of KDE and Gnome were to bring Linux closer to the desktop market.
I don't expect them to install it. They could reduce the headache though. And as Colin said previous, why do you have to download a whole new distribution of KDE just to fix one bug. It makes no sense.
You don't need a binary patch. A shell script can be made to do a compile and install. Give it some multiple choice options and away you go.
"We won't miss you." Hmm that's very community minded innit. I put a structured criticism together and u flame me LOL. As far as me providing the RPM myself, assuming I know how to why should I? Linux WON'T corner Windows the way most want it to if the ease-of-use feature responsibilities are hung on the end user. Hence why Linux hasn't already cornered it today.
You are asking KDE to do the job of your distributor. KDE provides the fixes. The distributor is responsible for keeping you happy with updates and upgrade paths. It's a very real distinction although it's hard for some people to understand.
KDE is developed towards the distributions
It's not developed towards the users that, afterall, it's the people that use it. Distributions can screw up KDE, but KDE as a project won't care.
I suggested them to keep just one version of binaries for i386 following a standard file system organization. This wouldn't be much but they allways say that KDE will NEVER release binaries!
Having a binarie version they could just say: "Hey I recompiled some soft/librarie that had a bug! Just overwrite xxx.so, etc....", but BINARIES produce alergies to some kde developers although it would be so great for ALL users :-(
I won't ask for that again... I will use KDE because it's good but I'll have my eyes open to new fresh ideas also.
> I suggested them to keep just one version of binaries
> for i386 following a standard file system organization.
for which OS? Linux, FreeBSD, NetBSD, OpenBSD, Solaris i386, SCO, what?
for which libc?
for which compiler?
this is (partly) why KDE produces source and leaves the binaries to the vendors.
Hey! I'm not an idiot.
You know already.
You could compile it for what would suit what big majority of KDE users have.
i386 and linux.
YES! Let's solve the problems of the big majority!
I don't know which libc. I'm not a developer, but I'm sure you know!
You might be a developer, but we, the users, are not stupid.
And other platforms could still use the soft to compile. The same way they do now... But at least joe users could take great benefit from that move.
> i386 and linux.
ok, now let's pick a version of linux. realize that binaries don't run on every version of linux out of the box.
> I don't know which libc. I'm not a developer, but I'm sure you know!
no, i don't know, because different distros (and different versions of the same distro) use different versions. that is part of the problem.
and why the LSB is so important for the future.
> You might be a developer, but we, the users, are not stupid.
hey now, i never called you stupid. i just pointed out the obvious problems. if you weren't aware of them before you are now.
now, that said, if you have a target platform in mind nothing stops you from building those binaries and uploading them somewhere for distribution. a fellow made a HUGE slew of KDE 3.0.2 binaries for RH7 himself, so it is quite doable.
Suse binaries run on RedHat and on Mandrake. I don't know debian, but it's quite probable....(I guess that makes 90% of Linux boxes)
And I was talking about delivering packages for LSB compliant distributions.
Of course I'm not saying to keep binaries for all the universe of distributions.
> fellow made a HUGE slew of KDE 3.0.2 binaries for RH7 himself, so it is quite doable
If it's so doable, why KDE developers cannot deliver the binaries and say, "KDE has binaries that the majority of our users can install". And also:"Distro X and Y and Z are LSB compatible, hence the binaries will install stright away"
> Suse binaries run on RedHat and on Mandrake. I don't know debian, but it's
> quite probable....(I guess that makes 90% of Linux boxes)
which versions of each? how are dependencies handled? how is it layed out in the FS?
yes, you are right that once the LSB is adopted and the distros are all using the same gcc version (or one that is ABI compatiable) then things will be peachy. that isn't the world we live in today. and people were asking why this isn't done today.
> If it's so doable, why KDE developers cannot deliver the binaries
the answer is in the question: they are developers. they are busy fixing the bugs and adding the features others in this thread are complaining they don't. some of them don't even use linux or i396 boxes!
no, the real question is: if it is so doable and you want it so badly, why aren't you doing it?
>no, the real question is: if it is so doable and you want it so badly, why aren't you doing it?
You said it first that is "doable". I don't find it so doable. :-)
If I knew how to do that I would have never asked to get those binaries ;-)
But thanks for your explanations.
Anyway, do you really think we'll have the LSB in future?
Will distros adapt it?
I don't know much bout distro politics, and I also don't know if it would be difficult for them to change their internal structure... ¿?¿?¿?
Any hopes?
the next major version of most of the linux distros will be LSB complian if they aren't already. so, yes, LSB is a definite thing.
No, SuSE binaries for KDE will not work reliably on Mandrake 8.x or RedHat 7.x since they use different g++ versions (2.95.3 and RH-2.96).
And, BTW, LSB is not anywhere near complete enough to make SB-compliance sufficent for one KDE install to work. It doesn't specify the C++ ABI for starters.
> If it's so doable, why KDE developers cannot deliver
> the binaries and say, "KDE has binaries that the
> majority of our users can install".
You make it sound like there is this mythical wonderful "KDE developer land" that is a separate place, where they all get together and laugh at users being unable to install KDE.
*YOU* are the developer. Anyone that uses KDE is the developer. If you want it done, make binaries, and show the world that they can get binaries for KDE! If you don't have the time to work on putting those binaries together, hire someone, or contact the discussion list with a *real plan* of how to make it happen.
Most of the great ideas that make it out into the discussion forums on KDE don't die because they don't scratch someone's itch, they die because the person asking for such-and-such feature isn't even willing to take the time to put together something other than a pie-in-the-sky "wouldn't it be nice" one-sentence e-mail.
If you sat down one evening, and said, "What do I need?" and wrote it down, and then came to the KDE list, not begging, but asking as an equal, for people to help make it happen, I bet you would have "official" KDE binaries.
The problem is that there are too many people willing to come up with ideas, but they're not willing to see them through.
I don't know a lick of C or C++, but I knew enough to get KDE ported to MacOSX by finding the right people to help me, saying, "Look, I got 30% of this working, so I know it's possible. Anyone want to help?" You don't have to be able to do everything to be a *part* of the KDE community, rather than just someone on the outside, wondering why his feature request never got implemented.
"and why the LSB is so important for the future"
I agree that the LSB is important for Linux and also KDE, and it is POSSIBLE to make ONE RPM for ALL LSB-LINUXVERSIONS.
If we don't make the installation (on all points and all programs) better and eassier we will NEVER stand a chance against MS-Windows. LSB is the key to a solution for this. Lets use it when ready.
Go for standards!
I am a developer, but i have no idea which version of libc the majority of linux distributions use.
You are asking for something that sounds simple, but it isnt. Hopefully with time these issues will go away, but for now pretty much the only way to make a set of binaries that will run on almost all systems is to statically compile the programs, which is AFAIK not even possible to do with kde, and even if it was im sure it would make it even bigger and slower.
The thing that spring to mind is getting Linux on the desktop and currently it does not stand a chance in the corporate enviroment because it does not have the ease of updates across all distros etc.
Is it not time that all this superior knowledge was put together to make Linux the best in the man on the streets eyes? To do this we need an easy to use update system for all distros, where joe bloggs can just press a button and it is done.
United Linux may be a start but more is needed to get it mainstream. The thing that killed Unix in the past was its diversity amongst dialects so lets not do it again. Alex is right above although he could have choose better words to express himself. Whether KDE or Gnome I dont care if one is more stable I use it but I prefer KDE.
We have to get away from this squabling amongst ourselves and make Linux the best there is and prove it the Joe Bloggs or we will stay as we are a minority forever.
http://www.linuxbase.org/
Im sorry, but this is just not true. I think linux is very much ready for many corporate uses. I think the biggest challenges is to make it ready for the average home user.
In almost all corporations they have an IT department that handles all hardware and software setups and upgrading, the normal employee dont have to worry about it. If they do then something is wrong and i wouldn't want to buy stocks in that company :)
Are linux systems hard to upgrade? Hardly. If you use redhat you have the redhat network. If you are using debian you can just make the machine do an apt-get dist-upgrade once a day. Im sure that suse and mandrake have something similar. It really is as simple as pushing a button. Well, in debians case, typing in a single command. Please stop claiming that it is hard to upgrade, it just isnt true anymore, and havent been for quite a while now.
There is no big problem with diversity, a company can just pick one and just one distribution for the entire company. It might be a problem for commercial software vendors, but if you are afraid you cant run their software then use redhat, which seems to be the one distro that everyone supports.
Im not saying linux is perfect, it isnt, (yet :) but upgrading is not its problem, i even find this easier to do than it is in windows.
At the moment upgrading or installing programs on Linux sucks. I use mandrake 8.2 and many times, (when installing the right rpm for this distro) I get an error-message and the rpm doesn't install. In a worser case I get a broken system. All those different versions of Linux is not good, there are to much of it.
Hopefully things will chance when in about a few months most of the distro's are based on LSB (Linux Standards Base).
I disagree, i think installing software on some linux systems are very easy as long as you use packages made for your exact system.
In debian all you gotta do is type "apt-get install programname" and it will download it and install it for you. I know that there are graphical frontends for it but i havent tried any of those.
I also remember mandrake having a graphical tool that would let you download and install packages as well as resolving the dependencies.
I think most of the problem is that many users are trying to install software the windows way, which is indeed not all that easy. The dependancy hell as many refer to is caused by the fact that a linux system is much more modulized than a windows system. This is one of the strengths of unix, but it does make single package installation more complex. But i think the main problem is not how things are working, but that the users need to be reeducated to use the new tools correctly, just as you had to learn how to do things in windows years ago.
I have never said that installing programs is difficult, but there are too many problems involved in that procces of installing a RPM, even when it is the right one for your system.
In fact I do believe that RMP has the potential to become the best and most easy to use system for installation and upgrading your system.
But at the moment things are goiing wrong because of missing standards. And if projects like the LSB succeeds, they will be solved.
> We have to get [...]
This is my big gripe. By "we" you mean "not me at all, but possibly some other people". Please, be part of the "we". If you want something like this to happen, be part of the process! Start emailing distros and organizing statdards discussions.
I think one of the ways of summing up a lot of the feeling of the KDE developers is this:
Most of us got involved because we saw something that we thought would be good to fix or add (not because of super human coding skills or because we just couldn't find anything else to do with the time). By people saying that they want the "KDE developers" to do this or that, they're basically saying that they expect something from others -- volunteers mostly -- that they aren't willing to do themselves.
You're saying "not only do I want, I expect you to help me", when you won't give us a hand. I think that's what a lot of this "we're not paid to do what you want" stuff comes from. It's not just that we're poor -- we just expect you, as an Open Source user to do your part. Every one of us has a todo list that is growing faster than we can get the things on there done.
And Open Source not only has a different kinds of developers, it has a different kind of user, and really the line is blurry. You can't come to Open Source and expect it to work the same way that you're used to. You have to be a part of the solution here (or content).
How does this map to large companies? Well, they need to have people working with them in Open Source if they want it to do things *exactly* their way. It still will cost them a fraction of what it costs to support Windows.