Linked by Thom Holwerda on Fri 21st Mar 2008 21:49 UTC
Editorial "I used KDE as my primary desktop from 1996 through 2006, when I installed the GNOME version of Ubuntu and found that I liked it better than the KDE desktop I'd faced every morning for so many years. Last January, I got a new Dell Latitude D630 laptop and decided to install Kubuntu on it, but within a few weeks, I went back to GNOME. Does this mean GNOME is now a better desktop than KDE, or just that I have become so accustomed to GNOME that it's hard for me to give it up?"
Thread beginning with comment 306224
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[9]: From GNOME to KDE and back
by leos on Sun 23rd Mar 2008 00:17 UTC in reply to "RE[8]: From GNOME to KDE and back"
leos
Member since:
2005-09-21

I fail to see what any of that has to do with my post. That said, I think it is absurd to say that you have to compare a current release of one project with a two year old release of another project to get a fair comparison.


I completely agree. No release in several years does not excuse one from comparison to more recently updated software.

One should compare the current offerings. If one project has chosen to destabilize their release by doing the inadvisable (http://tinyurl.com/4gus), that comes with real consequences.


I think it's funny that you compare KDE 4 to the rewrite of Netscape/Mozilla. Sure it took a while, but Mozilla is now an undeniably successful browser engine. Impossible to tell if that would have happened if they had stuck with the old codebase.

If KDE's rewrite actually pays off in the future, another fair comparison can be made then. But be aware that the bar will be higher at that time.


Indeed. Of course, you are probably aware that there is a lot of talk in the Gnome camp about breaking API compatibility in a serious way with GTK3. You see, eventually the incremental approach just can't go on forever, no matter which toolkit you're writing for.

Reply Parent Score: 3

sbergman27 Member since:
2005-07-24

I think it's funny that you compare KDE 4 to the rewrite of Netscape/Mozilla. Sure it took a while, but Mozilla is now an undeniably successful browser engine.

Leos, I'm not sure how long you have been observing, but when Netscape released the source code in 1998, their browser had 70% market share and IE had a paltry 22.7%.

http://tinyurl.com/g6e7u

It was because of the *4.5 years* that Mozilla spent rewriting and getting that rewritten code base into a competitive state, during which time they basically had no product, that IE was allowed to completely overwhelm the market. Arguably, it was more like 6.5 years, because face it, Mozilla Suite was a nonstarter with respect to market share. It was not until Firefox that things even started to improve.

We've crawled back up out of the gutter over the last few years and are cheering the 15% or whatever that Firefox has gained back... after falling from 70%. And just how clean and high quality is the resulting FF code base, anyway?

So you see, the Mozilla story is an excellent example of why rewriting is a bad idea. Even today, 10 years later, we have only just begun to recover, and likely never will fully recover what we lost due to impulsive decisions made by overconfident developers in 1998-1999.

you are probably aware that there is a lot of talk in the Gnome camp about breaking API compatibility in a serious way with GTK3.

There is some talk about incrementally breaking the GTK API in certain ways, and in a controlled and orderly fashion. Getting from here to there is very likely to be an evolutionary process. API changes are, indeed, necessary for progress. But dramatic, sweeping changes, and complete rewrites are for the foolhardy and overkonfident.

Edited 2008-03-23 01:42 UTC

Reply Parent Score: 3

Manuma Member since:
2005-07-28

overkonfident

lol.

Reply Parent Score: 0

_txf_ Member since:
2008-03-17

Moving to Qt4 changed too much of the api...so a lot of the kde libraries would have to be adapted. At some point it makes more sense to rewrite the whole system.

Whilst I agree that kde4 is ambitious (so much so that 4.0 is pretty much a dud in my eyes). I think trying to transition would be even more time consuming as kde3 libs would have to be ported and then again further down the line in order for them to bring new features.

The advantage gnome has is that they have a lot more control over what happens with regards to gtk. with gtk3 the transition will be a lot slower and there will be less breakage but applications will still have to be ported or fixed.

Reply Parent Score: 3

elsewhere Member since:
2005-07-13

Leos, I'm not sure how long you have been observing, but when Netscape released the source code in 1998, their browser had 70% market share and IE had a paltry 22.7%.

http://tinyurl.com/g6e7u

It was because of the *4.5 years* that Mozilla spent rewriting and getting that rewritten code base into a competitive state, during which time they basically had no product, that IE was allowed to completely overwhelm the market. Arguably, it was more like 6.5 years, because face it, Mozilla Suite was a nonstarter with respect to market share. It was not until Firefox that things even started to improve.

We've crawled back up out of the gutter over the last few years and are cheering the 15% or whatever that Firefox has gained back... after falling from 70%. And just how clean and high quality is the resulting FF code base, anyway?

So you see, the Mozilla story is an excellent example of why rewriting is a bad idea. Even today, 10 years later, we have only just begun to recover, and likely never will fully recover what we lost due to impulsive decisions made by overconfident developers in 1998-1999.


In fairness, you're taking kind of a revisionist POV on the history there. IE 4.0 was groundbreaking, regardless of what people say. That's the point at which I quit Netscape, because believe it or not, IE was fast and sleek (funny how things change, though) whereas Communicator had become a big download. It absolutely smoked Communicator/Navigator by that point, which had become a bloated monster. Look at how rapidly Netscape lost usershare in such a short period of time, on that very graph you referred to.

Why was it a bloated monster? Because while MS was furiously re-engineering their browser with each version, Netscape just kept tacking more and more code on without really doing anything differently.

So you're trying to argue that Netscape "lost" because they chose to re-engineer, while I would argue they lost because they waited too long to re-engineer. Though at that time MS was pouring cash into development and literally hiring hundreds of developers simply to engineer a better browser, and that can be difficult to content with. But regardless, Netscape blew the browser wars. They were focused on building a business around their server software, since they were aiming at a future where everyone was deploying applications on their own proprietary web servers for use with their wildly popular web browser. That's the whole reason MS went to war with them in the first place, it had nothing to do with control of the internet, they simply didn't want Netscape providing an alternative platform for applications.

So point being that Netscape took their dominance in the browser area for granted, and focused their attention elsewhere. And Microsoft ate their lunch for them.

I think there's a lesson to be learned there, because when you assume that your dominance is assured, it's easy to become complacent with making the hard decision about when to take the necessary steps to truly innovate your product, at the short-term risk to disruption or inconvenience for your users, and simply assume you don't need to... and that's when your competitors will ultimately strike. Doesn't matter whether you're talking about FLOSS or commercial software.

So I agree your analogy is valid, I just don't agree with the interpretation... ;)

Reply Parent Score: 6

leos Member since:
2005-09-21

Leos, I'm not sure how long you have been observing, but when Netscape released the source code in 1998, their browser had 70% market share and IE had a paltry 22.7%.


Yes, and then Microsoft used their monopoly to push IE on the masses. If you have a closer look at the market share for browsers, you will see that by 1998 Netscape was already on a rapid decline. The slope is pretty much constant there. http://en.wikipedia.org/wiki/Image:Layout_engine_usage_share.svg

It was because of the *4.5 years* that Mozilla spent rewriting and getting that rewritten code base into a competitive state, during which time they basically had no product, that IE was allowed to completely overwhelm the market.


No. That is one theory, but to say that the rewrite lost them the browser war is a stretch at best. It possibly contributed, but you have absolutely no way of knowing how it would have went if they would have just continued to try and improve it incrementally. A browser is not exactly black magic, and microsoft can make a decent one when pushed to do so. Microsoft would still have used their monopoly to push IE on people, and if there had been more competition that IE would have been better.

There is some talk about incrementally breaking the GTK API in certain ways, and in a controlled and orderly fashion.


Haha. C'mon. An API break is an API break. There is nothing more or less orderly about what Gnome are planning than what KDE decided to do several years ago.

Getting from here to there is very likely to be an evolutionary process. API changes are, indeed, necessary for progress. But dramatic, sweeping changes, and complete rewrites are for the foolhardy and overkonfident.


Since when is KDE4 a rewrite? The only thing that was rewritten was kicker/kdesktop. And that decision was made by the guy that had been maintaining it for several years, so he probably know more about the state of it than you or I. Everything else was simply ported and had a few features added. Even dolphin existed in KDE3.

Reply Parent Score: 3