At first I was thinking to title this report, "Report: KDE at Comdex". You're thinking, well, that's what you did. Big deal. However, I then changed my mind and started to write a report entitled "Microsoft is Afraid". I think this is the most appropriate title. Then I realized that no-one would know what the real topic is, so I changed it back.
As many of you know, KDE was one of the winners in the O'Reilly contest to send six open source projects to Comdex. I went to Las Vegas to represent KDE, and brought my brother Matthew to help with the hordes of interested people.
The Comdex floor show opened on Monday, November 17. However, Bill Gates had his keynote speech on Sunday evening. We were able to attend in person, and it was an enlightening experience. Others have pointed out the behavior of Microsoft, and how they appear to follow the words of Gandhi: "First they ignore you. Then they laugh at you. Then they fight you. Then you win." Well Microsoft did follow just this pattern. Bill Gates and Steve Ballmer made a rather sad looking video of them taking on the roles of Morpheus and Neo in "The Matrix". The agents were Linux users, and their "kernel crashed" prompting them to "rewrite all of their drivers". Neo then had to make the choice of the little red Microsoft pill, or the "big blue, hard to swallow, IBM/Linux" pill. Ridiculing Linux may seem like a decent approach to them, but I think it shows how afraid they really are. They are preparing for a big fight (more to come on this). That being said, I think Bill Gates could use some public speaking lessons. He's not very charismatic. Perhaps I just expected too much from a billionaire philanthropist.
The KDE presence consisted of a small podium in the "Open Source Innovation Center", and was of course funded by O'Reilly. We had two laptops, a Dell Inspiron and an Apple Powerbook, both running random snapshots of KDE from CVS-HEAD. It was daring, but I don't recall encountering any problems in the course of demos. We occupied the podium for half of each day we were there, sharing it with other projects for the rest of the day.
Our demos mainly consisted of overviews of KDE, showing Konqueror, Kontact, KDevelop, Kopete, and other applications. We also had Kolab setup, but I don't recall anyone interested in seeing it in action. They were just interested that we even had such a thing, and then wanted to see the client. Despite all of this, the most popular features by far were Kiosk mode - for administrators/IT people - and Qt Designer - for developers. People were completely amazed with the power of Qt and Designer. It seems that many don't realize how powerful it can be. As for Kiosk mode, it is truly a "killer feature" for KDE. However, we need a GUI for this badly. We also need better, more centralized documentation.
It seems as though there is a list of important features that people need for accepting KDE (or even just Linux) on the desktop:
- Office Suite - OpenOffice is the only choice presently
- Groupware Client - Kontact looks promising, perhaps better than Evolution
- Browser - Mozilla is there, but I don't see anyone having problems using Konqueror instead
- Easy Updates - Right now we leave this up to the distro. Perhaps a bad idea, perhaps not
- Remote Administration - We're getting there
- Kiosk Operation - Here, KDE is king in infrastructure. We just need a UI
- Basic Accessibility, Usability - We're as usable as other desktops, but we need better accessibility
Among the visitors to the KDE booth were CIOs, CEOs, VPs and Presidents of
major companies and smaller businesses, students, hobbyists, journalists, and
professionals. I was stunned to see executives from Fortune 500 companies
coming by for a demo of KDE, saying that it was their favorite desktop and that they hope that we continue to do such a good job so they can adopt KDE for desktop deployments in the future. I was most, and least, surprised by one class of visitor though. We had regular visits from Microsoft
employees! They wanted demos of KDE, to see how it works and what we have.
What an interesting situation. I soon discovered that this was not the only
place that Microsoft people were doing investigations.
On Tuesday there was a Linux Desktop conference in the afternoon. I decided it would be worthwhile to go to that, and while I was stuck at the booth for quite a while and ended up being 15 minutes late, I still managed to catch a significant portion of it. KDE (on SUSE) was used as the demonstration desktop on the screen, and the panel was quite supportive of KDE. I think this is partially because of Xandros and Lindows using KDE as well. I did not see any trace of Lindows people around, although my understanding is that they should have been there.
While the talk was going on, I heard some furious typing behind my back. I
turned around and could see someone two rows back writing on one of those
tablet PCs that nobody buys. I turned around further and there was someone
beside him typing on a laptop - whew, I wasn't insane! I wondered who was so
interested in this conference, and I tried to see their badges. They were
hidden, but as the talk ended I waited around until they got up. Microsoft.
I am not sure if I got their names right but according to LinuxWorld they were General Manager of Microsoft TV Marketing, Alan Yates, and Pascal Stoltz, director of Microsoft's Information Worker Group - the group producing Office, Visio, FrontPage, etc. Seems like someone is really interested in Linux on the desktop!
Overall the experience was great for KDE. I think our presence alone was
enough to help introduce KDE to a crowd that really wasn't too familiar with
Linux at all, let alone a Linux based desktop. I would like to thank O'Reilly
for their help and for giving us this wonderful opportunity. I would also
like to thank John Taber and the Las Vegas LUG for their help at the booth.
Comments
tuxreports.com made an interview with George at COMDEX: http://tuxreports.com/article2513.html
Yeah, I've been thinking that we'd need that for a while. Suprisingly so few people know about it.
I second that, third it even! I am deploying LTSP boxes in our school district and am constantly struggling with sometimes contradictory and often incomplete documentation on configuring kiosk mode. Once I finally got it, it's not too bad, but let's just say the process is not Pointy-Haired-Boss-Friendly. When it comes time to demo the process (which is coming up pretty soon - several other districts nearby have expressed interest in our LTSP project), the kiosk configuration is going to be the hardest part to deal with. Bosses >like< the GUI. Heck, so do I sometimes - it sure beats parsing some of the longer menu files.
Maybe after I get these things deployed, I will have some time to learn QT Designer and just write the rotten thing myself. If someone is already started on this, let me know.
Yeah, a common complaint. I believe Waldo is working on it. Stay tuned!
Nice work, George! Thanks for your work in promoting KDE.
Find anything interesting from the other Open Source projects that were present?
Now I know what Plone is. :-)
Actually OOo had quite a big crowd there. It's a very popular project. I also got to see a preview of the new Gimp. I still don't think I have the artistic skill to use it, but it looks nice. :)
Well done :) I'm glad North America has someone like you to represent it.
It's very interesting how many execs still go to Comdex. But it's great that some, at least, are running KDE now and thinking about it in the future. I honestly would have never guessed.
But what doesn't surprise me is the lack of knowledge of QT Designer. Litterally noone knows how polished, professional, functional, supported, and actively maintained QT Designer really is :) -- some don't even know that it exists -- they immediately think 'less capable than what VS.Net can do with WinForms'. I firmly correct them of course by throwing together a quick little dialog with some controlls. The miriad of properties one can edit per control is awe inspiring to them :) ... then I remind them again that it will run on Windows, Linux, and Mac.
There's been some talk the past day or so on the mailing list for HOWTO's ... Since I'm not too confident in myself writing one up (and no time since it's my last semester in school) ... I'd definatly like to see some QT Designer tutorials and how to use KDE components within Designer as well. That would probably gain a bunch of attention.
Have a look here:
http://women.kde.org/articles/tutorials/kdevelop3/
This is a tutorial for beginners.
Great writeup, George! I admit I've never tried kiosk mode (maybe this will prompt me to give it a shot) -- it's basically a restricted desktop, right? What sort of GUI do you envision it requiring?
As far as competition from Microsoft goes, except for the handful of people paid to work on KDE, who cares? If they see an opportunity to make Windows better, that's great. It's not like a little more or less competition for Linux makes my life any different.
What did the Microsoft people you talked to have to say? I'm curious what strengths and weaknesses they perceived.
It's always fun to meet some great programmers (George and Matt) and the chance to see KDE 3.2 running - Can't wait to use Kontact in my company. And KDevelop looked so impressive, I'm installing the CVS version on several computers to use asap. Keep up the great work, Team KDE. It was really great of Tim O'Reilly and company to sponsor the open source area - it was one of the busiest booth areas at the show. And at the small stage area, it was amazing to see people grabbing seats 30-60 minutes early to see the open source presentations with lots of people standing to watch.
It's ALL GREAT !!!
wow ! Happy Am I.
just to add my two cents, about "Easy Updates"...
I really think is good we let distribution take care of most of the work.
but KDE should have a package center able to play with KDE package and their require dependencies. So that we can't update and uninstall SINGLE app, not only entire package.
and we should not only count on the rpm/deb database, but also create our own database, so that every installed files are linked to app/package, and the we can know if we have useless .so file for example. or *ui.rc
It depends on the distribution if you can install single apps or not. At least you can do this in Debian.
I am definitely pleased with the direction KDE is going and that quality is the number one focus?
About the Kiosk mode however, isn't the Kiosk mode supposed to have a GUI in KDE 3.2?
Well I don't use the big-monster Mozilla, nut I do use Firebird on both Linux and Windows. I have to say that I can not use Konqueror on a few sites that I visit, because of some very incorrect rendering (grey boxes filling up most of the page yet the links are still clickable). However, Firebird works very well with the same sites. I would gladly report these sites and what the problems are, but (at least in the past) the bug-reporting guidelines stated the developers do not want a URL for a webpage that is not rendering correctly but want the specific HTML-code that is causing the problem. Well I simply do not have the time to narrow that down myself, nor do I have the time to debug khtml myself. If these policies have changed, I will gladly send in URL or a copy of the page that is causing the problems.
At this time, the sites that Konqueror can not render are ones I visit more than once a day. So I basically stopped using Konqueror. I am using the Konqueror from CVS-HEAD (cvs-head binaries for debian sid from opendoorsoftware.org)
The policy hasn't changed, and isn't going to change. It is a good policy. It is much better for the users of KDE betas to manage the bug reports and present them in a nice way to the developers, than to have the bugs.kde.org become simply a log of random user complaints and let the developers sort them out. Developers want to develop, not wrangle bugs on Bugzilla all day. Since there are many more KHTML users than developers, it works out very well: the developers can focus on what they're good at, and the much larger number of beta users can be the testers and bug-finders. If you don't have time to produce nice bug reports, that's perfectly understandable, but cluttering up Bugzilla with bad bug reports that you didn't spend any time on won't help.
Ok well I dont know if you mean it or not, but you kind of come across sounding like a jerk -- sorry but thats how you come across. I just got done saying how I'm fully aware of what the policy was/is and that I didn't submit the URL because of that. I understand that they might not want to track down the bug only from a URL or even from the HTML for the whole page, but again this will just be a hinderance. Perhaps as one other person suggested, the kde developers could put together a special section for KHTML problems where you can submit a URL and maybe it will ask for a (screen|window)shot a description and then it can download the page. That shouldn't be too hard to add, though I dont know the internals of bugzilla. Then later maybe some grateful souls could filter through these and try to narrow down the code to submit to the khtml developers. It maybe some work, but it can only help KDE in the end.
The sites that have the problems that I visit, are high-traffic sites and are used more by "regular" people than probably by the people that visit here and /.
The sites maybe even doing something that violates an html-spec, but since they render correctly in IE and firebird, it should be able to render correctly with khtml.
I wonder, is there anywhere where one can submit such sites so that others can check them? I'd be willing to look at a URL or two if there were such a place, and I'm sure I'm not the only one.
One could start a separate Wiki page at http://kde.ground.cz/ for that purpose. For others to be able to pick up an URL and making a testcase out of it additional information like used KDE version and a comparisation of expected and actual result need to be given as well though.
The problem is that there are way too few __khtml__ developers to be able to write testcases for every site that gets reported.
You can submit sites without testcases, any many people do that, but the outcome might be that it stays in UNCONFIRMED state for a year =).. I've been trying to write testcases for bugs w/o testcases, and there are other various people who do the same, but there are still probably more than 500 khtml bugs without testcases.
Hell, even Apple, who has a team of full time employees working on Safari is bogged down with bugs atm (see Hyatt's recent posts about this at http://weblogs.mozillazine.org/hyatt/ )
How about the solution Apple uses?
IIRC, the have a button on safari that allows the user
to submit a bug report for a page not rendered properly,
ALONG with the actual page (or url).
So, it's an instant testcase, isn't it?
You want to look up the meaning of "testcase".
After all everyone in the world I know installs it on OS/X NT XP Win98 and every other OS I know of ... because on those OSes it beats the crap out of the "native" browser.
"I was stunned to see executives from Fortune 500 companies coming by for a demo of KDE, saying that it was their favorite desktop and that they hope that we continue to do such a good job so they can adopt KDE for desktop deployments in the future."
Well, I suppose that's good news. Maybe sooner or later they'll even consider helping actively (e.g. by paying a couple of developers to improve weaker areas and deploy it within their company) :-)
Right now, I'm a bit disappointed to see that apart from very few "small" companies (Linux distros like Mandrake and Suse, Eric Laffoon, Trolltech and maybe TheKompany), there is not so much financial support going to KDE... while in the meantime Gnome gets much more. Support is not in proportion to popularity, which is normally not the way things work in the OpenSource world.
The problem with that is that China for instance will get Gnome-based desktops from Sun (see http://slashdot.org/article.pl?sid=03/11/18/0219249&mode=thread&tid=102&...) when they could (should) have had KDE.
What happens next with the thousands of Chinese programmers that will use this ? They will develop Gnome apps, not KDE :-(
Too bad !
what about the indian people ? the are now in the computer age , and half many people as in china , so what are they developing for ? linux/kde/gnome ?
please name me an indian opensource software project.
chris
Anjuta
Oops that a Gnome project ;)
But it not as good as Kdevelop
GNOME is already pretty dominant in India atm. Mostly lacking to no Indic language support in Qt until 3.2, and thus the fact that there aren't any translations of KDE apps in the various Indian languages. GNOME otoh, has pretty good translations in at least Hindi, and at least partially in Bengali.
Some of us are working on filling in gaps in this area. Infrastructure first, translations next.
> Right now, I'm a bit disappointed to see that apart from very few "small" companies (Linux distros like Mandrake and Suse, Eric Laffoon, Trolltech and maybe TheKompany), there is not so much financial support going to KDE...
Heh heh... It's still so weird to be reading and have my name pop up. It's like seeing a black helicopter outside my window. ;-)
Seriously, I would bet there are a lot of users that made more than I did in my little start up company these last few years. I think 2004 will finally be different. As an individual you direct your resources toward your passion and I'm a very passionate person when I have a cause. Recently I've seen FOSS in the perspective of a power struggle between naked greed for wealth and power for the sake of wealth and power and the promotion of freedom and opportunity. Today it is not of much consequence, but projected out 50 years it is choosing between the best and worst that can happen to civilization. I've never had an alternative to supporting open source that I could live with.
From the corporate standpoint I think it's different. It's all about keeping the stockholders happy and not being the first lemming over the cliff... it's okay to run with the rest of them. Fortune 500 companies need to be shown a successful chain from start to finish, supplemental questions and concerns need to be answered and then someone who speaks their language needs to have a bulletproof presentation and beat them over the head with it repeatedly until it sinks in that it is completely within their financial interests to sponsor FOSS and that successful mechanisms and organizations exist to demonstrate the benefit... in fact that's part of the whole thing.
For large companies it's the coming paradigm shift that is hugely different in how they view things... or it is missed for something that looks more familiar but has similar failings. The reality is that while it is simple in concept it is not so simple in practice and, like the ultimate desktop or the next killer app, it will require some patience to develop. What should be kept in mind is that errors in judgement here can end up being costly to the community. For this reason I don't see any reason to lament as to who is ahead of who here. An infusion of finance in FOSS can destroy as well as create if it's not done right. ;-)
SUN is boasting!
In fact, the chinse company which is cooperating with SUN is very weak in Linux, especially in desktop market. RedFlag, which is the largest chinese linux company, use only KDE as its desktop environment. According to IDC report of 2002, RedFlag shipped more than 1 million copy of its desktop version and occupy 30% of the world linux desktop market, and the growth rate is 270%! This shipment number makes RedFlag become the 1th linux desktop vendor in the world. In china linux desktop market, RedFlag occupies about 70% share. The company cooperating with SUN has only 3%.
So don't worry about that.
Come on... Don't you think you're being a tad bit self-indulgent here ? That they consider Linux to be competition hardly means they are "afraid", or even mildly worried. About the server market, may be, but on the desktop they are pretty safe.
They're worried about the future, not the present. They see what happened on the server, and what is possible on the desktop if we work hard enough. Come on, think about how rediculous it is to have wealthy, high power software industry leaders sitting in a room taking notes while hobbyists talk about the free software they're writing to replace the industry-provided solution. I also clearly remember the infamous letter by Bill. Imagine this, the hobbyists are now actually a real threat to him!
(yes Linux is not just for hobbyists anymore, but come on, a huge chunk of it is still developed by them)
MS is always worried about the future, that's how they got where they are :-).
They also know that writing for the desktop requires much more ressources than for the server. Let's be realistic, it's been 7 years since KDE was started and we barely have the feature level of a standard Win95 install (with Outlook and Office), let alone a comparable range of end-user applications. Yes, we're more stable, but experience shows that this comes way after features for end users on the importance scale.
To be a real threat, it would take IBM hiring the top 50 KDE contributors, then a couple hundred more coders, managers, doc writers, translators, UI experts, graphic designers etc..., relocate them all in the same building and wait a couple of years.
"we barely have the feature level of a standard Win95 install"
Troll? After using all three systems (Win95, KDE, WinXP) considerably, I would feel inclined to say that the last post could only be a troll (I guess I'm feeding it).
KDE has as many features over Win95 as XP does. Some of those features are in opposite directions, but there are as many. The only thing I've found XP good for over Win95 is better networking (out of the box with KDE), nice device integration (pretty good with kio_slave and linux distribution improvements - not as good as XP - but still not bad), and fancy, glitzy menu options for doing things like slideshows (which are actually buggy and are "magic" in that there is no way to control when they turn on or off and most likely violate all sorts of anti-trust settlements for incorporating too much into the monopoly OS). Oh - and XP crashes less than 95 - but I can still crash it.
I don't use Outlook (OK I do at work when I forget my Linux Laptop with Kmail and IMAP support). I don't use the "new" offices (except for a few tweaks there is nothing improved over the 6 year old Office 97).
Taking all of that into account and adding in the freedom and openness that I get in using KDE, there really isn't any competition in my mind.
The only reason I still dual boot is because of external applications that run only on WinXP (that is where the only innovation in the windows world has really come from).
As a final blow - my 4 year old son prefers booting into Linux and the game selection he has there, even though we have a dozen proprietary kids games under windows.
> Troll?
Search my name through Google. You really think I'm a troll ?
As for the rest of your comment, well, it's the standard "why I think Linux is better" stance. Basically we take our own case and postulate that everyone would think the same if they "knew what we knew", or "used what we use". The reality is that we use Linux because we *like it*, and we avoid Windows because we *don't like it*. There is very little rational thinking in this decision.
As you say there is no competition in *your* mind. Try that on someone to whom a computer is just a tool to get the job done and not a hobby.
As for the rest of your comment, well, it's the standard "why I think Linux is better"
Isn't that the way it should be - with some evidence? Your statement that Win95 is more featureful than KDE is complete assumption. You seem to be simply making a blanket statement of Win95 being better than KDE or anything else. It is simply "why I think Windows is better."
"Try that on someone to whom a computer is just a tool to get the job done and not a hobby."
I already have. A KDE based, open source desktop wins hands down because people can see what they actually need. I've seen people use Windows and apart from latching onto a few applications they use or have heard about they haven't got a clue how to use it. Having been in business and supporting people, I have seen first-hand how little of Windows people actually use and how confusing it is for them. I thought because everyone knew Windows, computer use for most people wasn't so bad. I never realised how bad it was before - I do now.
> Your statement that Win95 is more featureful than KDE is complete assumption.
I meant "typical Win95 install", i.e. with Office and Outlook. Office 95 can be used in a workplace, KOffice unfortunately can't. Outlook's level of integration was what we're trying now to get with Kontact or Kolab.
> I already have.
Have you made, say, a small business move from Windows to KDE ? Without using Open Office ? Many people here would be very glad to hear about your experience, myself included.
I've also been on the support side and witnessed how little people actually
make of their Windows computers. In fact, one lady (wealthy retired executive) actually told me that she'd pay more for a small, instant-on, energy efficient computer that'll give her the ability to simply access e-mail and write a few letters here and there. The computer she had at the time was a custom-built Pentium 4 machine with 3 fans, 1 7200 rpm hard disk and a CD burner. I shudder to think that she actually turned that thing on just to check e-mail. People don't really like having to use a horrible Windows machine at work and then go home to do the same thing. That'd be like driving a tractor trailer at work and using the same thing to go pick up groceries. I know it's a stretch but it gets the point across. The nice thing about cars is that they most of them have 4 driving wheels, a body and a steering wheel, but manufacturers are allowed to build a seemingly infinite variety of machines based on that, all instantly recognizable by the general public. The software world needs such basic standards so that people can use they like while still being able to get something accomplished.
> Let's be realistic, it's been 7 years since KDE was started and we barely have the feature level of a standard Win95 install (with Outlook and Office), let alone a comparable range of end-user applications.
Nice troll. If I spend an hour installing windows, at the end I just have windows. If I spend and hour installing SuSE Linux, at the end I have hundreds of applications (including two office suites), plus my choice several email clients all of which are superior to Outlook in mail handling and security. Anything of the MS world that I might need already has a superior Open-Source alternative, and everything else runs fine under CrossOver Office (including my favorite Windows application devel environment: Progress Software Corp's fantasticly wonderful Provision). ;-)
> Nice troll.
FYI Guillaume Laurent is one of the most knowledgeable developers on KDE. Even more interesting he came by way of developing on Gnome's C++ bindings and is one of the strongest advocates and apologists for C++ as opposed to C in Linux application programming. He's also one of the core developers of Rosegarden, a program to complete with professional level music sequencers.
Just because his assesment of KDE's competative posture to Windows is somewhat critical is no reason to call him a troll. From what George said about fortune 500 execs watching development and a current review of desktop penetration KDE has not broke into a clear lead in features. Given how much less time KDE has had it would be astonishing if it was that far in front. He also said that most users don't prioritize stability like we do. I am inclined to think his estimate for what it would take to get ahead is exaggerated, but I'm not inclined to argue with him because I hate losing arguments. ;-)
hrrmmm, um ok, so maybe no troll. Still I think Mr. Laurent should be more encouraged by the state of KDE. And I think he would lose the argument about Win95 feature set vs. Linux/KDE. FWIW, a non-geek windows-usin' friend of mine switched to SuSE 8.2 On her heavily used home PC. Result: She loves KDE! She would use it at work if NASA would allow it. (Ironic that). The success of non-geeks using linux/kde is how we should measure the success of KDE. From this little corner of the front, it appears that well-groomed linux distros ARE ready for the desktop. Rosegarden looks fantastic.
> I think Mr. Laurent should be more encouraged by the state of KDE
But I am. That I don't think MS is worried doesn't mean I don't like KDE. I think OSS is a vital ressource to humanity, there has to be a counter-power to MS. But the old "we got MS on the run" tune is still complacency and self-delusion IMHO.
For the record: I'm not defending calling him a troll.
He is, however, grossly inaccurate in his assessment of KDE. Certainly, the core technology of KDE is way ahead of Win 95, and in the important places, comparable to Windows XP.
- Win95 had nothing comparable to DCOP.
- OLE/COM in Win95 wasn't used as prevasively as KParts is in KDE.
- Win95 had nothing comparable to Konqueror.
- Win95 had nothing like the KDE multimedia framework.
- Windows didn't have a (standard) clean OO API until .NET came out.
- Windows still doesn't have anything comparable to KIO.
- Windows still doesn't have unified toolbar/menu customization.
- Win32 still lacks layout management.
And much much more!
Maybe in terms of UI polish, we're back near Win95, but its certainly an exaggeration to say that KDE as a whole is comparable to Win95.
You're talking about developper features (except for Konq). I'm talking about end-user features. That KDE is a better development platform than Win95 is a given, no argument on that.
"I'm talking about end-user features."
And exactly this area is very blurry and subjective at best.
Your comparison was ill concieved. The user-visible differences between Win ME and Win 2k were not great, but the difference in underlying technology was huge. That core technology takes a long time to build, and good technology is ultimately expressed in the resulting applications. Even today, you can see it in the tight integration of KDE apps (via DCOP), the excellent component-architecture of KOffice (via KParts), and the location transparency of all KDE apps (via KIO).
To say that KDE is at the level of Win 95 implies that KDE is 8 years away from catching up to Windows XP. Because the underlying technology is present, however, that's a very misleading impression.
> To say that KDE is at the level of Win 95 implies that KDE is 8 years away from catching up to Windows XP. Because the underlying technology is present, however, that's a very misleading impression.
I disagree. I think KOffice is clearly years away from Office XP, simply because so few people are actually working on it. We always think that since the underlying tech is here, we're 80% done, because that the hardest part. In fact it's just the opposite. The underlying tech is not the hard part, is the "sexy" part we all like to work on. The apps are the real hard part.
Then finally stop writing about KDE and MS Windows when you actually mean KOffice and MS Office, simple as that.
As for apps being the hard part and the underlying backend being the easy part I consider you either being nut, kidding or simply unaware of the situation of many products and projects where they cared a s**t about the underlying backend in the past and are now forced to "unify" their single apps across several apps by dublicating every single UI feature and behavior. You tell them please they shouldn't care, the underlying backend is easy to do after all.
> As for apps being the hard part and the underlying backend being the easy part I consider you either being nut, kidding or simply unaware
Do you have direct exprience of what you're talking about, or are you just referring to hearsay and mailing-list browsing ? Please explain me further how writing code for the end-user is easier than writing code for programmers. My company is moving from writing libraries to full-fledged end-user applications. We would love to hear how "easy" it is, because at the moment I assure you it's not, and quite frankly nobody is the least bit surprised by that.
Ask around, what do KDE core devels think was the harder to do : DCOP and KXMLGUI, or Konqueror and KWord ?
BTW yes, retrofitting a new infrastructure into an existing GUI is hard, of course. That's not what I'm talking about.
I used to think that Windows 98 had a lot of stuff. After going back to W98Land for a visit, it looks very barren....
Only used WinXP for a little and saw little to urge me to go back.
KDE is the best!
MarkP
> Guillaume Laurent is one of the most knowledgeable developers on KDE.
Thanks, though I think you're giving me way too much credit here :-).
> Rosegarden, a program to complete with professional level music sequencers.
Let's say we try to offer a cheaper alternative for hobbyists. We can't realistically be pitched against Cubase and the likes. The embryonnic state of the Linux audio framework doesn't help either. Once everybody runs 2.6 things should get quite better though.
> I am inclined to think his estimate for what it would take to get ahead is exaggerated
3.5 years of Rosegarden development and my professional exprience have taught me one thing : the Devil is in the details. That's where you spend 80% of dev time. Usually we stop at the 20% because, as geeks, we're happy with even alpha versions as long as they're stable enough. :-)