Red Hat CEO Decries Open Source Pretenders 171
OSTalent writes "The Register has an article about Red Hat CEO Matthew Szulik's recent remarks...'For all his enthusiasm about the community and sever-side Linux, Szulik provided something of a reality check on the much debated theme of a Linux desktop. According to Szulik, the huge presence of legacy infrastructure like Microsoft's Exchange and PowerPoint has prevented a lot of people making the move.'" From the article: "It's very difficult to shape the development agenda of the community... every day people comment to us on the quality of our products through Kerrnel.org. What's important is staying true to the premise of the GPL model ... It starts with the APIs now, then it moves into content. Try to put [Microsoft's] Windows Media Player into Firefox and see what it looks like. In a world where application-to-application interaction becomes the norm, where does that innovation come from and who owns it?"
Bitching doesn't help, action does. (Score:5, Insightful)
Pick an open project for calendaring/mail and make Outlook work with it.
Create better tools for identity management.
The problem with people not embracing open source is not with open source, its that nobody knows what they're looking for with open source. Focus on what small business needs, and what open source can offer. Create small, turn-key packages. Create an LDAP authentication server. Create an LDAP mail server that operates as a drop-in replacement that works with the identity server. Create a Document Management System that works with OpenOffice, so that you have it part of the file-save dialog. Give business the tools it needs to work, and work efficiently!
The tools are better. Everyone keeps saying that they are. The design is sound, the pieces are there, but nobody has stepped up to the plate and sewn it all together. Stop the development of new tools. Take the tools that we have already and put them together. Industry needs more than Google and a Howto posted on an undergrads website.
Everybody knows that there are a million ways to authenticate a bunch of workstations to one or more server. LDAP, LDAP and Kerberos. GSSAPI, Radius, whatever, but for the love of all things sane and holy, pick one! Pick one, and build the turnkey solution to do it.
Re:Powerpoint?? (Score:4, Insightful)
http://www.wired.com/wired/archive/11.09/ppt2.htm
New Virtualization HW may be key (Score:3, Insightful)
Mark my words: the biggest threat to Microsoft is having the "either-or" argument disappear. (And I acknowledge that VMWare and others can do this today, but they 1) aren't free, 2) are already growing in use.)
Visio (Score:3, Insightful)
Re:Bitching doesn't help, action does. (Score:3, Insightful)
Novell is very pricey for the SME business. They are a large installation company, or a second step company. No manager that is used to windows is going to bet the farm on Novell right out of the gate. They need an open source package to try out first. Let them grow into a Novell.
As for your last statement, that's exactly what I'm saying. It can be done, but you need somebody to do it. a lot of these companies don't have tech guys, they have Bob in accounting that's pretty good with computers. If there was a distro out there that you could drop in, it was configured as a server, with file store, e-mail, calendar, document store, and it was reasonably easy to add new users, add e-mail addresses, and find and save documents easily, then business would snap it up.
Actually there is, and it's called Microsoft Small Business Server. Make it for Linux, and make it free. Then extend it.
Re:non-sequitur (Score:5, Insightful)
Redhat. Lets think about this.
Oh yah. They do large amounts of development work and stabilization for the 2.4 and 2.6 series kernel.
Hrm. I don't seem to be remembing anything abotu rewriting stack smashing protection and getting it actually incorporated into the GCC 4.x series.
Oh, and making SELinux usable. Na. That couldn't be Redhat. Could it?
Getting OO.org 2 ported to work with the gcj compiler instead of requiring a java runtime for many of it's features. THat couldn't be Redhat, eh?
Or how about GCC? Redhat couldn't be putting developers and resources into that project either, right?
And open sourcing GFS.. or netscape directory services, or developing and improving the ext3 file system.
Or how about Cygwin? I bet Gentoo did a lot of work on that one. Didn't they? That couldn't be Redhat could it?
I guess that doesn't amount to jack shit compared to your massive contributions to F/OSS software.
This couldn't be the company that allows projects like CentOS and Whitebox to download source code to their entire operating system and build 100% compatable clones either. Gee since they don't do that I would expect that Redhat would be big hypocrites.
Hey, how about this. Maybe Redhat has a business, and has employees and stockholders that they are responsable for. Hrm. Seems to me that each peice of software they buy or develope ends up being open source, isn't that funny for a company that doesn't give a shit?
Seems that they would behave more like original Suse did and rely on closed source management tools like Yast, or be like Gentoo, whose founder now works for Microsoft.
Give me a break. All Redhat does is:
1. Charge money for support
2. Protect their trademarks (which if you don't protect you loose unlike copyrights and patents. It takes a active effort to protect trademarks or they are invalid and anybody could use Redhat icons and call themselves redhat; including MS or IBM)
3. Don't provide binary downloads for free, except thru Fedora and Rawhide.
But they do provide the source code for everything they use... which is pretty open source, isn't it?
Mixing oil and water (Score:4, Insightful)
Free markets are about freedom. When people have it, they tend to use it to create wealth and prosperity where none ever existed before. Closed software is not about freedom, copy it and you can be sued or go to jail. Some people call that an "intellecutal property" right, but just because someone calls something a property right doesn't mean that it is.
True property rights don't derive from incentive, they derive from just allocation of things that have limited supply and demand. Just property rights lead to strong incentives, but coerced incentives do not lead to just property rights.
Re:Likewise for Visio (Score:2, Insightful)
I have used both, and I hope you are not serious.
By that same logic, I don't think Photoshop is worth the money either, since any image made with Photoshop can be made with MS Paint. It may require a teeny bit of manual labor (mmm, blend two images by averaging their RGB values), but hey, it's free! That's great!
Sometimes people forget time is money.
Re:PowerPoint (Score:5, Insightful)
1. Speakers use the PowerPoint as a substitute for actually knowing the topic; they just go over whatever it says on the screen, rather than being able to articulate the topic.
2. PowerPoint is a one-way communications mechanism; you can't readily make drastic changes to a PowerPoint presentation on-the-fly, the way that you can with a whiteboard. When I hold team meetings, I generally just write down the key points on a whiteboard, and as ideas get brought up, they get written down. Sure, it's low-tech, but it works a hell of a lot better than PowerPoint.
The smaller, the truer... (Score:1, Insightful)
Re:Powerpoint?? (Score:4, Insightful)
Really, though, the existence of weird and wonderful Excel applications is usually a bigger obstacle to a conversion than the need to display some PowerPoint slides.
It boils down to Gninertia (Score:3, Insightful)
It doesn't matter if it's a holy GnuWidget. People don't know (F)OSS from dog poop. They know Microsoft because that's what came on their machine. There are people that swear by Microsoft Works, perhaps the most awful 'office suite' ever written, because they finally figured out how to make it work. There's a lesson in that for the community.
FOSS has no marketing department, and will always battle those with budgets that can spread the word, or make it part of a bundle on a newbie's PC. Fight that, and you'll win, if winning is important.
Wrong tool for the job, not a bad tool! (Score:1, Insightful)
A spreadsheet isn't a statistical tool. You're misusing it and saying it sucks. For statistics use something like SPSS or such, not excel. It might be OK for small stuff, but I've never seen anyone even use it for that. Next thing you know, someone will be complaining about the 65 thousand row limit, and how he can't manage his data with it, just because he's never heard of databases...
What you're complaining about is like a carpenter saying his hammer isn't very good at dealing with screws.
linux has only been made for developers (Score:1, Insightful)
Well, it helps me to understand why usually companies which offers linux as preinstalled system, espect to be paid for preinstallation.
Everytime i look for some new stuff, especially cheap stuff, i am really frighten, it would take weeks, sometimes month before it could work.
What i could understand 5 years ago, as linux was growing, now start to bother me so much, i have a windows partition, to see if everything is working and how is it working with windows, before linux installation.
I personaly think, this kind of issues, will always stop common users to join the linux community. I have worked in the IT sector, and i am 100% sure, But common users will never be able to join the linux community, if this issue is not solved. Also what disapoints me, is that it hasn't been solved since the start of linux, which was in the 90's. So will it ever be solved ? and linux to stay a free developper product for developpers ?
Re:Err... AutoCAD? (Score:3, Insightful)
Think about it: most OSS programs either have a general appeal for a broad audience (internet programs, multimedia programs, desktops, office suites, etc.) or a specific appeal to the computer-technical audience (networking utilities, system administration, groupware, databases, etc.) AutoCAD doesn't fit into either category. It's not something a lot of casual users will play around with, nor is it something that a lot of programmers will have much use for either. The primary users of AutoCAD are design professionals: engineers, architects, etc, who don't need to know a thing about programming to use it, and in large part wouldn't have the time or expertise to begin to develop a replacement. Finally, if you can't win those primary users over with the replacement, the project simply dies.
I'm not saying it's a good thing that AutoCAD rules the roost; it's sitting in an Office-like position in its field, no question, and has been playing the changing-file-format game for a few years now. I just don't see a replacement on the horizon any time soon, if ever.
Re:legacy products (Score:3, Insightful)
My experience with Exchange has been that it falls over frequently. That was a long time ago now, and I imagine it's much improved, but instability was the hallmark of my Exchange experience for some time. The MTBF was probably about 3-5 days.
That said, I will never ever again deploy an Exchange server to one of my client sites. If a client insists, I will terminate the contract or require additional compensation due to the headache. It would never come to that though, because I would simply outsource the entire deployment to an Exchange service company (and it sounds to me like Dan Bercell works for one of these). Exchange is not trivial to manage, and requires careful setup and tuning. It has a reputation for this amongst IT professionals (at least in my circles). This is why Exchange service providers have so many clients: IT professionals in general don't want the headache.
There are many products that integrate with Exchange. It would be overgeneralization to say that they all suck; clearly, there is variance as with any group of software from diverse vendors. However as third-party products they can't afford to suck to the degree that Microsoft can afford to: MS users are locked-in (as I will elaborate below).
I disagree; MS has shown that it is quite happy to foist buggy products on its users repeatedly until eventually responding to outcry by beginning development of a better solution. Witness, for example, Windows 98SE and Windows ME. This isn't to say that this is what is going on; rather, it says that it's entirely possible. MS does not have to respond quickly to customer complaints, because its software is compatible with MS software and third-party products written especially for it... and nothing else. MS software is a tremendous pain in the ass to integrate with other products. If you're using Exchange, you'd better be running SQL server, and your clients had better be Outlook, and if you want webmail it had better be IIS, and you would be crazy to use a non-MS CRM solution - or you may find you have a tremendous headache on your hands. Of course, this all has to run on Windows. Want to swap out Exchange? Well, now you have to give up all the extra bells and whistles for all the other components, which depend on having MS products end-to-end. Outlook loses shared calendaring, you'll require additional third-party stuff to integrate IIS, it's probably nearly impossible to integrate well with Active Directory, and I don't even think CRM will work without exchange, etc etc.
MS locks its users in by keeping its integration in-house or to licensees. That's a big business for them, and they're doing a good job of selling access to their monopoly without giving up their lock on it.
This is also unsound. There may be another reason why corporations all over the world continue to use Outlook. I propose that that reason is that there are no alternatives, because only Outlook has the groupware and calendarin