What do other, more experienced readers think about the problems and improvements related to dropping the current Linux approach for a 'bundle-like' one in Linux distributions?
OK.. this question is really 1st year CS material, so hopefully this will set all y'all newbie young'ns straight. "Bundling applications," as defined as giving every app it's own copies of used libraries, is just plain stupid if at all avoidable. Here's why:
1.) What happens when a bug or security flaw is found in a library? Without a shared copy, you must figure out which apps are using it (which may be thousands) and then upgrade every application "bundle" instead of one library for the whole system. And what if some apps are using an older version of the library which nobody bothered to patch?
2.) Disk caching. Today's hard disks may be really large, but they're still really slow (compared to the rest of the system). If you have to load separate copies of a library for each app, you lose all the benefit of disk caching.
3.) Memory usage. Shared libraries allow a single copy of the library in memory to be used by multiple applications. This also reduces load time if the library is already in memory. (ie. this is why it makes sense efficiency-wise to use either KDE or GNOME and not a mixture of apps from both) It's also partly why OpenOffice and Firefox take so long to load on Windows compared to Office and IE. (they don't use all the standard windows libraries.)
4.) Shared libraries are a major driving force in pushing application developers to stay on their toes and keep up with the progress of the library developers.
5.) You shouldn't be compiling your own apps unless you're their developer or have very specific security or optimization needs. It's a waste of time unless you're learning something in the process. Leave that job to distro package maintainers and do something useful with your time like becoming a better programmer and/or contributing to your favorite app. Once Linux ceases to be a toy for you, you'll avoid compiling everyday software like the plague.
I could go on for several points, but that should be enough to convince ya. (:
Over the next 10-20 years, monolithic groupware like Outlook will disappear and it will be replaced by a dozen separate servers and applications.
Just a clarification: although there may be a dozen separate servers and applications, they will appear as one integrated application to the end user. I'm talking about proper, natural integration here, not crude cobbling. Users will no longer have to worry about import/export of data between applications because it will all be one unified workspace. Under the hood, all the software will talk using open protocols, making it easy to extend or customize the functionality of said workspace for particular needs. This is the future of business software: Perfect modularity that is perfectly invisible.
Incidentally, this is also going to make the desktop OS almost completely irrelevent for the majority of cases..
Give me another solution that offers seamless integration of proccesses.
I see where you're coming from.. It can be done today by integrating dozens of Open Source tools, but it requires way too much geek elbow grease for the typical organization. So, listen up any of you enterprising geeks out there! This is what the business world of IT wants: seamless integration.. solutions.. simplicity.. Make it happen with OSS and you'll have a goldmine on your hands.
In contrast, transitioning to Java usually takes greater skill in object orientation.
I count that as a benefit of Java, hands down. Java itself and the surrounding tools lend themselves towards proper object oriented design. Honestly, a programmer who isn't already OO savvy is not who you want to develop your enterprise software. Send them back to school or find somebody else. This is the 21st century folks. Unless you're doing low-level coding, you should be using proper OO design.
Really, though, these technologies are so similar that you should probably just pick one and go with it.
Sure, you can knock out a solution with either platform. (or even PHP/Perl if you're a masochist) But why lock yourself into an expensive, proprietary platform?.NET doesn't offer you anything that Java is missing. If you're not familiar with either platform yet, pick Java and don't look back.
If you want the easiest and most productive development environment, stick with.NET.
This is rather a myth unless your team is a bunch of MS junkies who've never used anything but Visual Studio. Java development tools, overall, are more powerful than those for.NET today. And the Java development tools are mature and time-tested.. For those getting in on the ground floor of enterprise software development, just choose Java and don't look back. There's no compelling reason to choose.NET if you're just starting out. All the problems that have plagued Java in the past (early 90's, dot-bomb era, etc.) have been solved.
But do note: You should be largely looking to the Open Source community for your Java tools, not Sun. Sun is driven by committee. OSS is driven by what works in the real world. Look to Spring, Hibernate, Struts, and the rest of the Apache Jakarta project. Forget about heavyweight J2EE app servers unless you really need them. (And you can always refactor later)
Then I tried it on Windows and I can certainly see a problem. In most any Linux windowmanager, you can force windows to stay on top (and shade them). In Windows, you lose those two options and that's when the interface really becomes a nightmare to use.
This is why Gimp needs to have a *user option* so that both groups are satisfied.. Even as a seasoned Gimp user, there are a handful of cases where I would prefer an MDI interface for certain types of tasks. Can this really be so hard to accomplish? It's almost a silly political issue from what I can tell. I mean.. screw it.. Give the users what they want if it will help the project grow faster. Just make it optional so that people who use an X11 window manager and know what they're talking about can still use it the Right Way. (-:
It only sucks for those who use photoshop extensively before switching to the GIMP. I remained free of Photoshop long enough that now I am completely used to the GIMP's interface, and I don't see why it is any better/worse than Photoshop's.
I would generally agree with you as this was my experience as well. On the other hand, not everyone's brains works the way ours do. It is quite possible that some people do legitimately have more trouble learning Gimp than Photoshop. As a result, I think the Gimp project needs to make MDI vs. multi-window an *option* for the user. Some people just refuse to change no matter what. Maybe there's a good reason, maybe there's not. But ignoring the issue is not the right approach. I mean.. come to think of it, there are times when I would prefer to be able to switch Gimp into some sort of MDI mode for certain tasks. Can this really be that hard to accomplish?
As I see it, we shouldn't try to convert the professional full-time users of Photoshop, but rather the people who pirate it.
We should also try to convert the people who casually use Photoshop but don't even know how to use the advanced features that pro users demand. You wouldn't believe how many wasteful legit copies of Photoshop there are in the corporate world.. in the hands of complete novices who just need to touch up photos for a website or prepare graphics for Powerpoint slides. It's pathetic.
Piracy is a bigger threat to Free Software than it is to entrenched industry standard software, IMO.
Indeed.. I can't tell you how many Windows and Mac junkies I've seen who refuse to switch to Linux because, "I want to have my Photoshop, Premiere, SoundForge, Cubase, Dreamweaver, etc." And, of course, they could never afford to license all those programs. But to them, it's just as free as Linux with Gimp, Cinelerra, Ardour, Rosegarden, Quanta, etc. These people are pretty much hopeless to convert (although a really anal DRM scheme in Windows itself might actually help us.. heh) The only way these people will be reached is by the trickle down effect. If we can capture the corporate market, where licensing is enforced, the Free alternatives will become the big popular names and the same people who currently warez Windows/Mac crap will simply use the Free software instead "because that's what everybody is using."
Incidentally, it may also be said that P2P sharing of major label music only really hurts the small, independent artists who are often willing to freely share their music online. The original promise of P2P was that it would eventually be a way for indie artists to get their music out there with zero distribution cost. However, by the time P2P was popular with the masses, most users were just a bunch of lamers who wanted free copies of the (major label) songs they heard on the radio. And the music cartel continues strong as ever..
Re:"GIMP is also not meant to be like Photoshop"
on
Paint.NET: The Anti-GIMP?
·
· Score: 4, Insightful
I'd say that Gimp on Linux is for 90% of Photoshop users (I say users including all the random kiddies who downloaded it because they want to be "pros") an absolutely solid replacement. I know that in all the years I've used it for commercial web design, photo manipulation and UI development it has never yet left me wanting.
Exactly.. This is what most people seem to ignore.. Gimp is not Photoshop. But it does happen to meet the needs of probably 80-90% of Photoshop's target market. You would not believe how many wasteful copies of Photoshop are licensed in the corporate world because Joe Idiot says, "Hey, we need a photo editor.. go buy the best thing out there." And $699 later, there's the latest version of Photoshop. (And oops.. it's hard for newbies to use (just like Gimp), so go buy a copy of Photoshop for Dummies too) Sure, Photoshop is still the best thing out there (today at least), but most of the people using it would have been fine with PSP or Gimp.
And here's the real kicker: how fast would Gimp improve if those 80-90% that don't really need Photoshop contributed a few bucks each to the project? Granted that won't happen, but there are other ways to harvest this market. The Gimp folks need to take a look at how they can capitalize on what they've developed.
MS is sponsoring not just free software but Free software, Very interesting!
Microsoft has always been pro-BSD licensed software because it lets them take and not give back. They hate GPL because they can't use it to be greedy and monopolize. For example, MS could create their own branch of this Paint.NET with a handful of extra features, not release any source, and include it with Windows.
Oh and can we take this and shove it on Linux?
It uses proprietary MS-only libraries like GDI+, so the general answer is no.. unless somebody writes a.NET GDI+ compatibility layer for Mono. It would probably be a waste of time to do such a porting effort when the time could be better spent improving Gimp or KOffice's Krita.
I think you are mistaken. Linux does well right now as a server, but it is not very common as a desktop. Obviously, the numbers are hard/impossible to gather, but from personal experience I can tell you I only know one person who is not a programmer that runs Linux as their workstation OS.
It's true that not very many non-technical users, as a personal choice, use Linux on their desktops. However, there is a rapidly increasing use of Linux on desktops where there is support by IT staff. The "numbers" don't really show this kind of usage because it often isn't backed by sales figures coming out of RedHat, Novell, etc.
Neither Mac version is anywhere near usable, so Mac Users usually just buy MS Office.
I wonder what will happen first.. stable OO.org for Mac or KOffice (or other) taking over instead. Thing is, StarDivision's StarOffice was originally developed for Windows and Linux only. It was a very sloppy codebase and the results are still obvious today. I liken it to early Mozilla milestone releases when we were dealing with the messy Netscape code. The cleanup work on OO.org will eventually make the project very portable, but for now, it still has a long ways to go. Somebody really needs to start a serious fundraising effort for OO.org in general. There are so many people and organizations who would save big bucks if they didn't need MS Office any longer.
The number one complaint I've heard about open source games is that while the engines are often cool, the graphics are always crap.
Agreed.. it's an obvious problem. Most OSS games, for example, have absolutely awful 3D textures and it completely ruins the appearance, even where the game engine is doing a good job with geometry and lighting.
I know 4 professionals (including as I said the best graphic artist I know) who all tried to work with multiple OS projects and all gave up in frustration because no one would even talk to them. This is symptom of a problem.
That's a real shame. Have you thought about teaming up and taking a stand? Nothing will change if people aren't made aware of the problem.. If there are comments / flames that were posted to a public forum, bring them to light. "opengamegraphics.org" anyone?
One of the reasons OS X is more popular on the desktop than Linux is that no one in Mac OS X forums ever tells you to learn how to program or RTFM.
There's no such thing as just "the desktop." I wish people would stop being so generic. On the corporate and education desktop, Linux is now more popular. On the home and graphics professional desktop, Mac is more popular. Overall, Linux is rapidly pulling ahead and it's because of the fact that, using the right distribution, newbies don't have to program or Read Fine Manuals anymore. At the same time, sysadmins deploying it have a platform that's easier to maintain (on a large scale) than either Windows or OSX. When was the last time you saw a network booted Mac?
It is much cheaper and easier to just buy MS office (which is what most people do).
Are you talking about Linux or OpenOffice? You seem to have switched topics, because there's no programming or manual reading required to make OpenOffice work on its supported platforms. In fact, it's faster to get working than MS Office in Windows. (no license key, no product registration, no rebooting, no service packs, etc.)
The hostile and elitist attitudes of many open source zealots are really hurting the community.
There are a handful of jerks out there -- like any other real-world community. There are far more people who have not had the kind of experience you mention, so please don't spread bad rumors about what contributing to OSS projects is like. And remember: more than just the source code is open. If you run into a jerk, you don't have to let that stop your own participation. Most likely other people in the same project are not jerks and may even talk to the jerk in private, telling them to apologize and lighten up. I've seen it happen. If your friend had set up his own site with alternative graphics, they would have eventually been integrated regardless of the wishes of the jerk. Rule by the masses means jerks don't ultimately hold much sway. Whoever gets the job done right becomes accepted into the community.
your attitude that non-programers have no right to make comments is just the sort of attitude that pushes talent away.
I agree that many OSS projects need to adopt formal means for non-programmers to help out -- perhaps even separate from the programmers. KDE, incidentally, does a great job of this. Usually the harsh attitude of programmers towards non-programmer upon receiving comments is because they are already overloaded. And I say this with personal experience as a developer who receives a lot of unprofessional offers of help or advice. Casual, non-technical help and advice is far more difficult to integrate into the process. Good projects should have liaisons to distill non-technical help into a form directly usable by the core developers.
That being said, there is something to say for people who have complaints contributing financially. Money has a way of getting people's ears very quickly.
Are not sys admins (or capable of) and as such cannot, and should not, use linux
Linux is just a kernel. There are some software distributions that utilize Linux which are difficult for non-sysadmins to use. There are also some which are nearly idiot proof today. (moreso than Windows, almost as good as OSX) Don't make silly generalizations about who should and should not use Linux.
At least if an opensource product becomes dominant, there will still be some incentive for new features, since oss software is written by and for users, those users who desire new features and are capable of writing them, are free to do so..
The difference with Open Source is that dominance consistently leads to increased innovation rather than stagnation. Anyone who actually thinks OSS is not about innovation is a total idiot or has been living in a cave the last 5 years. Unfortunately, it's not always polished innovation, so average people tend not to notice until something like Firefox all of a sudden rises on the scene.
In an ideal world, opensource apps will commoditize the current feature sets and commercial vendors will have no alternative but to offer new and innovative feature in order to make sales.
In an ideal world, commercial developers will all be working alongside successful open source projects to meet the needs of their own customers. And meeting those needs will often occur in new and innovative ways which everyone will collaboratively benefit from. Reduced wheel re-inventing will shrink the software industry, but this is the inevitable progress of technology anyhow.
Rather than open source putting MS out of business, it seems far more likely to me that someone or something else -- the proverbial next paradigm -- will pull MS down.
What the Open Source community needs to realize is that the key to our success is, in fact, to innovate the "next paradigm." The great MS fall from power will happen naturally if the paradigm shifts faster than they can adapt. So these goals are linked. Unfortunately, most people in the community are stuck in the mindset that we have to *emulate* what is happening in the Windows arena. (OpenOffice, Samba, etc.) Meanwhile, MS is trying to find ways to make the current paradigm obsolete in a way that also maintains their control. (Take a look at Office 2003 and Sharepoint..) But they are limited at the same time -- they can't completely change all at once like we can. And there are much better ways of creating the next generation of well-integrated, XML and web-services driven IT solutions than what MS has come up with so far. But unfortunately the OSS community is largely asleep at the wheel.
That's a fundamental lesson for anyone who wants to draw new customers: Offer something that they didn't know they want until they see it.
Nail head.. meet my friend hammer. (You're right on the mark here..) What we need to do is develop a solution that is so brilliantly architected that people flock to it because it is the obvious best choice. Business customers today want their IT to be cheap, zero-hastle, and a natural fit for their organizational structure. This is IT 2.0. Individual technologies themselves are not the drivers. Complete solutions are.
So what is the new paradigm? Simplicity. Users should not have to be trained on various software packages to get their work done. Everything they need should be blatantly obvious -- like the experience of browsing a well-designed webpage. Administrators should not have to worry about integrating complex tools to come up with a good solution. The foundation should already be in place such that work can be focused on customizations and improvements.
First, you claim that an open source solution is going to pop out of thin air and replace the software that Microsoft already delivers with every one of their desktops as a part of Windows.
Hardly thin air. The relevant technologies have been available for years and are now very mature in the Java arena. It is integrating them in a manner that is user friendly which hasn't happened yet. And incidentally, Word does not come pre-installed on every Windows desktop. No, Wordpad doesn't count.
I can give you many reasons why it won't work the way you want it to. The main one is this: most desktops are Windows based, most of them contain Word or Wordpad, and most people learn to edit documents with these tools.
Word is a purchasing decision at $150-200/copy of Office depending on bulk licensing. Wordpad is of no value whatsoever. You've obviously never been a sysadmin in a real-world medium/large organization have you? And let me give you one very real reason why what I'm talking about will happen: corporate america is discovering that it costs a heck of a lot of time, money and lost productivity to maintain heavyweight desktop machines that each contain hundreds of points of failure. And worse, they're very hard to secure. The future is a hybrid of thin client technology and today's desktops. All mission critical software will reside on the server where it is easily maintained. Users will have more appliation software than ever before, thanks to Open Source, but will have much less administrative control over their "own" machine, therefore reducing the PEBKAC factor significantly. Upgrades will be rolled out with virtually zero effort and go into effect immediately.
And by the way, users won't need to "learn how to edit documents" with CMS based technology so that's a moot point. They will create content and the final documents will be generated close to automatically.
They are used, they are accepted, and within the context of purchasing additional software above and beyond the PC and OS that comes with it, they are free.
You obviously don't have any knowledge of the theoretical and real-world TCO of today's office suites. Don't think like a home end user.
Students don't have a reason to switch to a web-based CMS because it's overkill, and they certainly won't do it for the sake of supporting open source software.
You're thinking way too narrow. Students won't have CMS software but their universities will. Kids won't have CMS software, but the application server in their parents basements will allow every computer terminal in the house to access the same software.
They'll STILL be using Word based formats, because it will continue to do what they need it to do.
And what exactly do they need it to do that can't be done better via web-based technologies? One of the biggest fundamental flaw in "word processing" software is that users should not be doing their own typesetting. This is more significant in a business environment where there are stricter regulations on document formatting. Another huge flaw is document revision management. There is no easy way to integrate Word into your document production workflow.
They change with the times, and if CMS systems become all the rage, they will change with the times and continue to own the market.
CMS systems are a whole new field with many players and no entrenched standard. (other than XML perhaps!) Microsoft's monopoly works solely on the principle of vendor lock-in via secret API's and document formats. But they have no leverage in this open-standards based market and they can't afford to quickly throw away their Office cash cow in the process. Also remember that no need for Office will quickly mean no need for Windows. This is the beginning of the end of Windows' dominance. It's also the beginning of the possibility for paperless offices when combined with high-dpi display technologies. As I said, mark my words and watch the next 10 years. The tech field is about to get a whole lot more interesting.
Well, OSS is essentially a "race to the bottom" to see who can devalue the software market the most.
OSS is a race to devalue the non-innovative end of the software market and to create a free platform for developers and users alike. The faster that ordinary software is commoditized, the faster the industry will move away from wheel re-inventing to actually improving the state of the art. Proprietary software has been a shackle on true progress for the last 25 years thanks to M$ and friends. That's not to say that progress hasn't occured, but it has been a lot slower than it could have been.
It is simply a race to the bottom and frequently results in cheap "free" copies ruining the chance for quality comercial software.
Mozilla is commercial software. People are being paid to write it and other people have indirect financial gain by supporting them. Don't confuse commercial with proprietary. The best Open Source projects are almost always commercially supported.
Do I feel sorry for the Opera folks? Of course not. They knew for years that the desktop browser market was being completely commoditized by Open Source alternatives. Opera never had a chance to be more than a niche player compared to IE vs. Firefox. So Opera took what it could get and then more-or-less got out of the desktop game as Mozilla/Firefox took off. Now they do mostly mobile browsers. And when mobile browsers become commoditized, they'll have to move on again. Wouldn't be my choice of business models, but somehow it manages to pay their bills.
You obviously don't have any clue how many companies are still using old versions of MS Office and have no intention of ever upgrading because it meets their needs just fine. And incidentally, these are the companies that will be first to break from the aging "word processing" paradigm itself as soon as a viable alternative is available. It doesn't make sense to upgrade MSO if it doesn't improve the business. But true innovation will be accepted and this time it will be coming from the Open Source camp, not MS.
Microsoft Word documents are one of these formats that will remain the gold standard because it just plain works.
Except that it doesn't "just plain work." It's a major hindrance to future innovation. As soon as modern web-enabled software that actually improves office productivity starts to take hold, that "gold standard" is going to disappear very quickly. What I'm talking about here is the fact that "word processing" itself is obsolete. Mark my words: in 10 years, the only people still using "word processors" will be casual home users who don't benefit from web-enabled enterprise CMS technology. (and even then, it'll all be Open Source, so many will use the same technology they have at work in a scaled down fashion..)
Jem just needs to face the fact that MS rules the generic office productivity product space and it will never be unseated. One of the keys to the castle is that there is very little innovation left in office productivity software.
You're exactly right. In fact, I would go further to say that office productivity software as defined as "word processing, spreadsheets, and presentations" is a very outdated paradigm in the modern, web-enabled world of IT. Office workers should not be "creating documents." Instead, they should be creating content via a web-enabled interface that may be stored in a database and automatically transformed into any form necessary on the server side.
The way to replace MSO is not to make OO.org into a perfect clone. That will take too long and won't actually innovate. (ie. why upgrade? what we have works.) What we need are quality web-based *true* content management systems. (ie. not blog or web-design software) Likewise, we need a new generation of flexible database software that makes spreadsheets (aka. poor man's databases) entirely obsolete. Today's presentation software is inherently flawed anyhow, so there are dozens of replacements ideas in this arena.
OO.org is a transition tool. It's not a long term solution.
I love the open document format concept. I think it is vitally important. I can't believe that enterprises and governments are willing to store critical archival documents in Microsoft Office format, and put them selves at risk of being unable to open these documents as little as 10 years hence.
Ironically, the existance of Open Source import filters for MSO's file formats now guarantees that these documents will be readable forever. There are good argument for OO.org, but this is no longer one of them.
Of course, one exception may be if MS ever manages to popularize their concept of "trusted computing" and document "DRM." But this is not likely to happen any time soon and it will be too late by the time they try.
Are you sure that Word loads dll's at boot time? IE does, because it's largely the same program as Windows Explorer. I can't think of any reason why Word would do so.
In Windows, OO.org has a lot of its own "custom" libraries that it loads at startup instead of using the "standard" proprietary Windows libraries. This is the biggest factor in why it takes so long to load. Remember that, unlike MSO, OO.org is multi-platform.
That being said, there is a lot of room for improvement. OO.org load time is improving as its custom libraries are replaced by cleaner, tighter, and "standardized" multi-platform libraries. Compare the earlier days of Mozilla to Firefox today.
Lets make another comparison: If OO.org was ported to use strictly KDE or Gnome libraries, it would load rather quickly if those environments were already in memory.
A useful feature of OO.org would be the ability to automatically save to multiple file formats. You could have a "word versions" directory for all the stuff you need to email to MSO users.
It doesn't matter so much that most people will still use MSO file formats for now. People who switch to OO.org have an open format for their own internal usage. The real key is finding ways to make the XML-based OO.org formats a killer feature and not just a philosophical nicety. Right now, the OO.org formats add no value in and of themselves. However, open formats are much easier to support in other software. This lends itself to automated document generation, searching, and modification tools. (Think of how many HTML authoring and management tools there are!)
Let me give you a quick example: how easily can you extract all the images from thousands of Word docs? Or how about a tool that does search and replace of a certain text string or image on those files? (say, you're company changed it's name and logo..)
Also, don't forget that the other major Open Source office suite, KOffice is moving to the OO.org (OASIS) file formats. I personally think KOffice has the potential to surpass OO.org because it is a fresh start and not a cleanup effort. (StarOffice was an awful piece of code)
1) The article could just as easily be titled "Failing to Profit from Open Source Software".
I agree. Fact is, anybody could have written a 3rd party proprietary interface to Snort. (Just like there is 3rd party software that tries to make Windows more tolerable) This is not a case of directly profiting from OSS. The success of Snort and the proprietary products are closely linked, but you really have to separate them. And what happens when the various open source interfaces to Snort improve to the point where the proprietary products are no longer necessary?
2) What it seems to suggest is that hybrid models combining some open-source goods and a general use of the "open-source culture" with some proprietary products is the way to go, especially for a product where you can't expect to create a lucrative consulting business.
Indeed, this is the direction that most server / security OSS developers have gone -- Open Source to seed the market, proprietary to harvest it. But there are many other options that do not sacrifice the core philosophies of the movement. Besides consulting, which indeed is not always feasible, you can sell various types of documentation (like training manuals), hold conferences, be paid for the initial creation of new features, run certification programs for techs and consultants, etc. When it comes to profitting from support services, however, developers should realize that this is generally something they should hire somebody else to do. Their own time is too valuable.
it's something I've failed doing time and time again..
The key is to develop the software from day one with users in mind. It has to be a good solution to a clear problem. You can't just write software that you think people might possibly want and expect it to magically take off. F/OSS requires just as much market analysis as proprietary software. Obviously that analysis includes a thorough understanding of the technical challenge at hand.. and this is where many traditional companies go wrong.
Slashdot Mirror
What do other, more experienced readers think about the problems and improvements related to dropping the current Linux approach for a 'bundle-like' one in Linux distributions?
OK.. this question is really 1st year CS material, so hopefully this will set all y'all newbie young'ns straight. "Bundling applications," as defined as giving every app it's own copies of used libraries, is just plain stupid if at all avoidable. Here's why:
1.) What happens when a bug or security flaw is found in a library? Without a shared copy, you must figure out which apps are using it (which may be thousands) and then upgrade every application "bundle" instead of one library for the whole system. And what if some apps are using an older version of the library which nobody bothered to patch?
2.) Disk caching. Today's hard disks may be really large, but they're still really slow (compared to the rest of the system). If you have to load separate copies of a library for each app, you lose all the benefit of disk caching.
3.) Memory usage. Shared libraries allow a single copy of the library in memory to be used by multiple applications. This also reduces load time if the library is already in memory. (ie. this is why it makes sense efficiency-wise to use either KDE or GNOME and not a mixture of apps from both) It's also partly why OpenOffice and Firefox take so long to load on Windows compared to Office and IE. (they don't use all the standard windows libraries.)
4.) Shared libraries are a major driving force in pushing application developers to stay on their toes and keep up with the progress of the library developers.
5.) You shouldn't be compiling your own apps unless you're their developer or have very specific security or optimization needs. It's a waste of time unless you're learning something in the process. Leave that job to distro package maintainers and do something useful with your time like becoming a better programmer and/or contributing to your favorite app. Once Linux ceases to be a toy for you, you'll avoid compiling everyday software like the plague.
I could go on for several points, but that should be enough to convince ya. (:
Over the next 10-20 years, monolithic groupware like Outlook will disappear and it will be replaced by a dozen separate servers and applications.
Just a clarification: although there may be a dozen separate servers and applications, they will appear as one integrated application to the end user. I'm talking about proper, natural integration here, not crude cobbling. Users will no longer have to worry about import/export of data between applications because it will all be one unified workspace. Under the hood, all the software will talk using open protocols, making it easy to extend or customize the functionality of said workspace for particular needs. This is the future of business software: Perfect modularity that is perfectly invisible.
Incidentally, this is also going to make the desktop OS almost completely irrelevent for the majority of cases..
Give me another solution that offers seamless integration of proccesses.
I see where you're coming from.. It can be done today by integrating dozens of Open Source tools, but it requires way too much geek elbow grease for the typical organization. So, listen up any of you enterprising geeks out there! This is what the business world of IT wants: seamless integration.. solutions.. simplicity.. Make it happen with OSS and you'll have a goldmine on your hands.
In contrast, transitioning to Java usually takes greater skill in object orientation.
I count that as a benefit of Java, hands down. Java itself and the surrounding tools lend themselves towards proper object oriented design. Honestly, a programmer who isn't already OO savvy is not who you want to develop your enterprise software. Send them back to school or find somebody else. This is the 21st century folks. Unless you're doing low-level coding, you should be using proper OO design.
Really, though, these technologies are so similar that you should probably just pick one and go with it.
.NET doesn't offer you anything that Java is missing. If you're not familiar with either platform yet, pick Java and don't look back.
Sure, you can knock out a solution with either platform. (or even PHP/Perl if you're a masochist) But why lock yourself into an expensive, proprietary platform?
If you want the easiest and most productive development environment, stick with .NET.
.NET today. And the Java development tools are mature and time-tested.. For those getting in on the ground floor of enterprise software development, just choose Java and don't look back. There's no compelling reason to choose .NET if you're just starting out. All the problems that have plagued Java in the past (early 90's, dot-bomb era, etc.) have been solved.
This is rather a myth unless your team is a bunch of MS junkies who've never used anything but Visual Studio. Java development tools, overall, are more powerful than those for
But do note: You should be largely looking to the Open Source community for your Java tools, not Sun. Sun is driven by committee. OSS is driven by what works in the real world. Look to Spring, Hibernate, Struts, and the rest of the Apache Jakarta project. Forget about heavyweight J2EE app servers unless you really need them. (And you can always refactor later)
Then I tried it on Windows and I can certainly see a problem. In most any Linux windowmanager, you can force windows to stay on top (and shade them). In Windows, you lose those two options and that's when the interface really becomes a nightmare to use.
This is why Gimp needs to have a *user option* so that both groups are satisfied.. Even as a seasoned Gimp user, there are a handful of cases where I would prefer an MDI interface for certain types of tasks. Can this really be so hard to accomplish? It's almost a silly political issue from what I can tell. I mean.. screw it.. Give the users what they want if it will help the project grow faster. Just make it optional so that people who use an X11 window manager and know what they're talking about can still use it the Right Way. (-:
It only sucks for those who use photoshop extensively before switching to the GIMP. I remained free of Photoshop long enough that now I am completely used to the GIMP's interface, and I don't see why it is any better/worse than Photoshop's.
I would generally agree with you as this was my experience as well. On the other hand, not everyone's brains works the way ours do. It is quite possible that some people do legitimately have more trouble learning Gimp than Photoshop. As a result, I think the Gimp project needs to make MDI vs. multi-window an *option* for the user. Some people just refuse to change no matter what. Maybe there's a good reason, maybe there's not. But ignoring the issue is not the right approach. I mean.. come to think of it, there are times when I would prefer to be able to switch Gimp into some sort of MDI mode for certain tasks. Can this really be that hard to accomplish?
As I see it, we shouldn't try to convert the professional full-time users of Photoshop, but rather the people who pirate it.
We should also try to convert the people who casually use Photoshop but don't even know how to use the advanced features that pro users demand. You wouldn't believe how many wasteful legit copies of Photoshop there are in the corporate world.. in the hands of complete novices who just need to touch up photos for a website or prepare graphics for Powerpoint slides. It's pathetic.
Piracy is a bigger threat to Free Software than it is to entrenched industry standard software, IMO.
Indeed.. I can't tell you how many Windows and Mac junkies I've seen who refuse to switch to Linux because, "I want to have my Photoshop, Premiere, SoundForge, Cubase, Dreamweaver, etc." And, of course, they could never afford to license all those programs. But to them, it's just as free as Linux with Gimp, Cinelerra, Ardour, Rosegarden, Quanta, etc. These people are pretty much hopeless to convert (although a really anal DRM scheme in Windows itself might actually help us.. heh) The only way these people will be reached is by the trickle down effect. If we can capture the corporate market, where licensing is enforced, the Free alternatives will become the big popular names and the same people who currently warez Windows/Mac crap will simply use the Free software instead "because that's what everybody is using."
Incidentally, it may also be said that P2P sharing of major label music only really hurts the small, independent artists who are often willing to freely share their music online. The original promise of P2P was that it would eventually be a way for indie artists to get their music out there with zero distribution cost. However, by the time P2P was popular with the masses, most users were just a bunch of lamers who wanted free copies of the (major label) songs they heard on the radio. And the music cartel continues strong as ever..
I'd say that Gimp on Linux is for 90% of Photoshop users (I say users including all the random kiddies who downloaded it because they want to be "pros") an absolutely solid replacement. I know that in all the years I've used it for commercial web design, photo manipulation and UI development it has never yet left me wanting.
Exactly.. This is what most people seem to ignore.. Gimp is not Photoshop. But it does happen to meet the needs of probably 80-90% of Photoshop's target market. You would not believe how many wasteful copies of Photoshop are licensed in the corporate world because Joe Idiot says, "Hey, we need a photo editor.. go buy the best thing out there." And $699 later, there's the latest version of Photoshop. (And oops.. it's hard for newbies to use (just like Gimp), so go buy a copy of Photoshop for Dummies too) Sure, Photoshop is still the best thing out there (today at least), but most of the people using it would have been fine with PSP or Gimp.
And here's the real kicker: how fast would Gimp improve if those 80-90% that don't really need Photoshop contributed a few bucks each to the project? Granted that won't happen, but there are other ways to harvest this market. The Gimp folks need to take a look at how they can capitalize on what they've developed.
MS is sponsoring not just free software but Free software, Very interesting!
.NET GDI+ compatibility layer for Mono. It would probably be a waste of time to do such a porting effort when the time could be better spent improving Gimp or KOffice's Krita.
Microsoft has always been pro-BSD licensed software because it lets them take and not give back. They hate GPL because they can't use it to be greedy and monopolize. For example, MS could create their own branch of this Paint.NET with a handful of extra features, not release any source, and include it with Windows.
Oh and can we take this and shove it on Linux?
It uses proprietary MS-only libraries like GDI+, so the general answer is no.. unless somebody writes a
I think you are mistaken. Linux does well right now as a server, but it is not very common as a desktop. Obviously, the numbers are hard/impossible to gather, but from personal experience I can tell you I only know one person who is not a programmer that runs Linux as their workstation OS.
It's true that not very many non-technical users, as a personal choice, use Linux on their desktops. However, there is a rapidly increasing use of Linux on desktops where there is support by IT staff. The "numbers" don't really show this kind of usage because it often isn't backed by sales figures coming out of RedHat, Novell, etc.
Neither Mac version is anywhere near usable, so Mac Users usually just buy MS Office.
I wonder what will happen first.. stable OO.org for Mac or KOffice (or other) taking over instead. Thing is, StarDivision's StarOffice was originally developed for Windows and Linux only. It was a very sloppy codebase and the results are still obvious today. I liken it to early Mozilla milestone releases when we were dealing with the messy Netscape code. The cleanup work on OO.org will eventually make the project very portable, but for now, it still has a long ways to go. Somebody really needs to start a serious fundraising effort for OO.org in general. There are so many people and organizations who would save big bucks if they didn't need MS Office any longer.
The number one complaint I've heard about open source games is that while the engines are often cool, the graphics are always crap.
Agreed.. it's an obvious problem. Most OSS games, for example, have absolutely awful 3D textures and it completely ruins the appearance, even where the game engine is doing a good job with geometry and lighting.
I know 4 professionals (including as I said the best graphic artist I know) who all tried to work with multiple OS projects and all gave up in frustration because no one would even talk to them. This is symptom of a problem.
That's a real shame. Have you thought about teaming up and taking a stand? Nothing will change if people aren't made aware of the problem.. If there are comments / flames that were posted to a public forum, bring them to light. "opengamegraphics.org" anyone?
One of the reasons OS X is more popular on the desktop than Linux is that no one in Mac OS X forums ever tells you to learn how to program or RTFM.
There's no such thing as just "the desktop." I wish people would stop being so generic. On the corporate and education desktop, Linux is now more popular. On the home and graphics professional desktop, Mac is more popular. Overall, Linux is rapidly pulling ahead and it's because of the fact that, using the right distribution, newbies don't have to program or Read Fine Manuals anymore. At the same time, sysadmins deploying it have a platform that's easier to maintain (on a large scale) than either Windows or OSX. When was the last time you saw a network booted Mac?
It is much cheaper and easier to just buy MS office (which is what most people do).
Are you talking about Linux or OpenOffice? You seem to have switched topics, because there's no programming or manual reading required to make OpenOffice work on its supported platforms. In fact, it's faster to get working than MS Office in Windows. (no license key, no product registration, no rebooting, no service packs, etc.)
The hostile and elitist attitudes of many open source zealots are really hurting the community.
There are a handful of jerks out there -- like any other real-world community. There are far more people who have not had the kind of experience you mention, so please don't spread bad rumors about what contributing to OSS projects is like. And remember: more than just the source code is open. If you run into a jerk, you don't have to let that stop your own participation. Most likely other people in the same project are not jerks and may even talk to the jerk in private, telling them to apologize and lighten up. I've seen it happen. If your friend had set up his own site with alternative graphics, they would have eventually been integrated regardless of the wishes of the jerk. Rule by the masses means jerks don't ultimately hold much sway. Whoever gets the job done right becomes accepted into the community.
your attitude that non-programers have no right to make comments is just the sort of attitude that pushes talent away.
I agree that many OSS projects need to adopt formal means for non-programmers to help out -- perhaps even separate from the programmers. KDE, incidentally, does a great job of this. Usually the harsh attitude of programmers towards non-programmer upon receiving comments is because they are already overloaded. And I say this with personal experience as a developer who receives a lot of unprofessional offers of help or advice. Casual, non-technical help and advice is far more difficult to integrate into the process. Good projects should have liaisons to distill non-technical help into a form directly usable by the core developers.
That being said, there is something to say for people who have complaints contributing financially. Money has a way of getting people's ears very quickly.
Are not sys admins (or capable of) and as such cannot, and should not, use linux
Linux is just a kernel. There are some software distributions that utilize Linux which are difficult for non-sysadmins to use. There are also some which are nearly idiot proof today. (moreso than Windows, almost as good as OSX) Don't make silly generalizations about who should and should not use Linux.
At least if an opensource product becomes dominant, there will still be some incentive for new features, since oss software is written by and for users, those users who desire new features and are capable of writing them, are free to do so..
The difference with Open Source is that dominance consistently leads to increased innovation rather than stagnation. Anyone who actually thinks OSS is not about innovation is a total idiot or has been living in a cave the last 5 years. Unfortunately, it's not always polished innovation, so average people tend not to notice until something like Firefox all of a sudden rises on the scene.
In an ideal world, opensource apps will commoditize the current feature sets and commercial vendors will have no alternative but to offer new and innovative feature in order to make sales.
In an ideal world, commercial developers will all be working alongside successful open source projects to meet the needs of their own customers. And meeting those needs will often occur in new and innovative ways which everyone will collaboratively benefit from. Reduced wheel re-inventing will shrink the software industry, but this is the inevitable progress of technology anyhow.
Rather than open source putting MS out of business, it seems far more likely to me that someone or something else -- the proverbial next paradigm -- will pull MS down.
What the Open Source community needs to realize is that the key to our success is, in fact, to innovate the "next paradigm." The great MS fall from power will happen naturally if the paradigm shifts faster than they can adapt. So these goals are linked. Unfortunately, most people in the community are stuck in the mindset that we have to *emulate* what is happening in the Windows arena. (OpenOffice, Samba, etc.) Meanwhile, MS is trying to find ways to make the current paradigm obsolete in a way that also maintains their control. (Take a look at Office 2003 and Sharepoint..) But they are limited at the same time -- they can't completely change all at once like we can. And there are much better ways of creating the next generation of well-integrated, XML and web-services driven IT solutions than what MS has come up with so far. But unfortunately the OSS community is largely asleep at the wheel.
That's a fundamental lesson for anyone who wants to draw new customers: Offer something that they didn't know they want until they see it.
Nail head.. meet my friend hammer. (You're right on the mark here..) What we need to do is develop a solution that is so brilliantly architected that people flock to it because it is the obvious best choice. Business customers today want their IT to be cheap, zero-hastle, and a natural fit for their organizational structure. This is IT 2.0. Individual technologies themselves are not the drivers. Complete solutions are.
So what is the new paradigm? Simplicity. Users should not have to be trained on various software packages to get their work done. Everything they need should be blatantly obvious -- like the experience of browsing a well-designed webpage. Administrators should not have to worry about integrating complex tools to come up with a good solution. The foundation should already be in place such that work can be focused on customizations and improvements.
First, you claim that an open source solution is going to pop out of thin air and replace the software that Microsoft already delivers with every one of their desktops as a part of Windows.
Hardly thin air. The relevant technologies have been available for years and are now very mature in the Java arena. It is integrating them in a manner that is user friendly which hasn't happened yet. And incidentally, Word does not come pre-installed on every Windows desktop. No, Wordpad doesn't count.
I can give you many reasons why it won't work the way you want it to. The main one is this: most desktops are Windows based, most of them contain Word or Wordpad, and most people learn to edit documents with these tools.
Word is a purchasing decision at $150-200/copy of Office depending on bulk licensing. Wordpad is of no value whatsoever. You've obviously never been a sysadmin in a real-world medium/large organization have you? And let me give you one very real reason why what I'm talking about will happen: corporate america is discovering that it costs a heck of a lot of time, money and lost productivity to maintain heavyweight desktop machines that each contain hundreds of points of failure. And worse, they're very hard to secure. The future is a hybrid of thin client technology and today's desktops. All mission critical software will reside on the server where it is easily maintained. Users will have more appliation software than ever before, thanks to Open Source, but will have much less administrative control over their "own" machine, therefore reducing the PEBKAC factor significantly. Upgrades will be rolled out with virtually zero effort and go into effect immediately.
And by the way, users won't need to "learn how to edit documents" with CMS based technology so that's a moot point. They will create content and the final documents will be generated close to automatically.
They are used, they are accepted, and within the context of purchasing additional software above and beyond the PC and OS that comes with it, they are free.
You obviously don't have any knowledge of the theoretical and real-world TCO of today's office suites. Don't think like a home end user.
Students don't have a reason to switch to a web-based CMS because it's overkill, and they certainly won't do it for the sake of supporting open source software.
You're thinking way too narrow. Students won't have CMS software but their universities will. Kids won't have CMS software, but the application server in their parents basements will allow every computer terminal in the house to access the same software.
They'll STILL be using Word based formats, because it will continue to do what they need it to do.
And what exactly do they need it to do that can't be done better via web-based technologies? One of the biggest fundamental flaw in "word processing" software is that users should not be doing their own typesetting. This is more significant in a business environment where there are stricter regulations on document formatting. Another huge flaw is document revision management. There is no easy way to integrate Word into your document production workflow.
They change with the times, and if CMS systems become all the rage, they will change with the times and continue to own the market.
CMS systems are a whole new field with many players and no entrenched standard. (other than XML perhaps!) Microsoft's monopoly works solely on the principle of vendor lock-in via secret API's and document formats. But they have no leverage in this open-standards based market and they can't afford to quickly throw away their Office cash cow in the process. Also remember that no need for Office will quickly mean no need for Windows. This is the beginning of the end of Windows' dominance. It's also the beginning of the possibility for paperless offices when combined with high-dpi display technologies. As I said, mark my words and watch the next 10 years. The tech field is about to get a whole lot more interesting.
Well, OSS is essentially a "race to the bottom" to see who can devalue the software market the most.
OSS is a race to devalue the non-innovative end of the software market and to create a free platform for developers and users alike. The faster that ordinary software is commoditized, the faster the industry will move away from wheel re-inventing to actually improving the state of the art. Proprietary software has been a shackle on true progress for the last 25 years thanks to M$ and friends. That's not to say that progress hasn't occured, but it has been a lot slower than it could have been.
It is simply a race to the bottom and frequently results in cheap "free" copies ruining the chance for quality comercial software.
Mozilla is commercial software. People are being paid to write it and other people have indirect financial gain by supporting them. Don't confuse commercial with proprietary. The best Open Source projects are almost always commercially supported.
Do I feel sorry for the Opera folks? Of course not. They knew for years that the desktop browser market was being completely commoditized by Open Source alternatives. Opera never had a chance to be more than a niche player compared to IE vs. Firefox. So Opera took what it could get and then more-or-less got out of the desktop game as Mozilla/Firefox took off. Now they do mostly mobile browsers. And when mobile browsers become commoditized, they'll have to move on again. Wouldn't be my choice of business models, but somehow it manages to pay their bills.
Don't bet too much on this.
You obviously don't have any clue how many companies are still using old versions of MS Office and have no intention of ever upgrading because it meets their needs just fine. And incidentally, these are the companies that will be first to break from the aging "word processing" paradigm itself as soon as a viable alternative is available. It doesn't make sense to upgrade MSO if it doesn't improve the business. But true innovation will be accepted and this time it will be coming from the Open Source camp, not MS.
Microsoft Word documents are one of these formats that will remain the gold standard because it just plain works.
Except that it doesn't "just plain work." It's a major hindrance to future innovation. As soon as modern web-enabled software that actually improves office productivity starts to take hold, that "gold standard" is going to disappear very quickly. What I'm talking about here is the fact that "word processing" itself is obsolete. Mark my words: in 10 years, the only people still using "word processors" will be casual home users who don't benefit from web-enabled enterprise CMS technology. (and even then, it'll all be Open Source, so many will use the same technology they have at work in a scaled down fashion..)
Jem just needs to face the fact that MS rules the generic office productivity product space and it will never be unseated. One of the keys to the castle is that there is very little innovation left in office productivity software.
You're exactly right. In fact, I would go further to say that office productivity software as defined as "word processing, spreadsheets, and presentations" is a very outdated paradigm in the modern, web-enabled world of IT. Office workers should not be "creating documents." Instead, they should be creating content via a web-enabled interface that may be stored in a database and automatically transformed into any form necessary on the server side.
The way to replace MSO is not to make OO.org into a perfect clone. That will take too long and won't actually innovate. (ie. why upgrade? what we have works.) What we need are quality web-based *true* content management systems. (ie. not blog or web-design software) Likewise, we need a new generation of flexible database software that makes spreadsheets (aka. poor man's databases) entirely obsolete. Today's presentation software is inherently flawed anyhow, so there are dozens of replacements ideas in this arena.
OO.org is a transition tool. It's not a long term solution.
I love the open document format concept. I think it is vitally important. I can't believe that enterprises and governments are willing to store critical archival documents in Microsoft Office format, and put them selves at risk of being unable to open these documents as little as 10 years hence.
Ironically, the existance of Open Source import filters for MSO's file formats now guarantees that these documents will be readable forever. There are good argument for OO.org, but this is no longer one of them.
Of course, one exception may be if MS ever manages to popularize their concept of "trusted computing" and document "DRM." But this is not likely to happen any time soon and it will be too late by the time they try.
Are you sure that Word loads dll's at boot time? IE does, because it's largely the same program as Windows Explorer. I can't think of any reason why Word would do so.
In Windows, OO.org has a lot of its own "custom" libraries that it loads at startup instead of using the "standard" proprietary Windows libraries. This is the biggest factor in why it takes so long to load. Remember that, unlike MSO, OO.org is multi-platform.
That being said, there is a lot of room for improvement. OO.org load time is improving as its custom libraries are replaced by cleaner, tighter, and "standardized" multi-platform libraries. Compare the earlier days of Mozilla to Firefox today.
Lets make another comparison: If OO.org was ported to use strictly KDE or Gnome libraries, it would load rather quickly if those environments were already in memory.
Does MS Office read .sxw files? No?
A useful feature of OO.org would be the ability to automatically save to multiple file formats. You could have a "word versions" directory for all the stuff you need to email to MSO users.
It doesn't matter so much that most people will still use MSO file formats for now. People who switch to OO.org have an open format for their own internal usage. The real key is finding ways to make the XML-based OO.org formats a killer feature and not just a philosophical nicety. Right now, the OO.org formats add no value in and of themselves. However, open formats are much easier to support in other software. This lends itself to automated document generation, searching, and modification tools. (Think of how many HTML authoring and management tools there are!)
Let me give you a quick example: how easily can you extract all the images from thousands of Word docs? Or how about a tool that does search and replace of a certain text string or image on those files? (say, you're company changed it's name and logo..)
Also, don't forget that the other major Open Source office suite, KOffice is moving to the OO.org (OASIS) file formats. I personally think KOffice has the potential to surpass OO.org because it is a fresh start and not a cleanup effort. (StarOffice was an awful piece of code)
1) The article could just as easily be titled "Failing to Profit from Open Source Software".
I agree. Fact is, anybody could have written a 3rd party proprietary interface to Snort. (Just like there is 3rd party software that tries to make Windows more tolerable) This is not a case of directly profiting from OSS. The success of Snort and the proprietary products are closely linked, but you really have to separate them. And what happens when the various open source interfaces to Snort improve to the point where the proprietary products are no longer necessary?
2) What it seems to suggest is that hybrid models combining some open-source goods and a general use of the "open-source culture" with some proprietary products is the way to go, especially for a product where you can't expect to create a lucrative consulting business.
Indeed, this is the direction that most server / security OSS developers have gone -- Open Source to seed the market, proprietary to harvest it. But there are many other options that do not sacrifice the core philosophies of the movement. Besides consulting, which indeed is not always feasible, you can sell various types of documentation (like training manuals), hold conferences, be paid for the initial creation of new features, run certification programs for techs and consultants, etc. When it comes to profitting from support services, however, developers should realize that this is generally something they should hire somebody else to do. Their own time is too valuable.
it's something I've failed doing time and time again..
The key is to develop the software from day one with users in mind. It has to be a good solution to a clear problem. You can't just write software that you think people might possibly want and expect it to magically take off. F/OSS requires just as much market analysis as proprietary software. Obviously that analysis includes a thorough understanding of the technical challenge at hand.. and this is where many traditional companies go wrong.