Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Plugins, Scripting and XUL: The Future of Commercial Software?

By Arkaein in Op-Ed
Mon May 06, 2002 at 01:54:30 PM EST
Tags: Software (all tags)
Software

The future of commercial commodity software is a murky one. With the rapid improvements in Open Source and Free Software, many advocates feel that commercial software of the future will be relegated to big iron and in house applications. I see a new potential market that could emerge: highly specialized plugins, scripts and interfaces developed by specialized consultants.


There can be little argument that Open Source Software (OSS) is starting to make inroads into everyday computing. Businesses are starting to look to Linux and BSD boxen as inexpensive mid-range servers. Web servers in particular are dominated by free software. Desktops are another matter, but progress is being made here too. Linux on the desktop has been a choice of geeks and not too many others for the most part. However, companies like Mandrake Linux (the OS on my home machine) are greatly improving the usability of Linux for average users through simple and effective installation procedures while packages such as Open Office (which recently made its 1.0 release) show great promise as alternatives to Microsoft Office for general productivity software. Although it is not quite there yet, the vast improvements over the past few years lead me to conclude that OSS will catch up to the current commodity software standard, probably in as little as five to ten years for a wide range of applications.

If I'm wrong then the Microsoft Offices and Corel Draws are here to stay. But if I'm right, what does that mean for the future of commercial commodity software? One school of thought states that it will largely fade away as OSS provides free and open applications competitive with the current state-of-the-art in increasingly specific fields, starting with common office software such as word processors and spreadsheets and eventually reaching into markets such as publishing and graphic design. Programming as a source of income would be primarily devoted to in-house projects. These projects already comprise the majority of software development (a good thing for us software developers who hope to have jobs ten years from now). The theory is that if a new feature is needed for an important application used by a company staff developers will be able to modify the software to meet the organizations needs. If the changes might be useful outside the organization they might be released into the public as well, as long as doing so will not put the company at a competitive disadvantage.

This scenario is a strong possibility for large multinational corporations with the resources necessary to maintain full-time development staffs. Smaller organizations may not have this luxury. Though in theory a small group of programmers should be able to modify any OSS application, the reality is that a lot of time and effort would be needed to learn the intricate details of any complex application's code. Adding a seemingly trivial feature may require a large investment of time if none of the staff developers is familiar with the code. This could open up a whole new market for commercial software: OSS customization.

My reasoning is that it will be far more efficient and cost effective for developers to customize software that they are intimately familiar with. Rather than having a small group of staff developers try to learn about every application used by a company, customization work could be contracted to companies or individual consultants who special in particular applications. The market for customization for common office software such as Open Office or the KOffice suite could be very competitive.

One potential problem with this situation is that of upgrades. How can software be upgraded to new versions without losing customized features? In a few cases custom work may be submitted to the maintainers of the core project and included in future versions, but in most cases custom code will be very specialized and not useful to anyone outside the organization that contracted it. I can think of three ways in which application features can be added in ways without requiring modification of the application code:

  • Plugins - applications capable of using plugins can easily add features that perform as well as code compiled directly into the application. Examples include graphics filters for applications like Adobe Photoshop, content viewers embedded in web browsers and CODECs used by multimedia viewers.

  • Scripting/Macros - large tasks can be automated by sophisticated scripts or macros. Simple macros are often created by normal users but in some cases complex tasks done by many users may justify hiring application specialists to implement a single, optimal solution.

  • Customized Interfaces - the Mozilla web browser uses XUL (XML-based User Interface Language) to manage its GUI. This allows the interface to be customized without modifying the application source code. Applications using XUL or an equivalent GUI toolkit would allow customized interfaces optimized for specific tasks to be developed easily.

These three methods for adding features to OSS applications obviously would not cover all types of desired modifications even if every application developed in the future was designed for such extensibility. In many cases there will no getting around modifying actual code. This would necessitate a support agreement with contractors to patch future versions of software (presumably at a minor fee compared with the writing of the original patch) or staff developers dedicated not to writing software modifications themselves but rather updating contracted modifications to work for later versions. This should require significantly less knowledge of the application code than is needed to make customized features from scratch.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
What do you see for the future of software?
o OSS will never replace commercial software 13%
o OSS coexeisting with commercial software 49%
o OSS replacing only commodity commercial software 8%
o Everything will be OSS 4%
o OSS with contracted custom features 11%
o Other 5%
o Not Sure 7%

Votes: 95
Results | Other Polls

Related Links
o Web servers
o Mandrake Linux
o Open Office
o KOffice suite
o XUL
o Also by Arkaein


Display: Sort:
Plugins, Scripting and XUL: The Future of Commercial Software? | 46 comments (39 topical, 7 editorial, 0 hidden)
Great idea! (3.75 / 4) (#3)
by e on Sun May 05, 2002 at 05:56:57 PM EST

A company called Cygnus, now part of Red Hat, thought of this (making for-pay extensions to OSS) over ten years ago. So, you should probably not be presenting this idea as novel today...

-- E
"You're not paranoid if they're really out to get you..."

Cygnus vs. commodity software (5.00 / 3) (#4)
by Arkaein on Sun May 05, 2002 at 06:34:06 PM EST

I wasn't thinking of Cygnus when I wrote the article mainly because my main points deal with
  • commercial commodity software
  • using plugins, etc. as an easy way of providing custom features
Also I have never had OSS customization done by Cygnus (I do have Cygwin installed on my work machine though) but as far as I know they don't deal with the type of everyday office, etc. software that people buy for their PC, nor do they add features primarily through plugins or other easily extensible architectures.

----
The ultimate plays for Madden 2003-2006
[ Parent ]

The reason there can be little argument... (2.64 / 25) (#6)
by elenchos on Sun May 05, 2002 at 08:00:10 PM EST

...about the future of Open Sores is that the groupthink on sites like this one prevents anyone who doesn't buy into the free beer mythology is modded down and called a "troll".

If you just look honestly at how speculative your claims are, you'd feel naked taking the death of closed source for granted. Traditional business models have a solid track record of success. Open Source does not. It has been tried and tried again, and a hell of a lot of investors were duped into throwing their money away on this pie in the sky.

I'm not saying Open Source is a proven failure. But if you care about telling the truth, you ought to be far less confident, and you have no justification for being dismissive towards the closed source. If you don't like it, then go get some results instead of cranking out yet another article predicting that Lunix or KDE or Abi Word is almost ready for prime time. The've been "almost ready" for years and years now.

I'm betting they will always be almost ready.

Adequacy.org

open sores offends me more than USian [nt] (1.00 / 2) (#16)
by infinitera on Mon May 06, 2002 at 12:30:17 AM EST



[ Parent ]
Subject line. (3.00 / 9) (#18)
by elenchos on Mon May 06, 2002 at 01:52:40 AM EST

That's nice. Why are you sharing this information with me? Have you mistaken me for someone who gives a shit about whether you say USian or American?

I think the continual obsession of Lunixists over the niceties of terminology -- like trying to get everyone to redefine "hacker" to give teen computer jocks a free self-esteem boost -- is one of the most revealing aspects of the Open Sores mentality. This isn't really about good software or about how we get our software. It is a geek identity movement, whose agenda is ruled by the need to give geeks warm lovey feelings about themselves. Geeks want to be important and they want respect, and they don't care what lies they have to tell to get it.

And then they have the nerve to criticize African-Americans and feminists for being "politcally correct". Take the political correctness crusade out of Open Source and there is next to nothing left.

Adequacy.org
[ Parent ]

nope (1.66 / 3) (#19)
by infinitera on Mon May 06, 2002 at 02:10:34 AM EST

I'm just biting. In my own little way. Take your Adequacy/old skool Slash troll terms, and stuff them where the sun don't shine. Thanks, HAND.

[ Parent ]
Must you disrupt every discussion? (3.28 / 7) (#20)
by elenchos on Mon May 06, 2002 at 02:24:20 AM EST

If you want to write about me, or about Adequacy.org, why not put it in a diary? People come to K5 to discuss the topics in the articles. They do not want to constantly hear about your beef with some other site, or or your feuds with other posters.

It is sad that your type of off-topic posting is tolerated now; it didn't used to be.

Adequacy.org
[ Parent ]

off-topic is a way of life, it's like zen [nt] (1.33 / 3) (#21)
by infinitera on Mon May 06, 2002 at 02:29:00 AM EST



[ Parent ]
Linuxists? Open Sores? (3.00 / 4) (#25)
by MrMikey on Mon May 06, 2002 at 10:33:38 AM EST

Is this what passes for reasoned, mature discussion where you come from?

You made a few good points in your first post

Traditional business models have a solid track record of success. Open Source does not.
. . .
I'm not saying Open Source is a proven failure. But if you care about telling the truth, you ought to be far less confident, and you have no justification for being dismissive towards the closed source.
but your "Open Sores" rants don't do anything to enhance the discussion.

[ Parent ]
Stop it. (none / 0) (#35)
by Lord Snott on Tue May 07, 2002 at 07:53:33 AM EST

I want to agree with you. But you're being a wanker about it.

I'm tired of the "free, open source alternative" being "ALMOST" ready forever as well, but you're diminishing our argument.

Open Source will NEVER compete in some areas, ever. It's really simple. Open source is for a different market. Maybe I should write my own essay on why, then I could make myself seem petty and contrary by making bad puns about "open sores" (actually, I quite liked that one :)

I don't know whether to cheer you on, or smack you in the head!

Oh well, until I decide, "Carry on Trolling"!
(I loved those old movies...)
.
~~~~~~~~~~~~~~~~~~~~~~~~
This sig in violation of U.S. trademark
registration number 2,347,676.
Bummer :-(

[ Parent ]

Universities (none / 0) (#40)
by epepke on Tue May 07, 2002 at 05:48:51 PM EST

Traditional business models have a solid track record of success. Open Source does not.

(sigh...) Once again, the Open Source production model is more closely related to academia than it is to traditional business. It certainly has its roots there. Are you willing to assert that universities do not have a solid track record of success? If so, gee, I wonder why all those traditional businesses want job applicants to have degrees. I also wonder whether you ever use anything developed in academia like, I don't know, maybe the world-wide web? Naw, it couldn't be.


The truth may be out there, but lies are inside your head.--Terry Pratchett


[ Parent ]
are you referring to http? (3.50 / 2) (#43)
by eLuddite on Wed May 08, 2002 at 03:58:09 AM EST

It's a protocol.

Open Source is not any of the following: computer science, academic research, business strategy, development model, technical stratagem, etc. That's right, ESR is full of shit; OSS neither is nor implies a unique, interesting or efficient development or any other kind of model. Open Source is strictly software distinguished according to its license. Everything else written about it, both pro and con, is empty politics. That means you cannot make claims for anything about OSS with authority except its list of copyright restrictions.

---
God hates human rights.
[ Parent ]

Everything (4.00 / 1) (#45)
by epepke on Wed May 08, 2002 at 01:22:36 PM EST

The whole idea of HTML started in academia, as a means of distributing papers more quickly than journals. The first reasonably functional browser came out of the University of Illinois.

Open Source is not any of the following: computer science, academic research, business strategy, development model, technical stratagem, etc.

And a pencil is not writing.

That's right, ESR is full of shit;

Back at you. ESR is not open source. He's a person.

I will try to explain my point so that you might understand. My point is twofold. First of all, "business models" do not explain the entire universe, nor do they even explain how everyone on the planet makes a living. Many, many people do things for a living that would not work in a strict ROI model but people still want it done anyway.

Second of all, open source (the basic concept easily understood by English speakers, not the opinions of one person) is an entirely natural model for acadamia, perhaps even the most natural model for the publically funded universities and government programs that make up the bulk of academia. In fact, when other models are used, people calling themselves "taxpayers" generally make a very big stink about it.

Academia operates on something called funding, not revenue. While this requires a high degree of competition and mastery of realpolitic, it does not require the attributes of a "business model." In fact, attempting to impose those attributes generally results in the destruction of a public research program.

Since academia continues to exist, despite vigorous efforts of many people, and because generic models that can accurately be called "open source" are natural in academic settings, open source software will continue to be developed and released. Arguing that open source sucks or is doomed to fail or whatever on the principle of a "business model" is therefore to miss a basic point of reality by several kilometers.


The truth may be out there, but lies are inside your head.--Terry Pratchett


[ Parent ]
yawn (3.00 / 2) (#46)
by eLuddite on Wed May 08, 2002 at 01:57:18 PM EST

I have no idea why you think academia didnt exist before OSS licenses were first elaborated on paper, but it did. Whatever remorselessly logical conclusion it is you are trying to coax out of a series of byzantine and unrelated opinions about this, that or the other thing, you have faild to do so. Again, the only incontestable inferences you can make about OSS software is a list of actionable copyright violations.

Arguing that open source sucks or is doomed to fail or whatever on the principle of a "business model" is therefore to miss a basic point of reality by several kilometers.

Well then dont do that.

---
God hates human rights.
[ Parent ]

XUL? (3.33 / 3) (#11)
by jazzido on Sun May 05, 2002 at 10:49:16 PM EST

XUL is dead. Long live XWT

BTW, I discovered that wonder thanks to a K5 textad

--
"Patriotism is the last resource of scoundrels" (Samuel Johnson)

and why is that? (none / 0) (#32)
by panserg on Tue May 07, 2002 at 01:01:15 AM EST

Give me just three reasons:
  • why does most of XML developers know or use XUL and never heard or don't use XWT?
  • why is XWT better than XUL?
  • what is the best example of XWT making it more preferable than XUL?


[ Parent ]
XUL and XWT have different goals (none / 0) (#37)
by valency on Tue May 07, 2002 at 01:06:17 PM EST

XUL/Gecko tries to be a completely generalized mechanism for rendering arbitrary XML documents. XUL/Gecko can style, transform, and render your employee database as a visual document. XWT cannot. XUL/Gecko can render pretty much any HTML document on the web; XWT cannot.

XWT tries to do just one thing and do it well: user interfaces. XWT is far simpler and much faster. It is easier to write user interfaces in XWT, and the XWT engine is much smaller (500kb vs 13MB). XWT downloads on the fly, so there's no installer to run. XWT runs on any platform with a JVM, and runs natively on Win32 (native Solaris/Linux/MacOSX support coming soon).

most of XML developers know or use XUL

Despite my respect and reverence for XUL as the primary inspiration for XWT, I have to say that this statement is so far from being true, it's downright laughable.



---
If you disagree, and somebody has already posted the exact rebuttal that you would use: moderate, don't post.
[ Parent ]
quick answers (none / 0) (#38)
by speek on Tue May 07, 2002 at 01:11:35 PM EST

* why does most of XML developers know or use XUL and never heard or don't use XWT?

Because it's much newer.

* why is XWT better than XUL?

It's simpler to learn and to use. It's small, allowing XWT to replace html and applets as an instantly downloadable client/server application. That said, it may not be "better", but it has some advantages that might make it more useful in the long run.

* what is the best example of XWT making it more preferable than XUL?

Like I said, it's pretty new, so I'm not sure there are big examples showing off it's abilities. However, there are some interesting demos, and you can find out all you want at the website.

--
al queda is kicking themsleves for not knowing about the levees
[ Parent ]

Missing (4.00 / 1) (#12)
by ucblockhead on Sun May 05, 2002 at 10:49:42 PM EST

How can you talk about a subject like this without going into component models like COM or CORBA, which are intentend to solve exactly the sorts of problems you are talking about?

Without that, it reads like a "how to extend software" article written in 1992.
-----------------------
This is k5. We're all tools - duxup

A little different? (4.00 / 1) (#13)
by Arkaein on Mon May 06, 2002 at 12:07:21 AM EST

Disclaimer: I haven't done any direct programming with various component frameworks such as COM or CORBA, so I may be incorrect in some of my following assumptions.

It seems to me that the various component models are more concerned with reusable components for new applications. An example would be khtml which could be easily used in an application's help system. Models like these are having a big impact on software, but more in terms of creation of new software, both open source and commercial. In both cases they mainly reduce the amount of time needed to create new applications.

I though about component architectures briefly while writing this article but decided that they don't really favor commercial or open source software. I suppose really easy to use components may make it feasible to easily produce entire applications made to order, but most of these would have little advantage compared to the OSS software that would already exist (and may be using the same components).

If anything, wide-scale reuse of components will make software more homogenized. My prediction is that in the future the money will be in differentiation and specialization.

----
The ultimate plays for Madden 2003-2006
[ Parent ]

Not true (5.00 / 2) (#14)
by carbon on Mon May 06, 2002 at 12:16:07 AM EST

Although, to keep your example, khtml can be used to make displaying HTML in new apps easier, it can also be swapped out with another html renderer in existing apps. Furthermore, khtml is part of a larger suite of viewers and readers in kio/kparts/etc, all of which are swappable in and out automatically, without any source modification. For example, if you add a new KIO plugin to support loading pictures off a digital camera (and this has been done, too) then any app that has a file open dialog and support for opening images can import off your camera.

Component systems are exactly what you're talking about.


Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
That's Plugins (none / 0) (#17)
by Arkaein on Mon May 06, 2002 at 12:53:23 AM EST

That's a little more flexibility than I realized was available in KParts. Anyway, your example is basically the same thing as the plugins I talk about in the article.

----
The ultimate plays for Madden 2003-2006
[ Parent ]

Actually (none / 0) (#22)
by carbon on Mon May 06, 2002 at 03:08:43 AM EST

That example was more about KIO then KParts, but you get the idea. KParts is even more flexible : it allows embedding one type of application's document inside another's document, and getting the viewing and editing of both handled automagically.


Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
I disagree (3.33 / 3) (#15)
by Neuromancer on Mon May 06, 2002 at 12:19:16 AM EST

Few companies would pay for software to be released into the public or OSS domain. It might not seem like it to 90% of the world, but software is actually hard to write well. I am a professional software engineer, and I know that many of those in the business resubmit their resumes once a year, because they think that they'll make more money. Do you think that we'll all just jump off, start working at 7-11, and write software for free? Doubtable. There will ALWAYS be commercial software out there.

Well, it depends on what the software is for. (4.33 / 3) (#23)
by gordonjcp on Mon May 06, 2002 at 04:45:00 AM EST

I've written stuff for people to do specific jobs. I'm quite happy to distribute this for free, and the companies who buy it can distribute for free too. Since the software is already free as in speech, it may as well be free as in beer. I got paid for it, and it works for what the customer wants. If anyone else goes to use it, it's there - I don't have to write it again. The only problem is that since a lot of the interfaces are covered by NDA it can't be released.
It's like this though - if I tell you how to make a tool, it doesn't necessarily mean you know how to use it. You might still have to pay me to come round and use the tool.

Give a man a fish, and he'll eat for a day. Teach a man to fish, and he'll bore you rigid with fishing stories for the rest of your life.


[ Parent ]
Open Source Hubris (3.80 / 5) (#24)
by Simon Kinahan on Mon May 06, 2002 at 07:51:03 AM EST

Isn't "Commercial Software is Dying" the message of 1999 ? :) I thought it might happen then. I don't now.

Don't get me wrong here, open source software has achieved a lot more than anyone would have thought back in 1990, but the great majority are implementations of a well understood specification, or imitations of an existing product. They work especially well when the software has a simple, machine-centric interface. Projects with heavy UI requirements do not work so well. The majority of projects that don't fit this description are the results of research or standards-setting exercises that released their code to the public.

Open source works well enough for these kinds of things that commercial projects whose sole remit is "An implementation of the <blah> API that happens to belong to XCorp" are a dying breed. However, open source projects don't do what commercial companies do: they don't go out and discover new user needs and attempt to fill them. There are at least two reasons: its much harder to convey a new vision to a disparate team of volunteer developers, and volunteers just don't have the incentive to try to do new things.

Simon

If you disagree, post, don't moderate

Always different, always the same! (3.00 / 1) (#28)
by thebrix on Mon May 06, 2002 at 04:17:28 PM EST

I don't see any problem with 'imitations'; great artists borrow, lesser artists steal.

Window, icon, mouse, pointer has been around for 30 years and has insinuated itself into popular consciousness for 10 of those; for that reason it won't go away easily, despite premature obituaries over the years, and contenders such as virtual reality and speech recognition have fallen short. So small improvements on the existing paradigm are worthwhile.

That said, a big problem with Open Source is the sheer logistics of activities such as usability testing: getting numbers of people together and monitored while performing tasks is difficult enough within a single organisation. However, I note that Sun has been doing usability work for Gnome; this is excellent news and more such work is badly needed.

I don't agree with the comment on user needs. Subscribe to any KDE or Gnome mailing list, IRC channel or forum, for example, and these fly around by the hundred with the volunteers there to process and prioritise them. If you work on commercial software you put out questionnaires, hold workshops and so on to do exactly the same thing; in many ways Open Source development is more effective as commercial organisations are very poor at utilising the Internet to capture information from users (more often, they tend to be slightly frightened of them).

As for 'volunteers not having the incentive to try to do new things', I refer you to my (former) campaign Web site.

[ Parent ]

Whole different world (none / 0) (#29)
by Simon Kinahan on Mon May 06, 2002 at 05:05:15 PM EST

I don't see any problem with 'imitations'; great artists borrow, lesser artists steal.

There is no problem with imitation except that it doesn't advance the state of the art. If the nature of open source software remains imitative, a completely open source world can't happen, because the rest of the world moves on.

I don't agree with the comment on user needs. Subscribe to any KDE or Gnome mailing list, IRC channel or forum, for example, and these fly around by the hundred with the volunteers there to process and prioritise them.

Thats not what I meant. Small usability improvements matter in desktop systems, but thats a pretty small portion of the market for software by value. Open source has gotten better at these things. Its still not good enough, but its better.

What is still way out of the range of open source projects is the kind of software that supports businesses. Many corporations now depend on software to support their business models, and having the software available is worth a hell of a lot to them. They have to keep evolving those business models to compete. A lot of that software is written during consulting projects, but some of it is also off-the-shelf. Indeed, a lot of what gets written while consulting for one firm is then sold to others. Anyone who think software products that sell for half a million to several million pounds (not dollars) per installation is going to be open sourced is living in cloud cuckoo land.

Simon

If you disagree, post, don't moderate
[ Parent ]

Straw man (none / 0) (#33)
by thebrix on Tue May 07, 2002 at 06:34:33 AM EST

On the first point, see the rest of the original post.

On the second, of course Open Sourcing software worth X thousand pounds per shot is absurd; no model is appropriate in all circumstances!

People like IBM and Sun sell commercial software - and support and consult on it - but also donate to Open Source projects because, that way, they can influence the environment on which their commercial products run.

Having done work for a bank which tried to get a somewhat customised version of Windows out of Microsoft (and failed after endless wrangling), and two projects which required customised window managers so ruled out Microsoft operating systems straight away, that control over the environment matters.

[ Parent ]

Err, no (none / 0) (#34)
by Simon Kinahan on Tue May 07, 2002 at 07:45:18 AM EST

A straw man is a logical fallacy consisting of assigning to your adversary views he did not express. I did no such thing. The original article expressed the view that, aside from customising OSS, commercial software "will largely fade away as OSS provides free and open applications ... in increasingly specific fields". That is the view I was originally arguing with. You may or may not support it, but you misunderstood what I was saying, so I corrected you. That is all.

My point was that there remain fields on which OSS has not a hope of touching, because they are too high value and require too much specialist knowledge.  I'm not debating that control over the environment matters. In infrastructure type stuff, I would be surprised if OSS did not come to dominate in the longer run.

Simon

If you disagree, post, don't moderate
[ Parent ]

releasing source? (3.00 / 2) (#26)
by jacoplane on Mon May 06, 2002 at 11:38:06 AM EST

"The theory is that if a new feature is needed for an important application used by a company staff developers will be able to modify the software to meet the organizations needs. If the changes might be useful outside the organization they might be released into the public as well, as long as doing so will not put the company at a competitive disadvantage."

You are saying that companies only release the source of software they have modified as long as doing so will not put the company at a competitive disadvantage. This situation would exclude GPL software from being used, since there the source must be released. I would not view this as a bad thing however, since companies who take useful software from the public domain should be encouraged to return to the p.d. You seem to be advocating a BSD license, which could be confusing since most OSS is GPLed.

Only if distributed (5.00 / 3) (#27)
by Arkaein on Mon May 06, 2002 at 12:47:45 PM EST

In most of the situations I was thinking of companies would want features added to software for use only within the organization. As long as the binaries stay within the org the source does not have to be distributed. I can see no reason why the GPL would prevent hiring a consultant to add a feature with an express agreement not to distribute binaries or code to anyone else.

In most cases I don't think it would be a big deal either way. If company X wants a special payroll error flagging tool added to Gnumeric 5.6 and the tool is customized specifically for their employee database, will they care if anyone else sees the code? Some companies overly protective of their IP might care, but really the code probably would be useless to everyone except that one company. Plus consultants would probably charge less if they can sell the same mods to other customers.

One place where the effect of the GPL is a little ambiguous right now is with the plugins, scripts, and dynamic interface idea. None of these things are specifically required to run the base application (unlike the LGLP'd shared libraries an app might use) so whether they are legally considered part of the application is a little murkier. I think there may have been some GPL's plugin based applications where this issue may have come up, but I can't remember any details off hand.

----
The ultimate plays for Madden 2003-2006
[ Parent ]

GPL & Source Release (5.00 / 3) (#31)
by Matrix on Mon May 06, 2002 at 05:20:06 PM EST

This is a common misconception about the GPL. Since the other reply didn't only address this point, I will attempt to provide a clarification here.

The GPL does not require that modified source be released to the public. It never has, and I really hope it never will. In fact, I believe that one of RMS' fundamental freedoms is the right to make modifications for private use.

What the GPL requires is mind-numbingly simple. Anyone you give a copy of a GPL'd work or derivative work to, even if that copy is in binary form, has the same rights you did. These include the rights to request the source for the binary they recieved and the right to give the binary (and rights) to other people.

So if you take emacs and hack in a kitchen sink, and then give it to all your employees, they're the only ones you're required to give source to. If they then pass it on, I believe that they're the ones who're required to provide source.


Matrix
"...Pulling together is the aim of despotism and tyranny. Free men pull in all kinds of directions. It's the only way to make progress."
- Lord Vetinari, pg 312 of the Truth, a Discworld novel by Terry Pratchett
[ Parent ]

This has always been the dream (3.33 / 3) (#30)
by epepke on Mon May 06, 2002 at 05:15:36 PM EST

This has been one of the Holy Grails of computing.

Consider the work of Paul Haeberli in the late 1980's. He had a system that essentially allowed one to build a complete user interface on the SGI desktop as a user, not a developer. The visualization packages AVS and APE (which was free) were essentially just plug-in frameworks. The Ray Dream Studio animation package is also heavily built on plug-ins; just about every aspect of the system can be changed. Great stuff. I was thinking of, for fun, sketching out such a system for Cocoa under OS X, and of course there's Open Step if OS X isn't free enough for you.

Of course, a lot of in-house software development proceeds essentially the same way. Where I work, we use an Oracle system with HR and financials and all the trimmings, but most of the daily work gets done using custom bags that we have hung off of it.

The hardest part of the technical job is establishing a solid framework for these things to live. I don't think it's practical to try to do this without a standards committee and a rather massive centralized effort. However, these things can be done.

Here's why it won't work. Most people don't care enough at the time of purchase or acquisition or whatever you want to call it to consider this a major factor in making a decision. A company that need not be named will produce something that looks to a purchase manager like much the same thing but isn't, or it will eat all the companies and people making plug-ins.


The truth may be out there, but lies are inside your head.--Terry Pratchett


This again? (1.00 / 1) (#36)
by Lord Snott on Tue May 07, 2002 at 08:42:40 AM EST

Surely the K5 crowd is smarter than this?

Open Source software has it's place. That's the enthusiast market. Big companies generally don't use OSS for servers because they DON'T CARE. They REALLY DON'T CARE. They just pay someone to deal with that area, and they pay so they DON'T HAVE TO WORRY ABOUT IT.

The K5 crowd, my group of friends, and even those icky /. people all enjoy playing with this stuff, coz thats what we like. We're like the Doctor, tinkering with the Tardis, always trying to get that extra bit of performance, or trying to do impractical stuff just for the nerd value.

At work (I work at a bank - don't hate me) I spend half the day telling and re-telling people how to attach a file to an email.
They complain they get messages about their inbox being full, so I tell them to save their attachments to d: drive, and they say "but I already saved it to <blah> folder," and I say "No, thats a folder within your LotusNotes, so your not freeing space, just moving stuff around." "But your just moving it to <foo> folder there, whats the difference?"

These people don't care. They can't even fathom Windows or a basic email client. Most of the pc using world is like this. OSS developers code for us, not for them.

I have a Linux server connected to my Win2k machine. I've also had a Wavelan card communicating wirelessly and seamlessly to another server (when Nemo could keep his box up and running :)

I love playing with this stuff, we all do. But they DON'T. And they have the money.

Open Source will not take over because NO ONE CARES. Linux has done very well, but it's not even close to being desktop OS for the masses. We love our Apache, our VNC, our Gnome, But they can't even pronounce Gnome! They have no chance of using it. Outlook confuses too many people, what chance does anything less have? (Less? that came out wrong, I meant less user friendly.)

If I see another bloody article talk about OS catching up to the proprietry stuff I'm gonna spew! "Almost at release 1.0" What the hell is that!

The mainstream closed stuff have had over twice as long establishing a consumer base, and because they get paid, they pay attention to what the masses want. OS developers give us what WE want. And whether we like it or not, we are a minority.

I wish I was trolling, this is depressing. I'm gonna get myself a cup of coffee.
.
~~~~~~~~~~~~~~~~~~~~~~~~
This sig in violation of U.S. trademark
registration number 2,347,676.
Bummer :-(

Development (none / 0) (#39)
by radeex on Tue May 07, 2002 at 03:59:03 PM EST

Open Source software has it's place. That's the enthusiast market. Big companies generally don't use OSS for servers because they DON'T CARE. They REALLY DON'T CARE. They just pay someone to deal with that area, and they pay so they DON'T HAVE TO WORRY ABOUT IT.

That's how the people at the very top think. Certainly, the management of the team of programmers who are developing your software care about the efficiency they work at?

These people don't care. They can't even fathom Windows or a basic email client. Most of the pc using world is like this. OSS developers code for us, not for them.

Ok, so far you know how the people at the very top of your organization think, and how the end-users of your organization think. These people do not care about OSS, of course. But perhaps the development department does?

I'm not arguing that people should use OSS here; perhaps I'll think about it some more and post that in another comment. This one is just to point out the obvious flaws in your rant. Get the coffee before you post next time. ;-)
--
I DEMAND RECOMPENSE!
[ Parent ]

Remorse. (3.00 / 1) (#44)
by Lord Snott on Wed May 08, 2002 at 05:38:01 AM EST

I will have that coffee first before I post next time :P

I had a bad day. Our "new" PC's at work arrive with NT on them. Why not? It's only TWO generations old. And they're running LotusNotes dated over 3 years ago, and they're configured really bad. Not that any of my co-workers noticed - their last machines were 486's running an early version of OS/2.

My last job was as Service Manager at a computer company, I was used to people setting up machines PROPERLY.

It annoys the crap out of me to see the largest Bank in Australia (our profits last year were over AU$4 Billion) purchasing these systems as an upgrade.

Then everyone looks at me (they know where I used to work) and blame me because I'm "one of those computer people" that are making their lives miserable.

A friend who runs an ISP supports linux (all his servers are Debian), but his systems are always going down, too. Like we always said at Cougar (my old job), "It's only free if your time is worthless." Proprietry software generally comes with support (some might say you need it when you go proprietry ;)

I feel that although OSS CAN be more powerful/flexible, it requires greater tech knowledge to get the same out of it as you would closed source (generally speaking). My earlier point was most people in the world don't care enough to learn, or put the effort in to get that extra bit out.

I think it's gonna take a bloody long time before the difference between open and closed software becomes irrelevant. Not 5 or 10 years, but more like 20. I hope I'm proved wrong! :)

Thanks to Radeex and Arkaein for the feedback, and the sensible argument!
~~~~~~~~~~~~~~~~~~~~~~~~
This sig in violation of U.S. trademark
registration number 2,347,676.
Bummer :-(

[ Parent ]

Businesses do care... (none / 0) (#41)
by Arkaein on Tue May 07, 2002 at 07:20:00 PM EST

just not about the same things that OSS advocates and programmers care about. Corporate management types care about words like TCO and ROI. OSS has done a lot of catching up in a short time, the differences in quality and usability are shrinking in just about every area.

If Microsoft's (or whatever company's) software is seen as managment as unequivocally superior to the major OSS competitors. As the percieved quality gap skrinks the up front costs start to be come a bigger and bigger factor in making new installation decisions.

Usability is improving too. Just look at where GNOME and KDE are compared to a few years ago. No they're "not there yet", in many cases, that's not the point. Extrapolation from current trends tells me that my Mandrake 11.0 distro with KDE 5 will be every bit as usable, easy to install and more stable than MS Windows XYZ 5 years from now. Undoubtably industry will come out with some killer apps that it will take OSS a few years to catch up with, but if we can get along without them now they won't be essential then.

Of course this is a big digression from the point of my article. I wanted to introduce some ideas on how the commercial software industry could successfully adapt to the new challenge of OSS. Instead the discussion has mostly devolved into yet another flame war between open and commercial software. I guess it's harder to add thoughtful discussion about how industry would adapt under sufficient pressure than just to dismiss the whole idea, but I expected better on K5.

You are right about one thing though, this whole discussion is pretty depressing.

----
The ultimate plays for Madden 2003-2006
[ Parent ]

Need to do more editing... (none / 0) (#42)
by Arkaein on Tue May 07, 2002 at 08:04:33 PM EST

...If Microsoft's (or whatever company's) software is seen by management as unequivocally superior to the major OSS competitors then they will continue to purchase the proprietary solutions regardless of the up front costs...

----
The ultimate plays for Madden 2003-2006
[ Parent ]

Plugins, Scripting and XUL: The Future of Commercial Software? | 46 comments (39 topical, 7 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!