Rumors are that Transdyn are negotiating to get back into writing the code for Big Dig. Hopefully they will have better luck the second time around. I'm sure there are lots of helpful comments in the current source revs in the ITS software for whoever develops it (particularly Transdyn:P)
Holy cow, what a story. I can't believe Transdyn would be let anywhere near the project again after what's happened.
On the other hand, for all the posters who say that the source code for all taxpayer paid for software should be required to be opened, Transdyn was a commercial software product modified for this project.
Governments also buy Oracle, SAP, Windows, and every other conceivable commercial software product. Governments are just another customer, and tax dollars just those customers way of paying for it. And open source just becomes a hopeless mantra then.
Of more realistic hope would be requirements by law to consider open source first to fill requirements, and for taxpayer funded work to be done by those taxpayers.
Even those common sense requirements are fought against with all the money it takes from Microsoft and outsourcers.
In anticipation of one concern I saw raised repeatedly, that is the concern of memory usage from multiple instances of the Gecko engine in each Mozilla component versus apparently one being shared in Mozilla Suite.
These concerns are likely to have been raised by non-programmers with it being seemingly sort of obvious that duplicated code is unnecessary bloat, or perhaps from loading and looking at memory resources and seeing more taken.
An operating system only pages into memory from disk that which is being used. We would have been in big trouble throughout the history of computing if memory had to hold the whole program, although nowadays such huge amounts of memory are required that people may think that.
It is more the case of large amounts of memory being allocated by programs and the operating system for data, not code. We used to have to work with small data buffers too, but with hardware now allowing nearly infinite amounts of memory at a shot to be accessed by programmers, ever approaching infinity amounts of memory are being accessed by programmers. It's painless to them now, and makes for much easier programming.
But in reality, only the code being executed is pulled into memory by the operating system as its needed. The redundant Gecko engine code will sit in multiple.exe's on disk, so if disk space were a concern then there'd be a legitimate concern. We used to have to worry about disk space too, but that is something that really has reached infinity, so no problem there.
One may then point out that more memory has been observed to have been acquired, and the multiple Gecko thing is a pig. Again, this is more data than code. Even with a single instance of code, if you use it for multiple things it's going to grab multiple sections of memory for each use.
So in reality, in theory, because I haven't tried this at home, there's really not that much difference between a single Gecko engine grabbing two memory areas for FF and TB, or two Gecko engines grabbing an area apiece from FF and TB. It's mostly data work space needed to do the job on a web page and email.
And not only is there really no memory difference, but there should be faster performance and less possibility of internal data conflicts with two clean instances of Gecko working FF and TB versus a shared instance, not that there is any conflict in working, debugged code. Someone may very well say that a bunch of Gecko code it is spawned off for each use, and then it becomes clearer that it really doesn't matter if it starts off as two or is spawned off a couple of times from one.
And lastly, the Firefox Classic distro will be able to take advantage of any Mozilla FF, TB, etc. component integration with the Gecko engine that will conditionally compile and bring in focused Gecko code for the component at hand, versus one monolithic engine shared by all. Whether done or not now, there is certainly potential for focused integration with each component that Firefox Classic would inherit on each release.
So the response that two Gecko engines is a liability is that, actually, no, it's a plus.
I thought I was done commenting but I came back to add this. From the two slashdot threads I've read in the last few days and all the links to Mozilla dialogue, I thought Mozilla/Firefox defenders did a good job explaining how to soup up Firefox to do the things being complained about as missing, sometimes with a config text file option, sometimes an existing plugin, and sometimes just a setting from an FF menu.
What little that couldn't be done was attributed to Firefox UI decisionmaking frustrating those who wanted to add professional features, along with some dislike for people having to piece together all these tweaks and plugins and redo it on every update, and even then not having some patches in there that they would want.
So I posted the blurb at bottom here this afternoon, but I wanted to add another thought. The distro would have -
all the config settings, UI changes such as pulling the little search field off the UI, classic skin, etc., and everything else described as how to implement what was wanted,
with all plugins installed and configured that add the Mozilla Suite functionality and usability desired,
and all the patches implemented to this branch that FF WONTFIX to add all the advanced keyboard behavior and such for desired Suite functionality. My understanding from reading is that most of those patches are already done or not that hard but just refused to be implemented in FF.
This would be a CVS branch of FF, TB, the new composer, etc. where only these functionality patches have to be reapplied to new releases. Most seemed to be triggered by keyboard shortcuts or additional menu options so seem fairly external and addon to core code.
In addition, with a classic skinned, UI tweaked FF and TB then create a shell menu control that FF, TB, and other Mozilla apps plug into, who knows, maybe a defined plugin interface that allows Suite users to plug other interoperable browser related apps made available for it by third parties. At that point you should have essentially classic Mozilla Suite (I'm a real classic Netscape 7.02/Mozilla 5.0 user, I'll have to be dragged kicking and screaming from it) with only an outer menu shell and mostly keyboard triggered functionality tweaks to merge to current Mozilla components FF and TB.
That's a distro that should be able to ship with each new release. However, in my opinion if that is done by Mozilla Suite advocates my opinion is that Mozilla Foundation should put their stamp of approval on that and call it Firefox Classic.
I would even upgrade off of 7.02 for that.
rd
My previous post: "I would say after reading all this that it possibly makes more sense to have distro of FF, TB, etc. with extensions that do such things as provide a GUI options panel and every other thing that brings FF up to Navigator standards, even patches that FF WONTFIX like CTL-ENTER preference to match Navigator.
They should bundle the distro as the equivalent of Mozilla Suite (with a different name) with all extensions included and tweaked, which is the whole point of a distro.
This makes much more sense to me than the more massive effort required to maintain and enhance SeaMonkey."
I would say after reading all this that it possibly makes more sense to have distro of FF, TB, etc. with extensions that do such things as provide a GUI options panel and every other thing that brings FF up to Navigator standards, even patches that FF WONTFIX like CTL-ENTER preference to match Navigator.
They should bundle the distro as the equivalent of Mozilla Suite (with a different name) with all extensions included and tweaked, which is the whole point of a distro.
This makes much more sense to me than the more massive effort required to maintain and enhance SeaMonkey.
As I specified in my post, the confusing announcements are a fumble. The software is good, and the strategy seems sound for producing more. You and I are unusual, because we're technologists, in deciphering the announcements to mean anything but "we quit".
yes, you're right, Doc, your analysis was right on. It looks to me like it was actually lack of announcements and more what the Mozilla Suite developers posted that got the bad publicity.
I think the actual Mozilla announcement at this point should only be encouraging to those 25 million Firefox and counting downloaders, but the handling of all this was fumbled badly.
I hope they do the things you recommend in grandparent and make lemonade out of this lemon.
But announcing the transformation in terms of the demise of the organization, and "I'm sorry there will be no next version", is a total fumble.
Actually, rather than slow down Firefox downloads, it's discontinuing an older suite because of the success of Firefox downloads.
That is emphasis on and encouragement of continued said downloads.
And those who support the older suite are free to continue development on a new branch frontending Gecko page rendering without the Mozilla or Firefox name, in the current development environment. And Netscape also has their front end to Gecko.
I see this as a solid foundation effort with interfaces to suit, just like you would hope from the best of open source efforts.
I meet Ray at a conference about five years ago. Seems like a nice enough guy but Groove unless it sold for 1 Billion dollars was a total loss for the investors. A total of 155 Million dollars of VC [vcdeal.com] went into the company. That's Right 155 MILLION. They had a FIFTH round of investment in 2003 of 38 Million. That's an insane amount of capital just to sell to MS for a few hundred million in stock. The investors would have been better off just buying MS stock. I cannot seem to find the terms of the deal online.
The largest portion of that $155 million VC capital is from Microsoft. They owned 40% of Groove already.
The terms of the deal were not disclosed, but one rag said that three exec's but not Ozzie gor $1 million bonuses. All employee incentive deals were cancelled.
Personally, doesn't the ability to cancel employee incentive deals when selling out sort of take the incentive out of it?
However, I am very suprised that he's going to be CTO of MS, he's always been super friendly and encouraging to me even when I was just a peon, it seems like the wrong position for a guy like him. You'd think the MS CTO would need to be a real ballbreaker. I really hope it works out for him.
From articles I read tonight he is just one of three CTO's at Microsoft, one of the other two I heard of, the other one apparently coming from ERP acquisitions.
Ozzie and Gates will be vision makers, not technology ball breakers as far as I can tell. Also Ozzie will remain in the Boston area but spend a lot of time in Redmond on the Senior Advisory Team. He has already been named by Gates in 1994 as one of a handful of Windows Fellows even though he worked for Lotus.
Regarding medical software I find it interesting to consider the degree to which government money is being channeled into Open Source initiatives in healthcare. If this medical Open Source initiative becomes consolidated then commercial companies will have a competing supplier in the form of a coordinated, government-funded organization, possibly multi-national in nature. Open Source would simply become a process used in the collaborative development done by the coalition of government bodies or agents. Medical software companies could face intense competition in the short-term, with economic decline or downfall, although long-term viability of a government funded development is questionable. Does the replacement of private development in medical software systems with a publically-funded loose coalition make sense? What is the accountability? Who is resonsible for safety, privacy, regulatory, and validation etc. Is it an efficient use of public money? Is it sustainable, since the professors, students, researchers, doctors, IT specialists and physicist creators all have other competing objectives? At an extreme, would you trust your life, or health, to unaccountable shareware supplied by enthusiasts? Etc.
And what would happen to medical innovation? Another contributor to/. has already pointed out that it is amazingly expensive to develop medical devices.
Anyway, I'm off to check those games....
Good post. Anything developed with taxpayer funded grants should be public domain except of course anything security sensitive.
On the other hand, something akin to a commercial product should not be commissioned by the government with taxpayer grants along the lines of entreprenurial software development.
But if we do fund software development to help the public good, it should be open source.
I also appreciate the free utilities I have downloaded, and I made my Double Deck Pinochle card game for DOS freeware a long time ago. It makes me feel good that people have used it, and makes me feel that I have contributed as well as borrowed freely.:)
There seem to be two main providers (authors) of Open Source software: volunteers, who contribute for kudos within their "on-line" community and possibly for altruistic purposes; and government-funded workers in universities, research centres, hospitals etc.
A third and major source is failed commercial closed source attempts, especially against a monopoly such as Microsoft. Releasing as open source is about all you can do with it then, and it at least then has a hope against undermining the commercial monopoly.
Unfortunately, this is a model in which to keep making money, you have to keep writing more software.
It's also a model that requires you to find companies that art stupid enough to pay you to write software and then give it away to everyone else free, including their competitors.
But what if I program a closed-source FTP client and give it away. Then let's say I charge for customer service. I haven't made any money with TTs lib. Or have I?
Then you didn't GPL your code and distribute it as they did. If you don't, then it's largely irrelevant how you plan on making money on the closed source program or whether you ever do. The code wasn't GPL'd.
But good theoretical for people to understand dual licensing. I haven't done it, but I've seen a lot of bashing of it by people who apparently don't understand it.
Thought I should add... I wasn't using "monkey" as a derogatory term at all. I literally meant hiring a monkey to do the coding is an improvement over some of the service offerings available.
Then I retract my code monkey post and forward it to those who do refer to me as a code monkey.:)
Well, dual-licensing is definitely confusing. I guess the solution to that problem is to hire lawyers who write less-confusing licenses...
I've always thought the distinction between "non-commercial" and "commercial" purposes was clear enough though. That is, if you intend to use some software for use in a commercial (i.e. for business) manner, then one license applies, but if not, then the other license applies.
So, say you download Debian and KDE and use it on 5 different boxes at home. That'd be a non-commercial use, so it'd be free (in my world anyway).
But if you do the same thing and use Debian and KDE at work, then you would owe TT money, because you're using it for business purposes, i.e., for commerce, i.e., to make a profit.
wow, that is confusing, I certainly hope that it is nothing like that,
As far as I know, it is based totally on additional software development, and whether that software that uses TT libraries is itself GPL'd or not.
If it is GPL'd, no charge for the TT libraries. If not, a commercial TT license is required.
That's all fine and dandy, but why would you write your own software, when you can offer the exact same services for other peoples Open Source software?
Maybe 1% of software is available as open source, and only a minutely fractional percent of potential software.
How would we developers make money writing the other 99.99% of software that people would want to use?
There was no answer here from a developer that made a case for giving a significant software development effort away and make a living off of it.
Being able to hire a monkey to write more/improved code for a product is an improvement over some of the stuff out there now.
Which brings up the other point about the software should be free crowd. They refer to programmers as monkeys. Software should be free because it's written by code monkeys.
Well, this code monkey just decided that people who think that I'm a code monkey won't get any free software from me.
Youi can pay for it, and I don't care what you call me then. But I doubt we'll be having a conversation, as I'm only a code monkey.
What if I'm a developer and I want to make money...?
For most people it's not an option, it's a requirement.
After reading through this thread I've come to the conclusion that there are a few theorists running around pointing at the decades of free work leading to the GNU, Linux, Apache, and the like software infrastructure and citing that as an inevitable replacement for the rest of software.
They point at tools and talk of all software. Those are two different worlds, and as I wrote in another post, the answer is based on proportion of effort to value.
Those of us developers who put in a large proportion of effort to the value of the software created are not likely to be wanting to give it away and hope that someone asks us to modify it. The many disincentives to good software development in that theory have been pointed out well in this thread.
I do advocate licensing the source. It's the way I'm accustomed to working in the real world, modifying commercial source on large computers is what I do for a living, but it's just normal commercial licensing of source as has always been done.
That's a business model for developers that they can at least live on. At least those who put in a large proportion of work to something of value.
38.8 million, on 96.5 billion (with a b) in sales in 2004
what percent is that? %.0402 or.000402 of sales.. less than 1 half of 1 tenth of one percent oh, I'm sorry, that's over four years? about 1 tenth of 1 tenth of one percent of sales
Is that fathomable? I laud IBM for it's participation in FOSS, but- it's not even a drop in IBM's revenues...
And that doesn't even take into account that IBM claims to have spent $1 billion dollars on Linux, part of that being the basis of the braindead SCO lawsuit against them.
On the other hand, they are investing for the future, Linux runs in virtual partitions on their mainframe OS's, yada yada.
Anyway, you're right that people are not used to buying software this way, but that doesn't mean it can't work. (Actually, I think it is almost inevitable, once a larger and larger fraction of software becomes Free.)
Customer one would have to be Santa Claus, and the developer clueless. And how will this larger and larger percentage of software become free?
I often wonder how many proponents of Open Source software ARE actually small business owners, or indeed programmers or programming consultants who are actually in the industry right now, and NOT just some bored college kid who really doesn't have to do real work to eat for a living. I have real bills to pay, and a business to keep running, and competitors. It's all well and good to talk about the idyllic benefits of free software, share it all, publish my code, but for a small company, the service model isn't really an option, if you want to stay viable.
Not many from the posts I'm reading. Post after post that says software should be free so I can charge for admining Linux, or something to that effect.
So open source is no different from closed source software in its role in making money.
I make a living modifying source for ERP's on the AS/400. ERP's are expensive software, of course, and I have never heard anyone refer to them as Open Source.
I happen to agree with you that including the source is essential to a good relationship with the customer, but the license would need to be commercial oriented, that is, you have licensed the source, you can do what you want with it but you can't resell it or transfer it.
Otherwise the source could be posted on the internet and given away or resold many times. That bears no resemblance to anything I've ever seen here about Open Source.
If it's just that source code be made available, then in my experience historically large, expensive business system software has come with source and this whole Open Source thing would be pretty much limited to an anti-PC thing, as PC software vendors rarely release source. A notable exception is the Delphi developer community.
Prior to the posts about selling source was the more typical/. response of giving source away and hoping someone will pay you to modify it.
Large source systems can be pointed to that are this way, but equally large numbers of developers participated in creating it, or commercial success was hopeless and it was given away for hoped for market share.
In any event, what I see here is proportionality of effort to the value. A great deal of effort could be done by a few and given away for fame, but those who have expended a great deal of effort to create a system of value had to live off something while doing it, and if they can license it they become commercial developers.
If they can't, sure they will Open Source it and see if it takes off, they have nothing else to do with it. But the whole Richard Stallman free software thing is the perception of Open Source, when the only thing that makes economic sense to me for a significant effort of value is commercial licensed source.
Nearly any name that will be popped up in response to that will have been people that got something started and then was paid by commercial companies for strategic reasons to continue their work.
There's a lot of people who have donated a lot of great work in software, but they surely didn't do it for the money.
But yes, ideally one would want an editing front-end on top of any data representation.
Below that post I suggest something along your lines of xml text, but in a central xml repository with history and versioning to take advantage of the richer xml, logged out by the editing app as it edits the appropriate config files.
Well... changing the gconf config in/etc is really out of the question, since I'm just a user on the system. Is this doable in a way, that allows config changes to be explicitly committed to ~/.gconf on demand (I do occasionally want to change stuff), and does not require root?
A program, even a friendly interface program, shouldn't be a bypass of security. If one (and in my line of work on the AS/400 this includes me in most places) can't do it at the command line, then you shouldn't be able to do it from a much nicer interface to that command line.:)
Not that the AS/400 has a much nicer interface or that I have a job, but still...
Slashdot Mirror
Rumors are that Transdyn are negotiating to get back into writing the code for Big Dig. Hopefully they will have better luck the second time around. I'm sure there are lots of helpful comments in the current source revs in the ITS software for whoever develops it (particularly Transdyn :P)
Holy cow, what a story. I can't believe Transdyn would be let anywhere near the project again after what's happened.
On the other hand, for all the posters who say that the source code for all taxpayer paid for software should be required to be opened, Transdyn was a commercial software product modified for this project.
Governments also buy Oracle, SAP, Windows, and every other conceivable commercial software product. Governments are just another customer, and tax dollars just those customers way of paying for it. And open source just becomes a hopeless mantra then.
Of more realistic hope would be requirements by law to consider open source first to fill requirements, and for taxpayer funded work to be done by those taxpayers.
Even those common sense requirements are fought against with all the money it takes from Microsoft and outsourcers.
rd
In anticipation of one concern I saw raised repeatedly, that is the concern of memory usage from multiple instances of the Gecko engine in each Mozilla component versus apparently one being shared in Mozilla Suite.
.exe's on disk, so if disk space were a concern then there'd be a legitimate concern. We used to have to worry about disk space too, but that is something that really has reached infinity, so no problem there.
These concerns are likely to have been raised by non-programmers with it being seemingly sort of obvious that duplicated code is unnecessary bloat, or perhaps from loading and looking at memory resources and seeing more taken.
An operating system only pages into memory from disk that which is being used. We would have been in big trouble throughout the history of computing if memory had to hold the whole program, although nowadays such huge amounts of memory are required that people may think that.
It is more the case of large amounts of memory being allocated by programs and the operating system for data, not code. We used to have to work with small data buffers too, but with hardware now allowing nearly infinite amounts of memory at a shot to be accessed by programmers, ever approaching infinity amounts of memory are being accessed by programmers. It's painless to them now, and makes for much easier programming.
But in reality, only the code being executed is pulled into memory by the operating system as its needed. The redundant Gecko engine code will sit in multiple
One may then point out that more memory has been observed to have been acquired, and the multiple Gecko thing is a pig. Again, this is more data than code. Even with a single instance of code, if you use it for multiple things it's going to grab multiple sections of memory for each use.
So in reality, in theory, because I haven't tried this at home, there's really not that much difference between a single Gecko engine grabbing two memory areas for FF and TB, or two Gecko engines grabbing an area apiece from FF and TB. It's mostly data work space needed to do the job on a web page and email.
And not only is there really no memory difference, but there should be faster performance and less possibility of internal data conflicts with two clean instances of Gecko working FF and TB versus a shared instance, not that there is any conflict in working, debugged code. Someone may very well say that a bunch of Gecko code it is spawned off for each use, and then it becomes clearer that it really doesn't matter if it starts off as two or is spawned off a couple of times from one.
And lastly, the Firefox Classic distro will be able to take advantage of any Mozilla FF, TB, etc. component integration with the Gecko engine that will conditionally compile and bring in focused Gecko code for the component at hand, versus one monolithic engine shared by all. Whether done or not now, there is certainly potential for focused integration with each component that Firefox Classic would inherit on each release.
So the response that two Gecko engines is a liability is that, actually, no, it's a plus.
rd
I thought I was done commenting but I came back to add this. From the two slashdot threads I've read in the last few days and all the links to Mozilla dialogue, I thought Mozilla/Firefox defenders did a good job explaining how to soup up Firefox to do the things being complained about as missing, sometimes with a config text file option, sometimes an existing plugin, and sometimes just a setting from an FF menu.
What little that couldn't be done was attributed to Firefox UI decisionmaking frustrating those who wanted to add professional features, along with some dislike for people having to piece together all these tweaks and plugins and redo it on every update, and even then not having some patches in there that they would want.
So I posted the blurb at bottom here this afternoon, but I wanted to add another thought. The distro would have -
all the config settings, UI changes such as pulling the little search field off the UI, classic skin, etc., and everything else described as how to implement what was wanted,
with all plugins installed and configured that add the Mozilla Suite functionality and usability desired,
and all the patches implemented to this branch that FF WONTFIX to add all the advanced keyboard behavior and such for desired Suite functionality. My understanding from reading is that most of those patches are already done or not that hard but just refused to be implemented in FF.
This would be a CVS branch of FF, TB, the new composer, etc. where only these functionality patches have to be reapplied to new releases. Most seemed to be triggered by keyboard shortcuts or additional menu options so seem fairly external and addon to core code.
In addition, with a classic skinned, UI tweaked FF and TB then create a shell menu control that FF, TB, and other Mozilla apps plug into, who knows, maybe a defined plugin interface that allows Suite users to plug other interoperable browser related apps made available for it by third parties. At that point you should have essentially classic Mozilla Suite (I'm a real classic Netscape 7.02/Mozilla 5.0 user, I'll have to be dragged kicking and screaming from it) with only an outer menu shell and mostly keyboard triggered functionality tweaks to merge to current Mozilla components FF and TB.
That's a distro that should be able to ship with each new release. However, in my opinion if that is done by Mozilla Suite advocates my opinion is that Mozilla Foundation should put their stamp of approval on that and call it Firefox Classic.
I would even upgrade off of 7.02 for that.
rd
My previous post:
"I would say after reading all this that it possibly makes more sense to have distro of FF, TB, etc. with extensions that do such things as provide a GUI options panel and every other thing that brings FF up to Navigator standards, even patches that FF WONTFIX like CTL-ENTER preference to match Navigator.
They should bundle the distro as the equivalent of Mozilla Suite (with a different name) with all extensions included and tweaked, which is the whole point of a distro.
This makes much more sense to me than the more massive effort required to maintain and enhance SeaMonkey."
I would say after reading all this that it possibly makes more sense to have distro of FF, TB, etc. with extensions that do such things as provide a GUI options panel and every other thing that brings FF up to Navigator standards, even patches that FF WONTFIX like CTL-ENTER preference to match Navigator.
They should bundle the distro as the equivalent of Mozilla Suite (with a different name) with all extensions included and tweaked, which is the whole point of a distro.
This makes much more sense to me than the more massive effort required to maintain and enhance SeaMonkey.
rd
As I specified in my post, the confusing announcements are a fumble. The software is good, and the strategy seems sound for producing more. You and I are unusual, because we're technologists, in deciphering the announcements to mean anything but "we quit".
yes, you're right, Doc, your analysis was right on. It looks to me like it was actually lack of announcements and more what the Mozilla Suite developers posted that got the bad publicity.
I think the actual Mozilla announcement at this point should only be encouraging to those 25 million Firefox and counting downloaders, but the handling of all this was fumbled badly.
I hope they do the things you recommend in grandparent and make lemonade out of this lemon.
rd
But announcing the transformation in terms of the demise of the organization, and "I'm sorry there will be no next version", is a total fumble.
Actually, rather than slow down Firefox downloads, it's discontinuing an older suite because of the success of Firefox downloads.
That is emphasis on and encouragement of continued said downloads.
And those who support the older suite are free to continue development on a new branch frontending Gecko page rendering without the Mozilla or Firefox name, in the current development environment. And Netscape also has their front end to Gecko.
I see this as a solid foundation effort with interfaces to suit, just like you would hope from the best of open source efforts.
rd
I meet Ray at a conference about five years ago. Seems like a nice enough guy but Groove unless it sold for 1 Billion dollars was a total loss for the investors. A total of 155 Million dollars of VC [vcdeal.com] went into the company. That's Right 155 MILLION. They had a FIFTH round of investment in 2003 of 38 Million. That's an insane amount of capital just to sell to MS for a few hundred million in stock. The investors would have been better off just buying MS stock. I cannot seem to find the terms of the deal online.
The largest portion of that $155 million VC capital is from Microsoft. They owned 40% of Groove already.
The terms of the deal were not disclosed, but one rag said that three exec's but not Ozzie gor $1 million bonuses. All employee incentive deals were cancelled.
Personally, doesn't the ability to cancel employee incentive deals when selling out sort of take the incentive out of it?
rd
However, I am very suprised that he's going to be CTO of MS, he's always been super friendly and encouraging to me even when I was just a peon, it seems like the wrong position for a guy like him. You'd think the MS CTO would need to be a real ballbreaker. I really hope it works out for him.
From articles I read tonight he is just one of three CTO's at Microsoft, one of the other two I heard of, the other one apparently coming from ERP acquisitions.
Ozzie and Gates will be vision makers, not technology ball breakers as far as I can tell. Also Ozzie will remain in the Boston area but spend a lot of time in Redmond on the Senior Advisory Team.
He has already been named by Gates in 1994 as one of a handful of Windows Fellows even though he worked for Lotus.
rd
Regarding medical software I find it interesting to consider the degree to which government money is being channeled into Open Source initiatives in healthcare. If this medical Open Source initiative becomes consolidated then commercial companies will have a competing supplier in the form of a coordinated, government-funded organization, possibly multi-national in nature. Open Source would simply become a process used in the collaborative development done by the coalition of government bodies or agents. Medical software companies could face intense competition in the short-term, with economic decline or downfall, although long-term viability of a government funded development is questionable. Does the replacement of private development in medical software systems with a publically-funded loose coalition make sense? What is the accountability? Who is resonsible for safety, privacy, regulatory, and validation etc. Is it an efficient use of public money? Is it sustainable, since the professors, students, researchers, doctors, IT specialists and physicist creators all have other competing objectives? At an extreme, would you trust your life, or health, to unaccountable shareware supplied by enthusiasts? Etc.
/. has already pointed out that it is amazingly expensive to develop medical devices.
:)
And what would happen to medical innovation? Another contributor to
Anyway, I'm off to check those games....
Good post. Anything developed with taxpayer funded grants should be public domain except of course anything security sensitive.
On the other hand, something akin to a commercial product should not be commissioned by the government with taxpayer grants along the lines of entreprenurial software development.
But if we do fund software development to help the public good, it should be open source.
I also appreciate the free utilities I have downloaded, and I made my Double Deck Pinochle card game for DOS freeware a long time ago. It makes me feel good that people have used it, and makes me feel that I have contributed as well as borrowed freely.
Many thoughtful questions in your post.
rd
There seem to be two main providers (authors) of Open Source software: volunteers, who contribute for kudos within their "on-line" community and possibly for altruistic purposes; and government-funded workers in universities, research centres, hospitals etc.
A third and major source is failed commercial closed source attempts, especially against a monopoly such as Microsoft. Releasing as open source is about all you can do with it then, and it at least then has a hope against undermining the commercial monopoly.
rd
Unfortunately, this is a model in which to keep making money, you have to keep writing more software.
It's also a model that requires you to find companies that art stupid enough to pay you to write software and then give it away to everyone else free, including their competitors.
Companies don't stay in business being stupid.
rd
But what if I program a closed-source FTP client and give it away. Then let's say I charge for customer service. I haven't made any money with TTs lib. Or have I?
Then you didn't GPL your code and distribute it as they did. If you don't, then it's largely irrelevant how you plan on making money on the closed source program or whether you ever do. The code wasn't GPL'd.
But good theoretical for people to understand dual licensing. I haven't done it, but I've seen a lot of bashing of it by people who apparently don't understand it.
rd
Thought I should add... I wasn't using "monkey" as a derogatory term at all. I literally meant hiring a monkey to do the coding is an improvement over some of the service offerings available.
:)
Then I retract my code monkey post and forward it to those who do refer to me as a code monkey.
rd
Well, dual-licensing is definitely confusing. I guess the solution to that problem is to hire lawyers who write less-confusing licenses...
I've always thought the distinction between "non-commercial" and "commercial" purposes was clear enough though. That is, if you intend to use some software for use in a commercial (i.e. for business) manner, then one license applies, but if not, then the other license applies.
So, say you download Debian and KDE and use it on 5 different boxes at home. That'd be a non-commercial use, so it'd be free (in my world anyway).
But if you do the same thing and use Debian and KDE at work, then you would owe TT money, because you're using it for business purposes, i.e., for commerce, i.e., to make a profit.
wow, that is confusing, I certainly hope that it is nothing like that,
As far as I know, it is based totally on additional software development, and whether that software that uses TT libraries is itself GPL'd or not.
If it is GPL'd, no charge for the TT libraries. If not, a commercial TT license is required.
If it isn't that way, I would be surprised.
rd
That's all fine and dandy, but why would you write your own software, when you can offer the exact same services for other peoples Open Source software?
Maybe 1% of software is available as open source, and only a minutely fractional percent of potential software.
How would we developers make money writing the other 99.99% of software that people would want to use?
There was no answer here from a developer that made a case for giving a significant software development effort away and make a living off of it.
rd
Being able to hire a monkey to write more/improved code for a product is an improvement over some of the stuff out there now.
Which brings up the other point about the software should be free crowd. They refer to programmers as monkeys. Software should be free because it's written by code monkeys.
Well, this code monkey just decided that people who think that I'm a code monkey won't get any free software from me.
Youi can pay for it, and I don't care what you call me then. But I doubt we'll be having a conversation, as I'm only a code monkey.
rd
What if I'm a developer and I want to make money...?
For most people it's not an option, it's a requirement.
After reading through this thread I've come to the conclusion that there are a few theorists running around pointing at the decades of free work leading to the GNU, Linux, Apache, and the like software infrastructure and citing that as an inevitable replacement for the rest of software.
They point at tools and talk of all software. Those are two different worlds, and as I wrote in another post, the answer is based on proportion of effort to value.
Those of us developers who put in a large proportion of effort to the value of the software created are not likely to be wanting to give it away and hope that someone asks us to modify it. The many disincentives to good software development in that theory have been pointed out well in this thread.
I do advocate licensing the source. It's the way I'm accustomed to working in the real world, modifying commercial source on large computers is what I do for a living, but it's just normal commercial licensing of source as has always been done.
That's a business model for developers that they can at least live on. At least those who put in a large proportion of work to something of value.
rd
38.8 million, on 96.5 billion
.000402 of sales..
(with a b) in sales in 2004
what percent is that?
%.0402
or
less than 1 half of 1 tenth of one percent
oh, I'm sorry, that's over four years?
about 1 tenth of 1 tenth of one percent of sales
Is that fathomable? I laud IBM for it's participation in FOSS, but- it's not even a drop in IBM's revenues...
And that doesn't even take into account that IBM claims to have spent $1 billion dollars on Linux, part of that being the basis of the braindead SCO lawsuit against them.
On the other hand, they are investing for the future, Linux runs in virtual partitions on their mainframe OS's, yada yada.
It's not even a drop in their investment.
rd
Anyway, you're right that people are not used to buying software this way, but that doesn't mean it can't work. (Actually, I think it is almost inevitable, once a larger and larger fraction of software becomes Free.)
Customer one would have to be Santa Claus, and the developer clueless. And how will this larger and larger percentage of software become free?
rd
Otherwise, you probably can't make much as a normal commercial software inventor, which is increasingly seen as an anachronism from the last century.
/. as if it's true, but I've seen no basis for it.
This keeps getting repeated on
rd
I often wonder how many proponents of Open Source software ARE actually small business owners, or indeed programmers or programming consultants who are actually in the industry right now, and NOT just some bored college kid who really doesn't have to do real work to eat for a living. I have real bills to pay, and a business to keep running, and competitors. It's all well and good to talk about the idyllic benefits of free software, share it all, publish my code, but for a small company, the service model isn't really an option, if you want to stay viable.
Not many from the posts I'm reading. Post after post that says software should be free so I can charge for admining Linux, or something to that effect.
rd
So open source is no different from closed source software in its role in making money.
/. response of giving source away and hoping someone will pay you to modify it.
I make a living modifying source for ERP's on the AS/400. ERP's are expensive software, of course, and I have never heard anyone refer to them as Open Source.
I happen to agree with you that including the source is essential to a good relationship with the customer, but the license would need to be commercial oriented, that is, you have licensed the source, you can do what you want with it but you can't resell it or transfer it.
Otherwise the source could be posted on the internet and given away or resold many times. That bears no resemblance to anything I've ever seen here about Open Source.
If it's just that source code be made available, then in my experience historically large, expensive business system software has come with source and this whole Open Source thing would be pretty much limited to an anti-PC thing, as PC software vendors rarely release source. A notable exception is the Delphi developer community.
Prior to the posts about selling source was the more typical
Large source systems can be pointed to that are this way, but equally large numbers of developers participated in creating it, or commercial success was hopeless and it was given away for hoped for market share.
In any event, what I see here is proportionality of effort to the value. A great deal of effort could be done by a few and given away for fame, but those who have expended a great deal of effort to create a system of value had to live off something while doing it, and if they can license it they become commercial developers.
If they can't, sure they will Open Source it and see if it takes off, they have nothing else to do with it. But the whole Richard Stallman free software thing is the perception of Open Source, when the only thing that makes economic sense to me for a significant effort of value is commercial licensed source.
Nearly any name that will be popped up in response to that will have been people that got something started and then was paid by commercial companies for strategic reasons to continue their work.
There's a lot of people who have donated a lot of great work in software, but they surely didn't do it for the money.
rd
yeah, I saw your other post on checking if it does it the right way, etc., and understand better now. thanks!
rd
But yes, ideally one would want an editing front-end on top of any data representation.
Below that post I suggest something along your lines of xml text, but in a central xml repository with history and versioning to take advantage of the richer xml, logged out by the editing app as it edits the appropriate config files.
cheers
Well... changing the gconf config in /etc is really out of the question, since I'm just a user on the system. Is this doable in a way, that allows config changes to be explicitly committed to ~/.gconf on demand (I do occasionally want to change stuff), and does not require root?
:)
A program, even a friendly interface program, shouldn't be a bypass of security. If one (and in my line of work on the AS/400 this includes me in most places) can't do it at the command line, then you shouldn't be able to do it from a much nicer interface to that command line.
Not that the AS/400 has a much nicer interface or that I have a job, but still...
rd