Slashdot Mirror
GRASS Geographic Information System now under GPL
Spatialy Challenged writes "The GRASS Geographic Information System (originally developed by the US Army Corps of Engineers) is actively being developed and has now been released under the GPL. GRASS has a good core architecture, but is missing the interoperability and GUI features of commercial desktop GIS. I would sure like to see this software evolve into a KDE/Gnome GUI plus OpenGIS CORBA/SQL/COM interoperability. I'm sure it has the potential to blow the socks off of the big commercial names. "
Anyone her know about important lots of data from arcview? I've got about 15 gigs of coverages nad 60gigs of 1/4ft pixel I want to use in grass.
I notice that the software is gpl and such. But what about the datafiles? Do we still have to pay 500 dollars to get the data? Or is that going to go up on a FTP site somewhere?
Jeff Knox
GRASS isn't really competing against ARC/View it's market is the same as ARC/INFOs. Having used ESRI products I can say it would be nice to have another option. ESRIs bug fixes are slower than MS and often the bugs are real show stoppers. (Metric units anyone? :( )
I'm presently looking into using GRASS for a project and if it has the functionality I require it should save me 10's of thousands of dollars in licenceing fees. (Both ESRI and OS -Soloris or TRU64- because ARC/INFO will not run on linux.)
Oh come now. It's not flamebair. It's a joke. A joke!
Have people no senses of humor!?
--
Max V.
There should be a moratorium on the use of the apostrophe.
Max V.
NeXTMail/MIME Mail welcome
When I was working for the VA, we released all the source code used (except for the encryption ones) on CD-ROM for anyone to use. We were required to sell it for the cost of duplicating the information. This wound up being $10-15, since it was about 5 years ago, and CD duplication was expensive.
-- Ever notice that fast-burning fuse looks exactly the same as slow-burning fuse? I didn't... (Edgar Montrose)
You've funded a great deal of research and development to make this software component and all of a sudden, the government releases a similar component and allows anyone to make a product that competes with yours at a much lower cost than you had to expend. A good question, in general terms. However in this particular case, there is no "all of a sudden", GRASS has been around, and public, since 1982. Does anybody here know if any of the current big GIS players have used any of the PD GRASS code?
The other thing that would be nice is if the government would start making the datasets we paid good tax money for available for reasonable fees.
Agreed. A lot of other people think the same way. New Zealand just drastically reduced their digital base date prices: from $2 million to $1,500 for the nation wide set. While $1500 is not pocket change for a student or small company, at least it's reacheable. See this url for details.
A group of us are working on the Canadian government to follow suit. There's a petition at: http://members.home.net/freedata/. For some press see this article.
cheers,
-matt
Grass Documentation Project: /
w .pdf
.e00 to GRASS: . html
http://www.geog.uni-hannover.de/grass/gdp
From Grass to ArcView:
http://www.geo g.uni-hannover.de/grass/gdp/tutorial/grass_arcvie
From
http://www.geo g.uni-hannover.de/grass/projects/m.in.e00/welcome
I cross posted your message to the GRASS mailing list. [Majordomo@cecer.army.mil]
I don't know what scripting facility GRASS has...
;-) GRASS follows the unix paradigm of being modular and command line oriented, so you can use perl, or python, or tcl, or whatever you like.
Any kind you want.
Yes, from GRASS 5.x onwards the "auto-conf" is implemented. configure make install That's all to compile the GRASS GIS. Cheers Markus
Use python with COM and it get's better. --Bernhard Reiter
However, in general, the work being done under contract to the government is never actually placed in the public domain. In general, the rights to the work belong to the contractor and the government is only granted a license to use the work.
Rather, I think the real reason government works aren't copyrighted is that there's no reason for them to be. The purpose of copyrights is to promote the development of valuable works for public consumption, by giving the creators the oportunity to profit from selling copies of the works, and thus an incentive to develop them. Works developed by the government are already developed specifically for the public good (at least in theory), so there's no need for an additional profit motive.
Of course, that doesn't explain government-owned patents.
I think you're confused here. If something is in the public domain, it cannot disappear from the public domain unless all copies disappear. Furthermore, government agencies are required to make their (non-classified, PD) products available to the public at a reasonable fee (to cover costs), so just because government-developed software is released under the GPL doesn't mean you can't obtain a PD version.However, anyone can take that PD software, make some trivial changes, copyright it and release the result as proprietary, GPL, or whatever.
I actually think in many ways the GPL serves the public better than releasing it PD. Look at the TIGER/LINE data that Bruce Parens released under the GPL. Before that, there were dozens and dozens of companies taking that data (acquired at considerable public expense), making proprietary modifications and reselling it. All perfectly legitimate, but also wasteful, since anyone who wanted commercial-quality maps that weren't subject to someone else's copyright would have to go back to the PD version and duplicate the work that has been done dozens of times before. On the other hand, any improvements made the the GPL'ed version will be free to everyone, so no one needs to reinvent the wheel.
Strangely, I found out about this yesterday, when I followed the lwn link to the FreeGIS web site newly created by Jan-Oliver Wagner in Germany.
This site has links to a number of other interesting free software GIS packages, as well as a couple of sources of data.
It is my hope that a real free GIS community will develop. I have a personal interest in this, as I think my libart 2D graphics rendering library has the potential to render maps at a much higher quality than most proprietary GIS packages today (i.e. antialiasing, semi-transparent layers, combining vector with image data). If there's anyone who's interested in integrating libart's cool rendering capabilities with the cool free software GIS apps, both current and future, please get in touch.
LILO boot: linux init=/usr/bin/emacs
The http://www.usgs.gov site has a spatial data link that will send you on your way to data.
I've worked with salmon population and fishery models for a number of years. This past year I was instructed to create "spatially explicit" population viability models using GIS. I checked out the ESRI stuff and another high end package. I eventually selected GRASS however. The big commercial packages offer lots of pretty features but provide lame extension languages. I've seen people hit walls with these systems and end up stuck. I use Grass to dump spatial data and process it using a combination of C++ and Python. This combination is proving very effective.
We use to routinely acquire various CDROMs from the Defense Mapping Agency for a small fee. You can also obtain data from the USGS and numerious other government agencies.
Well, niche markets. GPL will contribute to bugs
;) Back in 1991
being fixed, but better development of these big
projects... I'm sceptical.
I'm scared since the day I looked at the Mozilla source.
Sector sliding: usually all sectors of a track
are aligned, which means if the drive controller
has read one track, software has read the data,
and the next track needs to be read, you end up
waiting for close to a full revolution's time
until the head is back over sector 1 (assuming
full track reads). That wastes time.
What you want is to have sector 1 of track 1
start at say sector 3's position (sector 3 of
track 0), so the head will be just ahead of
sector 1. When the software has finished copying
data off of the drive controller, and wants to
read the next track, it can do that with minimal
time wasted waiting for sector 1 to show up
under the head.
To save time trying different offset values for
your machine, you feed it a blank floppy and it
will determine the best offset while you go
sip some cappucino or something.
it used to save some time, now it would appear
floppy disks just plain suck. I clone machines
El Torito style now, if I have to.
For more GIS checkout
http://www.remotesensing.org
http://www.fsf.org/software/guile/guile.html
Is fairly up-todate.
Bernhard Reiter
You young whippersnappers, complaining about gettin' GPL'ed software... Why, back in my day, if we wanted a GPL'ed version of a product, we had to write it ourselves! We bootstrapped Linux on our PCs, and we liked it! We had ta write the dang-nabbed code for emacs in vi, and we liked it! It took four days for each build attempt of cross-target gcc on a Sun, and we liked it! You damn whippersnappers complainin' about gettin' your source code fer free sure makes me hot unner the collar. Damn kids today, got no respect, they don't know the value of what they got because they never had to work fer a thing.
Beer recipe: free! #Source
Cold pints: $2 #Product
The deal with a multi-person cooperative effort is that no one person holds the copyright any longer.
Linus essentially no longer holds the copyright to Linux fully because much of the work has been contributed by others.
In order to change the license, one would have to notify and obtain approval from all copyright holders, or remove their code from the project.
I think it's horribly unethical, because when I contribute code to a project it is done under the rules defined by the license the project exists in at the date of my submission.
Just what is the legality of releasing something that was developed with public money under something like the GPL? Any comments from lawyers or others who might know?
Further evidence that the GNU folksa are working on creating an open army to take over the world by force. Revolutionary socialism, and all that rot.
--
Max V.
There should be a moratorium on the use of the apostrophe.
Max V.
NeXTMail/MIME Mail welcome
If it was developed with public money, it's a pretty good idea to let the public have access to it, especially under the GPL. This basically ensures that it'll be 100% public.
Now if they were to have deicded to charge $10,000 licensing fees (IE: License for Profit), and viciously go after anyone who used it without paying it, then I'd be questioning the legality.
I think you're underestimating unix people's interest in GIS. Don't forget that most software written for unix compiles on linux, BSD, slowlaris, etc - including the majority of boxen people use for scientific / military / engineering applications. Thus even niche interest apps tend to get developed quite vigorously - eg. FEA / CFD codes are often open-source these days, Varkon parametric 3D cad software is going open source (sloowly), Maverik virtual environment software is GPLd, the eros os has been GPLd. All these are large, serious applications, with limited markets. Nonetheless, they stand to benefit considerably from the opened development model.
From what I can see, most places that get GIS's end up either pouring tons of resources into them for doubtful return, or end up occaisionally playing with them, producing one or two interesting maps and then falling by the wayside.
The key I think to give the average user the ability to use spatial analysis is to develop custom built applications that support specific tasks and analyses. I'm doing that right now in the public health sector. The problem is the licesnsing is a bear. The vendors don't want it to be too easy to develop applications with GIS functionality because it affects there bread and butter business. One vendor requires you buy licenses in blocks of 40, for example.
The other thing that would be nice is if the government would start making the datasets we paid good tax money for available for reasonable fees. The fact that people take them and simply resell them at lower prices (which presumably is fair market price) means that the government's revenues are not maximized for these resources.
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
I currently work at CERL, where GRASS was originally developed and there are still a number of hard core GRASS fans around in the GIS community. I've seen an app developed in tcl/tk/GRASS and redeveloped in ESRI's ARC/INFO/Visual Basic and beyond VB database capability (which the developer found useless anyway) it looked isn't much you can't do in GRASS that you can do in the industry standard products.
10 years ago GRASS had the capabilities that ESRI is just implementing into ARC 8. All GRASS needs is the GUI and it should be a big competitor. I wish I had more time... I'd love to be part of that!
Bishop
-=================================-
"Computers are mighter than the pen, sword and usually the programmer."
what's wrong with Tcl/Tk ?
If I recall, it used to say it was public domain software, because it'd been developed with government funds. I haven't played with it since the days of DOS, though; I could very well be mistaken. I remember it being somewhat hard to use, but quite featureful.
I'm mildly intrigued. Could you explain?
--
Fuck the system? Nah, you might catch something.
I think they had a long article in Linux Journal about using GRASS under Linux instead of the commercial alternatives, and even in the state that it is in currently, it saved them a lot of money and allowed them to use scripts and tools like Tk and others to be able to automate things to allow them to process a lot more data a lot more quickly.
The project was completed tens of thousands of dollars under budget and they experienced none of the problems they used to see with other systems.
This is a good thing. They spoke very highly of GRASS and its potential. I'm not in the field at all, but this is another victory for free software.
It's the *chic* thing to do nowadays if you've got something of value:
It's really nothing more than a PR thing anymore (in my opinion). Nowadays, if you've got a nice lil' linux app, you've really got only 1 option: GPL it. Anything less would be uncivilized and most likely heavily criticized.
--Nsfmc
This story reminds me a little bit about the French senators who want computer programming done for France to be open sourced -- it demonstrates the value that can be salvaged from tax money that's already been spent.
Is there a good reason that it could not simply be a standard clause in the contracts surrounding comptuer programming done for any government agency that the result must be reusable, barring previous conflicting licensing terms? There are all sorts of other standards imposed on nearly every government contract, and this is one that might actually add some value.
Remember, the only way the government buys something is with money it's already taken from you for your benefit, or with money it promises it will take from you later. (Also for your benefit.)
cheers,
timothy
jrnl: http://tinyurl.com/c2l8yr / foes: http://tinyurl.com/ckjno5
It depends on whether the licence terms permit changing the licence. True public domain code or a typical BSD licence does allow incorporating the code into a produce under a different licence. Anyone using a licence that permits this is presumably willing for people to do that.
GPL does not permit this, that's the fundamental difference between GPL and BSD type licences.
Grass does have vector support. It just needs work. I would bet that somebody could patch the existing vector support and make it rock.
Umm, is freeware not a contraction of Free Software? And GPL is the license for free software, correct? So how is freeware incorrect? Sure, it would be horrific to call GPL software public domain, or shareware, or something, but whats wrong with calling it freeware? (other than it sounding lame as hell)
Hamsters are at least as feathery as penguins. HamLix
How could any sensible person on this world be so stupid not to agree with the GPL? As far as I understand it only requires one to include everything (including the source code, licencing agreement etc.) when redistributing. A non-developer user has no restrictions at all, he doesn't evn have to pay. I like that much better than the stupid EUL of M$ (and others) which also doesn't take any product responsebility, but let you even pay for this product.
I know this is such a "niche" thing, but personally, I would *love* to have a CD of GIS data for my area that I could use along with a GPS and my laptop to track some of the data I "create" from my hobby as an amateur botanist. i.e. being able to do things like record GPS locations of plant ranges for certain species I've been able to find and recording and accurately mapping locations of sample collection sites and so forth. As it is, if I do record them, I have to wind up plotting them on a paper map of the "Rand McNally Road Atlas" type of thing. It would be nice to be able to regenerate new maps electronically any time I wanted.
-=-=-=-=-
-=-=-=-=-
My mom's going to kick you in the face!
Maybe now they'll use autoconf and automake. GRASS was a nightmare to install when I did it for my former boss.
Remember that what's inside of you doesn't matter because nobody can see it.
Revolutionary? Certainly. Socialism? Hmmmmm, more like anarchy.
Yup. Those have been available for some time from Perens' FTP site and a number of mirrors -- or (if you've got a Federal Repository Library handy, which my university does) you can just get the unencumbered originals.
I'm not familiar enough with GRASS to say whether it'll work with that data, but if you want to know what you can do with your laptop, head over to Bruce's website (perens.com, is it?) and join the mailing list for software using the data.
I'm sorry, but having used GRASS and having spent 4 years working with Arc/Info.
No comparison... GRASS does not have a chance.
Yes, GRASS is fairly popular with the government and university segments because the licensing for Arc/Info is rather expensive. But Arc does so much more...
It's funny. This mediocre product mentions that it's now under the GPL and slashdot goes hyper for it.
"Copyright protection under this title is not available for any work of the United States Government, ..."
The feds can own copyrights only by transfer.
Too often, they get around this by hiring some company to do the work and let them profit off the copyrights in addition to paying them for the work. What a deal!
The idea behind the law should be obvious: stuff developed with public money should be free for unrestricted use by the public who paid for it. Certainly not just for the GPL mob who have paid a minuscule fraction of the taxes that pay for the stuff.
I hate to see this stuff disappearing from the public domain into the chains of the GPL. It's often marked "GPL" at its first publishing even though that has no legal standing. This unethical action usually works, though because nobody but those around the jerk that does it knows the stuff is in the public domain. Even someone knows the stuff is being mislabled, there is little that can be done because it is legal.
I think you'll find examples of this in the Linux kernel, some of which seems to have been developed by people on Federal payrolls and/or Federal computers.
What's wrong with it? Mainly because the leader of the FSF doesn't like it. He didn't coin it. Freeware was a term in existance long before the GPL was created. In another of the many redefinitions of common words (like free, proprietary, public, open) for Gnu-speak, it means only software that can executed without paying a license fee but without source code which may be modified under the GPL.
The theory is that since the data in question is useful to such a small percentage of the overall US population, charging just those who care about the information for the data (and transfer) costs is perfectly ok. No need for every person in America to subsidize the needs of 1% of its citizens.
That being said, a fair amount of data is available on the net at no cost. The USGS elevation data for the US is downloadable from the USGS. One company will package up to 20 MB of the US Census Bureau Tiger Line data sets for download as one zip file.
How's my programming? Call 1-800-DEV-NULL
NOAA Spends millions of your dollars to collect data for Nautical charts, then goes out and picks a single vendor to distribute them at $500 for a single CD (which only covers a single region).
For more info see www.bsbcharts.com. Every bit of data they publish and sell was collected by the U.S. Gov.....
Guess the same was done with landsat imagery, public spends millions on sattelites, the reagan picked a single company to distribute and sell for whatever the market would bear...
I've been playing with GRASS since last year. It's good to see GIS get coverage on slash, even though the change to GPL is not a big deal for GRASS, since it was free and public domain... I'd like to see the source code of the GIS's that owe their existance to GRASS.... Build your own GIS ... Friends don't let friends buy ESRI.
and i thought it was a good idea done bad (but this was the first time i used any GIS software or even heard about such a beast).
Why i think this was done bad may be because of my lack of experience and my lack of training on it. basically, all I learned was via the web or books I searched at the library.
While it's a great way to learn computer science for me (because i already have enough knowledge to understand new things), this wasn't the most effective thing to learn ArcView or GIS philosophy in general.
The scripting language included inside ArcView looked great on paper, it was easy to use and it even was object oriented, but it seemed to be designed like a Mac: it is easy for users knowing nothing to computer but when you know about computers and want to do real programming you're stuck.
Of course this may come from lack of information, but the simple fact that I couldn't find this information was very frustrating.
I hope some
I don't know what scripting facility GRASS has but one thing that would be cool would be to see someone add a Guile interface.
BTW, is there some not-outdated Guile informations? If there are I can't find them
"The obvious mathematical breakthrough would be development of an easy way to factor large prime numbers." Bill Gates,
Common Lisp is best
I've used some of the commercial GIS tools, and the worst aspect of most of them was the clunky interface. Why does GRASS have to have a GUI? I personally would be happier with a command line tool that could output data into different formats. One export format for a vector-based graphics displayer, one export format for a raster-based graphics displayer, one for VRML, one suitable for plotting with Gnuplot or whatever, one for importing into a database, etc.
Isn't that the Unix way? A series of small, fast, specialized tools to do something, not one huge, monolithic tool that tries to be all things to all people? If the developers of GRASS created a strong backend, with ties to SQL and customizable export capabilities, with a well-documented API and perhaps minimal Tk/GTK/Qt/whatever hooks, then GRASS would end up being a much more poweful and flexible tool.
darren
(darren)
I don't think the public money aspect is important.
The issue I have with an open source project suddenly changing licensing schemes, especially one which has been around for a long time is...
Did you bother to ask *EVERYBODY* who had contributed to the project since it started?
Let's say I have the open source project called 'widget', and it's released under an older typical open-source license which is BSD style with a clause for non-commercial use only.
Joe, and George, and Susie all contribute to the project at various points. Joe writes a major part of the project, and then gets a new job and disappears.
A few years later George suggests they change the license to GPL so that they can get slashdot.org to post articles about the widget project.
Can they do this? Without Joe's permission? I don't believe they can.
If you can do this, without getting the consent of every single developer who ever touched the code... Then I can also take the Linux kernel and rerelease it under a SCSL without getting anybody's permission.
The purpose of the GPL is to allow only those members of the public who will agree to the GPL's terms to have access to the material. That is only a small fraction of the whole public. If you say: well, the whole public MAY use it, I'll say: so what - the whole public MAY use M$'s software. Just the terms are different. Nothing with a copyright or a license is 100% public.
Anybody got a link to the last version in the Public Domain? Or even the last Army version? I'd like to make sure it will be saved for the public not that these code hoarders are claiming ownership of their fork and releasing it under license so it's of no use to most of the public.