James Fee GIS Blog

79 responses so far ↓

• 1 Straw // Feb 18, 2008 at 12:50 pm

  Don’t go overboard with this, James. It was mostly you who has jumped onto the petty argument about whether or not to call Google Maps a “GIS”. If it wasn’t for you, that horse would not have received nearly as much beating as it did.

• 2 James Fee // Feb 18, 2008 at 12:52 pm

  Please Straw, you are the on who can’t admit that people are “doing GIS” on the internet these days. *rolleyes*

• 3 Sean Gillies // Feb 18, 2008 at 12:55 pm

  Linux is like the Dread Pirate Roberts: it’ll power your business today but might very well kill it in the morning.

• 4 Sean Gillies // Feb 18, 2008 at 12:58 pm

  That was me channeling the haters, if you can’t tell.

• 5 Lefty // Feb 18, 2008 at 1:04 pm

  Manifold, which has to my mind been an appalling waste of a generation of programmers in a narcissistic effort to engineer a GIS system that was already grossly obsolete in the late 1990s, with no greater point, apparently, than being a front for NVIDIA CUDA to sell video cards to unsuspecting newbies.

• 6 Straw // Feb 18, 2008 at 1:11 pm

  OK, I think I’ll skip this thread. Nothing to see here.

• 7 Gretch // Feb 18, 2008 at 1:20 pm

  Think of how much better Manifold IMS would be if they dropped their crappy IMS product and went with MapServer?

• 8 KipterUh // Feb 18, 2008 at 1:23 pm

  Well put James, the freedom is in the choices you are able to make. The myth that seems to exist out there is that the prime reason to use Open Source is that you don’t have to pay licensing costs.

• 9 Paolo // Feb 18, 2008 at 1:26 pm

  James, couldn’t agree more with your last sentence in the post body

• 10 J.B. Churchill // Feb 18, 2008 at 1:31 pm

  I may not have a quality opinion about basing an entire operation on Open Source GIS tools (because I’m with an EDU organization and we get ESRI products handed to us cheaply enough). I still prefer Open Source in many cases. The best example I can think of is that Web standards based content is easy with Mapserver while the basic IMS site you get is etrocious. I prefer the Minnesota Mapserver.

• 11 Jean // Feb 18, 2008 at 1:43 pm

  @Gretch: Can MapServer do something Manifold can’t? Unless MapServer has spatial SQL, the reverse is certainly true.

• 12 FOSS4G // Feb 18, 2008 at 1:44 pm

  Who would have thought the “ESRI Blogger” would be the voice of reason on Open Source GIS.

• 13 Bill // Feb 18, 2008 at 1:56 pm

  James, you’re spot on about freedom. The luxury of being able to use a product exactly how you need to is incredibly compelling.

• 14 Good Buddy // Feb 18, 2008 at 2:39 pm

  Yup roger that, OSGEO good buddy, we got that dot net thang out here in the bumper lane with a clean shot here through to the gdal. y’all cooking good with the mapserver but looking at feeding them ArcServer bears on the dirty side. i’m hoping to skip that fugitive on the manifold. Keep the bugs off your glass and the trouble off your ….

• 15 Bill // Feb 18, 2008 at 2:46 pm

  @Good Buddy
  Best comment I’ve seen so far….

• 16 Gretch // Feb 18, 2008 at 2:51 pm

  @Jean

  Well for one it can run on Manifold.net’s webserver which runs Linux.

  In all seriousness I can’t figure out what Manifold IMS can or cannot do. Manifold obviously considers ArcIMS a compeditor and since MapServer runs rings around ArcIMS I’d say a MapServer user wouldn’t find Manifold IMS of any value at all. I’ve seen the demos of Manifold IMS that folks have posted on this blog and to me shocked at the web front end. I thought tables were depreciated in web design, but I guess not. If this is the premier Manifold IMS application, then MapServer folks have no interest in Manifold.

  http://www.onstar.com/us_english/jsp/coverage_map_2008.jsp

• 17 Paul Ramsey // Feb 18, 2008 at 2:57 pm

  @Gretch, that site doesn’t like Safari, alas. My own fault, for running a browser on an operating system based on antiquated technology from 35 years ago. I should be ashamed.

• 18 nvidia_noob // Feb 18, 2008 at 6:13 pm

  Lefty says:
  Manifold, which has to my mind been an appalling waste of a generation of programmers in a narcissistic effort to engineer a GIS system that was already grossly obsolete in the late 1990s, with no greater point, apparently, than being a front for NVIDIA CUDA to sell video cards to unsuspecting newbies.
  

  you’ve said some stupid sh** on this site before, but this one takes the cake. A front for NVIDIA CUDA to sell video cards to unsuspecting newbies? Good god man, don’t you read.

  Have you seen some of the postings about the performance improvements when using NVIDIA? Don’t you think that exploring massive parallel processing on the desktop is a worthy endeavor? Or, is your head in the ESRI and open source sand so that you don’t actually recognize potential innovation when you see it?

• 19 Luc Vaillancourt // Feb 18, 2008 at 6:17 pm

  I can’t control myself. I see a business model behind every offering, free or not, open or not.

  I think the “coup” of the decade is Autodesk giving MapGuide to the Open Source Community. It was the only way to get the attention of people interested in ArcIMS (right side) or MapServer (left side).

  Open Source offers great opportunities. Several thousands downloads per month … you just have to convert 1% to buy the “enterprise” version and/or professionnal services… and there you go. Sun bying MySQL is a good example, the other way around…

  By letting people download / try / prototyping with Open Source, your multiply your chance to convert them!

  Good business model!
  Look at this very simple 14 slides PowerPoint about the “Strategy of giving”…
  http://www.slideshare.net/Miikka/business-10-vs-business-20
  Giving creates action!

• 20 James Fee // Feb 18, 2008 at 6:40 pm

  @nvidia noob

  I’m pretty sure that was Lefty’s point given he copied what Dimiti wrote.

• 21 Jean // Feb 19, 2008 at 12:57 am

  @Gretch:

  “In all seriousness I can’t figure out what Manifold IMS can or cannot do.”

  So you can’t figure out what Manifold IMS can or cannot do, but you will still call it “crappy” and argue that MapServer is better.

  Just what I thought. Truly insightful.

• 22 Cellulose // Feb 19, 2008 at 9:16 am

  @James

  What do you mean by “freedom” when you talk about using Open Source products? I certainly *don’t* see the freedom in terms of implementation of/changing the product. I just left a job where we had an over-investment in open source products and yet we still rarely had the resources to make all but the most minor, cosmetic changes to open source products.

  The only place it really offered “freedom” was on the Research (and I mean real RESEARCH, not commercial ‘research and deelopment’) side of the house… but then once we were told to turn the technology into a product, licensing conflicts forced us to rewrite from scratch anyway.

  I suppose if you mean “freedom” in the context of competition, different feature sets, and often times an ability to interchange certain products in the pipeline, I can buy into that…

• 23 Tim Maddle // Feb 19, 2008 at 5:55 pm

  @Andrew,
  Keep it going! I just started (back) with PostGIS recently and hope that the Spatial SQL will help me perform some analysis faster than ArcMap does it.

• 24 dylan // Feb 19, 2008 at 7:33 pm

  I couldn’t post this on the Open-Thread-7 page, as it is a reply to <a href=”this comment. Sorry for the cross-post!

  Wow Dimitry, I think you might just win the prize for that last one. Again, your comments reveal just how little you know about any of the applications I listed. I’ll keep my reply short, as I am not payed to advertise.

  Some definitions:

  1. GIS: any tool related to the management / analysis / summary / visualization of geographic data. Simple, but good enough for an operational def.

  2. black-box: this is something that performs an operation in such a way that the user has no means of knowing what is going on. all you have is an input and an output. a good example interpolation in Manifold via kriging! nice one. I wouldn’t trust a non-specialist to use this technique, let alone suggest that a “surface via kriging” is just a point and click away!

  Open source GIS applications are by definition not black-boxes. (BTW the command windows on windoz is sometimes black, and the window is rectangular– but this is not what “black-box” means). If I am unhappy with / concerned about / interested iin the algorithms employed in say, interpolation by regularized splines with tension in GRASS, I can look it up, and ask the author / read their paper on it. How exactly would I investigate the algorithms in commercial products?

  Now onto the packages.

  PostGIS: yes, a spatial extension to PostgreSQL– looks like you read the front page, have you seen the manual? It does a lot more than buffers. In fact, it does just about every vector operation (albeit using the simple features model) that one might need. No GUI? No problem.

  R: have you ever used this, or seen this comprehensive list of spatial packages? I suppose not. Yes, R is often used for statistics. You can also perform statistics on geographic data (who knew?). If you spend a couple minutes with it, you can even make maps.

  Mapserver: data presentation (need I say more?), data management (image server via WMS)… Not a “complete GIS application”– but a nice component in one (I’ll get to that in a bit). I guess you also forgot about all of the languages other than PHP which have native bindings in Mapserver (C#, Java, Python, Perl, …).

  GMT: I guess you haven’t used this. Look at the cover of any AGU publication for some examples of what people do with it. sure the syntax is not for everyone– but it is one hell of an automated or even semi-automated tool.

  And then GRASS: abysmal vector support? when was the last time you have used GRASS– honestly ? I’ll agree that the vector model (topologically correct) is difficult for some to use at first, but c’mon it isn’t that hard. Yes the GUI still has a ways to go, but there has been very significant improvements in the underlying GUI toolkit. GRASS won’t work as a drop-in replacement for most people, that is true– but you are forgetting what GRASS strives to be; a dedicated geographic/image analysis engine.

  I guess GRASS can be difficult if you can’t read, or don’t choose to read instructions. For those that like clicking their way through life GRASS might be too much trouble. Here is one especially difficult procedure in GRASS. 6 lines of “scary code” and you have a nice classification done for you.

  Its no wonder that these type of operations scare people– after all typing in commands makes you think slowly about what you are doing, and is a self-documenting approach to transparent work-flow.

  By failing to realize that I gave that laundry list of application as a demonstration of small , highly efficient GIS tools your famous affiliation with UNIX should be called into question. Really. Surely you must be familiar with this approach to problem solving.

  Its like this. If I want to process text I would use:
  bash, awk, sed, cut, paste, grep, sort, … etc.

  If I want to work with a complex set of geographic data I would use the list of applications from my original post. The point is this– the collection of open source “GIS” applications are not drop-in-replacements for monolithic corporate types. However, combined they form a fully-function set of GIS tools. Note that they are not toys, and can be used to accomplish the same type of end results as one might expect with a COTs GIS.

• 25 Straw // Feb 20, 2008 at 6:04 am

  Dimitri posted a reply in the original thread.

• 26 Chad // Feb 20, 2008 at 7:06 am

  Well, a half attempt at a reply.

  I doubt he has tried any of the tools.. so really, all it is until proven otherwise … is a bloated opinion to be taken with a grain of salt.

• 27 Allan doyle // Feb 20, 2008 at 6:16 pm

  Somewhere along the line, I noticed the Freudian slip in the title…

• 28 Dimitri // Feb 21, 2008 at 2:13 pm

  “Dimitri’s assertion at Linux was a complete waste of effort could have been the most bizarre comment on my blog I’ve ever read.”

  Well, James, I’d be interested in hearing why you find it “bizarre.” It is a shame you don’t tell us that, because that forces us to guess at your thinking. You also somewhat unfairly comingle my comments on UNIX and Linux with a misleading headline regarding open GIS.

  Surely you know that UNIX is the original mother lode of open source work. If you know anything about UNIX you know that.

  If you are suggesting that somehow in any way people were prevented by lack of source code access from tinkering with UNIX itself to have “the freedom to respond to the needs to a project directly” you don’t know much about either the technology of UNIX or about the easy access to UNIX source code enjoyed by programmers worldwide.

  Part of the genius of UNIX in the first place is that you don’t need source code to craft massive new functionality for the OS, as endless new capabilities created by generations of programmers have demonstrated.

  For those very few things for which source code access is useful, as far as I can tell just about every university on the planet that had computers had UNIX source code, because AT&T gave source code away for free to universities. I trust every reader of this blog is worldly enough to understand that if you hand out source code to tens of thousands of universities as a practical matter it becomes easily available to anyone who wants it.

  Let me remind everyone that it is exactly this dual combination of easy expansion without needing source coupled together with free and easy access to source code that resulted in the proliferation of endless UNIX variants. Anyone who knows anything at all about UNIX knows about the endless variant distributions cobbled up by everyone and their brother.

  So this talk that you have to re-invent the wheel to be able to do what you want to do is utter nonsense from a technical perspective. Anyone who knows UNIX knows that.

  Let’s talk about re-inventing the wheel: James, surely you understand that Linux started out as a copy of UNIX and even today is more or less a copy? What’s the point of spending so much time re-inventing the wheel?

  From a technology perspective there never was any point to that. When Linux got started, UNIX on the 386 already existed in numerous, near-free distributions that were thoroughly debugged by the consortia involved in creating them, with a huge number of device drivers and endless thoroughly-debugged utilities and development tools. Why go to the effort of recreating and re-debugging all that?

  I think the only reason is the reason I gave, that people didn’t like AT&T’s ownership of UNIX and wanted to create their own copy of UNIX that AT&T didn’t own.

  Linux itself has evolved over the years, so now you see a new set of self-justifying rationalizations in addition to the above. People say things like “well, Linux has allowed new ideas to get implemented that are distinct from UNIX.”

  Yeah, sure, so what? Nothing except being antagonistic to AT&T’s ownership prevented you from doing all that within UNIX as zillions of roll-your-own versions of UNIX have demonstrated since time immemorial.

  Some folks have said “Oh, the open nature of Linux has allowed collaboration by many people over the net.” Yeah, so what? Comments like that reveal ignorance of how UNIX evolved under the contributions of programmers worldwide working first over USENET and then Internet as well.

  For all the variations and parallel evolution and changes over the years Linux is still more or less a copy of UNIX that is distinctive primarily because of who owns it.

  There is no reason, and never really was, a reason for Linux to exist except for a political desire to have a UNIX that was not owned by AT&T. I’m not saying that wasn’t a strong motivator in certain circles, but anything technical you wanted to do, any extension, any variant desired could easily have been done as yet another UNIX variant.

  Prove me wrong by giving me just one technical capability made possible by Linux that could not have been done as well as or better applying the same effort to UNIX. Just one.

  Let’s cover my last two points briefly, that re-inventing the wheel was a waste and further, that doing so was especially a waste because now you’ve ended up with something that costs more:

  If you already have an operating system like UNIX with millions of man years invested in it, to re-invent the wheel takes a lot of man years just to get to the same starting spot even if you cheat a bit at the beginning by using the functionality and ideas of UNIX to guide your copy, by recycling UNIX source code or by applying what you learned from studying UNIX source code into your copy.

  If you had invested those same man years into evolving UNIX or, better still, starting with something really cool and new, you’d not have wasted so many man years on copying something already done.

  Some people freak out when I point out that UNIX was old and obsolete back in the late 80’s. So true, as the creators of UNIX themselves often remarked in those days. It is not an accident they couldn’t bring themselves to touch UNIX because for years they had already moved on to newer ideas like Plan 9. It would have been so much more productive if the effort that went into Linux to ape an old operating system had gone into something truly new and modern.

  James, I trust you know enough about the Linux market to realize that the way most people buy Linux it is not free? Surely you realize that if Linux is going to ever approach Microsoft in popularity it will have to be distributed as what used to be called “shrink wrapped” product using the methods of commercial vendors such as Red Hat?

  I grant you that in the early days of Linux it might have made sense to go to the effort of making a copy just to get out of ownership issues with AT&T. Although UNIX was darn near free, it was not completely free so I can see how there might be some appeal to having to pay zero royalties to anyone.

  But there is also a “penny wise and pound foolish” element to that, because if you count up the labor involved in duplicating UNIX to roll your own “free” Linux unless you value your time at zero you can easily end up spending a lot more than what a near-free UNIX would have cost you.

  In fact, the complexities of generating usable distributions of Linux have transformed the Linux available for most users into a highly commercialized Red Hat or similar distribution for which they must pay a *lot* of money in what is indisputably a commercial transaction. But now instead of having to deal with AT&T’s licensing they deal with a variety of licenses from GNU to the commercial licenses imposed by commercial, for-profit operations like Red Hat who want to make sure no one rips off their IP.

  As it is commonly sold, Linux is not remotely “free.” Take a look at the Red Hat store and you see an unlimited sockets distribution begins at $1500 - this is “free”? At these prices, even ESRI begins looking like freeware!

  I therefore stand by my thesis:

  1. You could do whatever you wanted to with UNIX, and people did, creating endless variations.

  2. Source code access to UNIX was easily available to anyone who wanted it.

  3. Linux today remains more or less a copy of UNIX, with no significant technical distinctions that could not have been accomplished had the same or less effort been deployed in real UNIX.

  4. The main distinction is who owns the operating system, but the commercialization of Linux blurs that distinction.

  5. The way most people must buy it, Linux now costs more than UNIX did.

  By the way, does the fact that many people today use Linux change any of the above? No. Can I point out that it was a waste to re-invent the wheel yet still continue using the new wheel? Of course.

  Linux has evolved into a fine product so, other than the high cost of commercial distributions, there is no reason not to use it. What’s done is done. My complaint is that going the long way around to get back to square one was a waste of time that could have better been spent building on what was already a strong position.

  In fact, I’ll go beyond that and point out the sadness UNIX advocates (like myself, to the end) feel at how the incredible waste of effort invested in Linux had a prime effect of further Balkanization of UNIX, of further erosion of what had a chance to come together as a mass market solution. For all those guys who felt proud about spending a programming lifetime just to show AT&T you don’t need them anymore, well, brothers, it could be well true that you put a gun to AT&T’s head and blew them out of business, but while you were pushing your UNIX variation in a civil war against your own role model all you ended up doing was waste precious time that allowed Microsoft to catch up and then pull ahead.

  It would have been far smarter for anyone so strongly interested in UNIX that they would spend years engineering their own copy to have contributed to forming a mainstream UNIX effort, to elaborate the operating system for mass market hardware platforms, to assiduously create installations and device drivers and GUIs to make UNIX work on PCs effortless, to take advantage of the huge lead of X Windows (years ahead of Microsoft Windows) and other UNIX windowing GUIs to craft a decisively better operating system for the immensely large world of Intel PC clones that was selling in the hundreds of millions of units. Had all those people who worked through the 90’s to do Linux made those contributions to UNIX, I believe that UNIX in the early 90’s could have become so strong that it could have made Windows impossible.

• 29 Gretch // Feb 21, 2008 at 5:18 pm

  Dimitri, you sure you don’t work for SCO? I’m pretty sure you do as you’ve completely misrepresented the reasons why Linux was created in the first place.

  Sure Linux was created to be a “UNIX-like” OS which was the basic idea behind “modern” OS at the time, but despite your assertion that UNIX was available, it wasn’t available under the GNU license.

  Your inability to get beyond your petty ideas about why folks use or donate time to “free” projects is disturbing, but not unexpected.

  Why wouldn’t folks pay for Linux? Get support and fund development? That is the freedom that you get when you join/support an open source community.

• 30 Dimitri // Feb 21, 2008 at 6:13 pm

  “[...] open source GIS [...] gives you the freedom to respond to the needs to a project directly, rather than working around them. ”

  and

  “It is the freedom that it gives both developers and the users of the applications developed with it, that is the “power” of Open Source GIS.”

  James - If you say the above, I believe you. But I do not quite understand the implications you intend, so let me ask some questions to try to understand what you so emphatically state.

  First things first:

  “open source GIS gives you the freedom to respond to the needs to a project directly, rather than working around them. ”

  That is a wonderful thing. In my own working life I make it a point to always choose software that gives me the freedom to respond to the needs of a project directly. I hate having to use work arounds, so I choose my software carefully to make sure I don’t have to ever hack up a workaround. If you have achieved that objective with whatever open source GIS tools you are using, you have chosen well.

  But why don’t you think commercial products also give their users the freedom to respond to the needs of a project directly? Isn’t that exactly what a well designed product is supposed to do, without inflicting workarounds upon the customer?

  Do you think the buyers of commercial products are eager to buy products which do *not* give them the ability to respond to the needs of their projects directly? Do you think buyers of commercial products look forward to workarounds? I assure you that is not the case, at least not in the mainstream where people have choices.

  From many years of experience observing the habits of buyers of commercial software, I can assure you that they are usually very demanding people. They absolutely, hands down, want to make sure that the tools they buy without doubt give them the freedom to respond to the needs of their projects directly, without any mucking about with workarounds. Fail to do that for them and you’ll end up out of business right quick.

  If anything, I assure you that buyers of commercial packages are an especially prickly lot because they are spending real money for those products. They expect to get what they paid for. Once you take money for a product it better work or the buyer gets angry.

  It’s not like they can say it is some free thing they downloaded off the web so it if doesn’t work there’s no money lost. And, presumably unlike the case with open source, if you sell something to your customer and it doesn’t work you don’t have the luxury of telling him “No problem, hire some programmers to make it work like you want. You didn’t pay anything for it anyway.”

  One other thing in your sentence that I would ask to be clarified. I don’t in any way want to disrespect the merits of open source GIS, but by way of understanding the cultural approach the question naturally arises: If you think that a commercial product somehow requires a user to work around the needs of a project, I get the impression that you are saying open source GIS does not require a user to work around the needs of a project. Not ever. If I understand you correctly, you can make such a strong statement because you feel able to say that because if anyone’s project ever has special needs, well, rather than having to work around those special needs the user can take advantage of their source code access to the open source GIS to re-program the open source GIS tool so it can handle those special needs. Right?

  But, here is the point I don’t understand… if you think it bad for a GIS user to have to be “working around” the needs of their projects because those specific needs are unserved by their commercial package, isn’t it much worse and more difficult if the open source GIS user is compelled to avoid such working around by having to reprogram his GIS software? That is, isn’t this just replacing any workarounds in process with workarounds accomplished through new code and new software devlopment? Help me out here, if that is what you are saying and the culture of your comment is that this is OK and better.

  Let’s move on to “power”:

  “It is the freedom that it gives both developers and the users of the applications developed with it, that is the “power” of Open Source GIS.”

  OK. I’ll take that at face value and trust your statement this is a benefit of open source GIS. But once again the question arises why you think that need not be the case with commercial products.

  If it is “freedom” and “power” to accomplish through programming whatever you want to accomplish, why do you think commercial products that offer extensive programming capabilities do not also offer such “freedom” and “power.”? If you are suggesting that the opportunity to tinker with source code is the source of that freedom and power, I absolutely believe you that it grants to those folks able to do that additional freedom and power. But are you suggesting that a commercial product could not be crafted in such a way as well? Why do you think the developers of commercial products would neglect to include open doors, support for user-programmed extensions and the like that would enable users in any material way they desired to use whatever source code they darned well pleased?

  Also, how does this grant such “freedom” and “power” to the *users* of applications developed by people who have access to source code? The developers of commercial packages also have access to their source code, so how could the binary result handed off to users be different in terms of “freedom” and “power”?

  I grant that there is a theoretical basis for supposing that some user might feel more reassured to know that a de facto source code escrow arrangement consisting of his ability to get his hands on open source GIS source code might be comforting. But there’s no reason a commercial product could not also escrow source code.

  For that matter, most experienced users understand that in the case of truly significant software packages that involve millions of lines of code it is not realistic for them to self-support years after the fact, no matter how modularized or now well commented that code may be, whether it be open source or commercial code. Experienced people know that it can take a very long time to understand the innards of code, and to hire someone who already does (whether it is open source or commercial code) is not usually cheap.

  I guess that is one reason that most open source GIS tools *users* have no more interest in rooting about in the innards of code written by someone else many years ago than commercial code users do. Although it seems that most open source GIS tools users, just like most commercial users, are not experienced programmers, I’ve noticed that even those who are (perhaps *especially* those who are) experienced programmers are the first to take advantage of tried and true software development methods of programming against accepted well-defined interfaces while usually refraining from tinkering with the innards of the objects they use. Experienced programmers understand that maintaining a divergent distribution is a total pain in the neck, so they don’t get themselves on that hook if they can avoid it.

  What I’m saying is that when you look at packages like R or GDAL the observation is that most folks are *users* of those packages and not *reprogrammers* of those packages. They pretty much use them without changing the innards, and they leave the re-wiring of innards up to the specific communities developing those packages.

  So how is that different from people who use sophisticated commercial packages or objects libraries and leave the re-wiring of innards up to the commercial teams that maintain them? I mean, besides that in one case people are getting paid for what they do and in the other case people are doing it for free?

  In both cases, the great majority (in some cases, I suppose to the 99% level) of users are consumers who are working against what is effectively a compartmentalized object developed by someone else.

  Unless you yourself are willing to dig into the source code innards of whatever tool you use, there is no more or less “freedom” or “power” for either you or the users of applications you develop.

  Note that I do not in any way challenge your declaration that open source GIS gives you the freedom to respond to the needs of a project directly rather than working around them. Likewise, I do not in any way challenge your statement that the “power” of Open Source GIS is the freedom that it gives to both developers and the users of applications developed with it.

  But I would like to get educated on why you think that commercial products do not also give such benefits, such freedom and such power.

• 31 Dimitri // Feb 21, 2008 at 6:21 pm

  Gretch:

  “despite your assertion that UNIX was available, it wasn’t available under the GNU license.”

  OK, I hear you. You are saying that the ownership licensing was the motivation. That’s what I said, so I am grateful you agree with me.

  You’ve said I misrepresented the reasons for why Linux was created. I did not give any other reasons except my sense that it was the above reason. So here is your chance to correct my omission and tell us the other reasons.

  “Your inability to get beyond your petty ideas about why folks use or donate time to “free” projects is disturbing, but not unexpected.”

  ??? Easy to argue when you make things up. Please quote my words where I offer any ideas about “why folks use or donate time to “free” projects.” …as is the case with any complex societal phenomenon, there are usually a lot of reasons for that but I’ve not offered an essay dissecting those.

  “Why wouldn’t folks pay for Linux? Get support and fund development? That is the freedom that you get when you join/support an open source community.”

  I never said that folks should or should not pay for Linux. I just pointed out that Linux is not free for most users, and that today’s costs for Linux are much higher than they were for real UNIX.

  If you think it is OK that folks should pay for Linux because that gets support and funds development, I’m OK with that. After all, the reason people pay for commercial products is to get support and to fund development, so why would I be against you paying for Linux for the same reasons?

• 32 Chad // Feb 21, 2008 at 8:35 pm

  @Dimitri: I have said it before and I will say it again.. if you are going to post stuff like;

  “I never said that folks should or should not pay for Linux. I just pointed out that Linux is not free for most users, and that today’s costs for Linux are much higher than they were for real UNIX.”

  You have to post links to this “factual” information. The only time I have ever really heard about the costs of Linux over other OS’es is from Microsoft sponsored research.

  You are very bad about spouting out “facts” and not backing them up with a reference. If I was still grading reports… I would have to give you a D- because of lack of references.. sorry, that is how it works now.

  And I am also amazed you have a post that I didn’t have to scroll to read.

• 33 Tim Bowden // Feb 21, 2008 at 10:13 pm

  Dimitri, you need to learn the difference between free as in free beer, and free as in freedom. Even with Red Hat linux (which you may pay quite a bit for), you still get free as in freedom. That’s the free that’s important. It was never about cost (though that’s a great bonus when you get it). Maybe that’s not so important to you (and that’s fine), but to freetards it is. If you’d like an example, have a look at Paul Ramsay’s latest blog (http://blog.cleverelephant.ca/2008/02/risk.html) and please Dimitri, treat it as an /example/, not a defining instance.

  I think we all need to respect that different people have different values, and move on. Agree to disagree about how important that type of freedom is.

• 34 Dimitri // Feb 21, 2008 at 11:00 pm

  Chad:

  Gretch responded to my post in which I cited Red Hat’s price of $1500 for Linux, hence the talk about prices for Linux. If you doubt that “fact” I invite you to visit the Red Hat online store to see their pricing directly. Red Hat seems to be the most popular packaged distribution so I used their pricing as an example. It could have just as easily been SuSE or someone else. You are not seriously suggesting that Red Hat or SuSE are a front for Microsoft sponsored research?

  Recall that this thread is about UNIX and Linux so the comparisions are between them, and not to anything Microsoft. [ I only brought Windows into this to suggest that had the effort invested into Linux been put into UNIX then perhaps Windows might never have happened. ]

  Before Linux, UNIX too had a bifurcated pricing structure. For most technical UNIX users, the hardcore programmers moving UNIX and open source and software tools boldly into the future, it was free or effectively free. In fact, it was so free that the social contract, the ethos of creating free code for UNIX and giving it away was so strong that surprisingly many UNIX programmers did not know UNIX itself was not freeware. I can’t tell you how many times during the management of a UNIX development project I would have to explain to an engineer that he couldn’t just give source code away to a buddy - we’d have to first make his buddy a “contractor” so then we could give it away to him.

  For commercial users there was a wide range of pricing depending on the distributor, but contrary to uninformed thinking none of that was AT&T’s fault. It was not difficult to get your actual cost of UNIX for redistribution down to zero on the margin, especially if you could ride on someone else’s redistribution license, as Intel allowed many distributors to do. And I mean a marginal cost of zero, absolutely no royalty at all that needed to be paid to anyone. That is why commercial pricing on mass market machines had dropped in some cases to under $100 or even under $50 . At least one vendor was giving away free UNIX licenses pre-loaded onto mail order, heavily discounted hard disks.

  In modern times, Linux has a similarly bifurcated price structure. If you are technical you download Debian and Linux also is free for you. Non-technical folks buy Red Hat or some other commercial distribution. Gretch correctly notes there is nothing wrong with that. It seems like every day there is another new product, such as Oracle’s new Linux distribution. Those commercial distributions are not free and they are so full of non-GNU proprietarizations that for all practical purposes they are full bore commercial products, breathtakingly high in price.

  Everyone in Linux knows the above, and you even hear people crow about it, for good reason I think, since people realize they are never going to get traction against Microsoft without emerging powerhouses like Red Hat or SuSE to provide packaged Linux distributions to the unwashed masses. For every guy with the technical skills to utilize a totally free distribution, there are a lot of rubes out there, I suppose the thinking goes, who will need a packaged commercial distribution to get started.

  “The only time I have ever really heard about the costs of Linux over other OS’es is from Microsoft sponsored research.”

  Well, that’s nothing short of astonishing. I can’t believe that the Linux community would fail to study and discuss the costs of Linux over other operating systems all on their own, without leaving such research exclusively to Microsoft. Surely there are many careful studies and analyses based on fact, none of this wishful thinking stuff, conducted by economically-minded members of the Linux community that expertly dissect all aspects of the costs of Linux over other operating systems. I just don’t believe that the Linux community could be such slackers that the only such analyses would be those funded by Microsoft.

• 35 Dimitri // Feb 21, 2008 at 11:14 pm

  Tim:

  I know what you mean and I’m a strong believer in the freedom you cite. It is worth fighting for. But if you want that you get Debian, not Red Hat, because the Red Hat distribution is so loaded with non-GNU licensing of Ret Hat proprietary things that as a practical matter it is not freedom to do with it as you please.

  Look, I agree that even the symbolic appearance of freedom is important. But sometimes it is so easy to get worked up over symbols that you don’t realize you have simply traded one master for another. And I don’t mean the simple case of thinking you have GNU licensing when so many details are under a separate proprietary license. I mean the more sophisticated case where you think you’ve acquired freedom by holding source code and the freedom to do with it whatever you like but then you discover that what has happened is that you have traded the modest imposition of someone else holding an original license, as AT&T did, but otherwise providing a rock solid intellectual property structure upon which you could count, for what is at times an uncertain ground or the lack of freedom imposed by having to hire a crew of experts to actualize any practical ability for you to exploit the theoretical freedom of doing anything with that source code.

  Remember my comparison is against UNIX, which set an extraordinarily high standard for social compact. Even AT&T, which many firebreathers wanted to be eliminated for daring to bring any IP control into the matter, actually did a very good job of holding UNIX almost in the public trust.

  And last but not least keep in mind that with total freedom comes anarchy. Linux is not “free” in your sense of the word: it is not in the public domain but rather is constrained by the GNU license. That license only works to deliver many freedoms because it constrains some freedoms. For example, you won’t ever be able to apply certain levels of massive capital to core improvements because you’ll never be able to enjoy a period of exclusivity required to earn the cost of that capital due to the copyleft nature of the GNU license. I’m not saying that’s wrong: I like copyleft - I’m just pointing out that sometimes to get an overall balance of more freedom you have to give up some freedom. In the context of that thought, AT&T’s stewardship of UNIX was not so badly done.

• 36 Tim Bowden // Feb 22, 2008 at 12:15 am

  Dimitri, for fsck sake, RH is /not/ loaded down with proprietary stuff. It’s got their trademarks in it which you aren’t allowed to copy, and that’s fair enough. The code /is/ free. If you’re going to babble on about with nonsense like that you won’t get taken seriously. Your argument is also self contradictory. RH is proprietary, but gpl licensed code can’t be distributed with a period of”exclusivity” (ie, in a proprietary state) so you can’t get a return? For $DEITY’s sake man, get a grip.

  Sorry, but I’m not going to waste my time anymore if that’s the quality of your arguments.

• 37 Chad // Feb 22, 2008 at 7:32 am

  “Well, that’s nothing short of astonishing. I can’t believe that the Linux community would fail to study and discuss the costs of Linux over other operating systems all on their own, without leaving such research exclusively to Microsoft. Surely there are many careful studies and analyses based on fact, none of this wishful thinking stuff, conducted by economically-minded members of the Linux community that expertly dissect all aspects of the costs of Linux over other operating systems. I just don’t believe that the Linux community could be such slackers that the only such analyses would be those funded by Microsoft.”

  http://www.google.com/search?source=ig&hl=en&rlz=1G1GGLQ_ENUS243&q=cost+of+linux&btnG=Google+Search

  What other OSes though? You can’t really compare to Unix as Unix depends on hardware lock-in. You can’t take a Unix install and put it on my desktop (Well, ok.. there is ONE that I can think of). But what other OS is there to compare too? You are left with Windows and MacOSX and OSX is a Linux base. But again, with the MacOS you are hardware locked like Unix. So really you have Windows and Linux to do an honest comparison with.

• 38 DirtySanchez // Feb 22, 2008 at 8:07 am

  Anyone that has tried to run SDE/Oracle on Windows compared to Linux (RH) knows the real power of Linux. I’d never go back. Oracle has actually tuned their product to work with Linux. And that goes for either the RH version or their “Oracle” version. All they did was take the Red Hat logo off the box. Good call by Chad on the hardware also.

• 39 Dimitri // Feb 22, 2008 at 8:49 am

  Tim,

  “RH is /not/ loaded down with proprietary stuff. “

  In point of fact it is, although you do have to read the fine print. Consider, for example, the list of “third party” EULAs embedded into the Red Hat package enumerated at

  http://www.redhat.com/licenses/thirdparty/eula.html

  Go follow the links and see what those licenses say. It is not GNU, yet such third party items embedded within Red Hat (along with Red Hat’s own proprietary value-added and software patents) are part of the essence of what makes Red Hat a successful commercial product.

  I don’t criticize that. Red Hat has set out to make a commercial Linux and they’ve done an outstanding job.

  But the assertion that to any meaningful degree people can just take Red Hat’s products and copy them willy-nilly without paying Red Hat is silly. If you want to do that, use Debian.

• 40 osjunkie // Feb 22, 2008 at 9:08 am

  its obvious by Tim’s web link that he is vested in Open Source consulting. So, that is why he is defending linux to the death. Tim, while you crap out on Dimitri, its obvious you are basically the same people: protecting your economic interest.

  You have a vested financial interest in open source since you are making money using it. So, I think its impossible for you to be objective. At least within Dimitri he admits being a shameless marketeer. You are the same thing. At least admit that up front.

• 41 Dimitri // Feb 22, 2008 at 9:25 am

  Chad,

  You appear to have confused me with Tim. Tim is the guy who needs help with using Google, since he is the one who wrote:

  “The only time I have ever really heard about the costs of Linux over other OS’es is from Microsoft sponsored research. “

  It is good that you provided a link to a Google search so that Tim can see there is indeed research on the costs of Linux sponsored by folks other than Microsoft.

  “So really you have Windows and Linux to do an honest comparison with.”

  Yes, well, that’s very interesting but it has nothing to do with my quotation that James flamed at. My point is that taking a 20 year circumnavigation to get to more or less the same place was not the wisest use of time.

  If you disagree, surely you have numerous examples of technical work made possible by Linux that could not have been accomplished with the same effort or less with real UNIX.

  Note that I don’t argue with or criticize in any way the notion that Linux as it as evolved is a fine operating system. I’m not saying it is worse than or better than Windows. None of that.

  For that matter, I’m not even saying Linux is worse than or better than UNIX. I’m fully willing to grant that today you have Linux packages that are “shrink wrapped” installable on PCs and that there are fewer than such available for UNIX.

  The point of my argument is that it is today 2008, over twenty years after 1988 when UNIX was already old. In 1988 there were dozens of shrink wrapped UNIX packages that were installable on PCs with full suites of user-friendly installation routines, automatic configuration of device drivers, etc.

  So, yeah, sure in those twenty years both UNIX and Linux have evolved and the relative fortunes, driven by the economics of ownership issues, of both have changed.

  What James flamed me about has still gone unanswered in this thread: Except for Gretch agreeing with me that it was indeed a matter of ownership, the GNU license, I don’t hear anyone else stepping forth to give a reason what, exactly, 20 years of programming effort were required to do that could not have been done with the same effort or less for UNIX.

  I hear the usual comparison between Linux and Windows but I don’t hear any response to my thesis. Look, if people feel that “well, OK, it was only about ownership and that’s enough for us” so be it.

  So let’s stop dodging the tough question. James slammed me for daring to say that re-inventing the UNIX wheel was a waste, except for taking AT&T out of the ownership picture. If it was not a waste, give me an example, a real, nontrivial example, of some material capability made possible by Linux that could not have been done with the same effort or less with real UNIX.

  What you’ll find is that to answer the above challenge is that you’ll have to learn something about the power, freedom and open source flexibility of the UNIX ecosystem as it was before Linux got traction. And the more you learn about that, the more you will realize what I wrote is absolutely correct. That is not a tribute to me, it is a tribute to the hundreds of thousands of people who worked together to create that rich, fertile and predominantly open source UNIX ecosystem.

  And it is also a slam, because if you dig into it far enough to attempt an intelligent, informed answer to my challenge you will be struck by how many different UNIX implementations there were. Think Linux has given us a rich array of choices? There were hundreds to choose from in UNIX. In fact, you will soon see that a hallmark of the UNIX ecosystem were the internicine wars that erupted over exotica, where groups would decamp in favor of one UNIX flavor or another. Look at it more closely and you can see that Linux was a continuation of that tradition, yet another schismatic group decamping to pursue their own vision of what UNIX should be, apparently driven by ownership issues more than anything else. I don’t criticize the results of what they created, I just point out that the 20 years of effort were a waste compared to what might have been accomplished had the same effort been put into making the existing wheel a much better new wheel or replacing it with something entirely new rather than re-inventing it.

• 42 InTheTrenches // Feb 22, 2008 at 9:35 am

  I manage a network of 60+ GIS data repositories scattered all over my state. The average size of each one is 150+ gig. 5 years ago I started out with Windows servers, but over the past several years have been switching over to to linux servers with great results. This fall I rolled out 45 new linux machines.

  I have complete control over my systems. I am not subject to the windows update schedule, with it’s constant re-booting, security holes, etc.

  I update my target data sites very effeciently using rsync. I can manage and mount any of the machines securely to my central repository using various secure shell tools.
  Sure, I run SDE (on Linux) for multi-user enterprise data editing apps (served via citrix): it works well. However Linux has enabled hundreds of my end users to get at up-to-date GIS data without the need for me to pay for Workgroup SDE at all my target sites. Ever wonder why one-way replication to file-geodatabases still hasn’t been implemented? I can replicate thousands of shapefiles, images, file geodatabases, etc. with a one-line rsync command and avoid SDE altogether.

  I choose to do this with SLES, an enterprise class offering from Novell but that is only because my organization already has a Novell relationship. Everything I do can be done with OpenSuse or other free option.

  Linux has saved us plenty, and put us back in the driver’s seat. Windows has it’s place, but the value of Linux has been proven in our shop.

• 43 Dimitri // Feb 22, 2008 at 9:44 am

  osjunkie:

  I appreciate your comment but should clarify: while I am indeed not shy about being a shameless marketing guy I do not have any economic interest in this debate - I no longer have any investments in UNIX so as far as economics are concerned the discussion of history does not impact me one way or the other.

  If anything, my suggestions to the open source GIS community that they should try to learn from history are exactly opposite to my current economic interests, which are deeply tied to Microsoft and to commercial products.

  As far as economic interests go, my pointing out that had the cards been played better there would be no Windows is borderline treason for a Microsoft courtier.

  For that matter, if I was to be cynical and to play exclusively to the economic interests of my “day job” in commercial GIS you bet I would be the first to say, “Hey, this attitude that awk and grep are the way to do text processing? Yeah, brother, you are absolutely right. That command line stuff is exactly what GIS users around the world want to do. Focus on that.”

  But this notion that somehow people would pervert their discussions in intelligent settings to conform to their economic needs is really pessimistic. How can people be so beaten down, so depressed, so unempowered not to realize that intelligent, effective people have choices and that they usually pursue those business and career strategies that make sense to them exactly because of what they believe?

  Why would anyone expect that someone who so strongly believes in the intelligence and value of a particular path as to bet their working life on it, the precious commodity of a single life’s time spent eight or ten hours a business day, would not also support, advance and explain that chosen path in forums such as this?

• 44 Cellulose // Feb 22, 2008 at 9:48 am

  @InTheTrenches

  Sounds a bit like you weren’t using your Windows Servers to their full potential. Everything you describe can be achieved with Windows Server with out-of-the-box tools and a tiny bit of WSH.

  I don’t understand your example… What does copying shapefiles and other static files have to do with one-way FGDB and Workgroup SDE? If the files existed in a file-based format to begin with and were fine in that format, then why were you setting up Workgroup SDE? And I don’t see how Windows is to blaim for a sub-optimal workflow?

  It sounds like you’re not applying Linux patches in a timely manner… The frequency of critical or high-priority patches is certainly less than Windows Server, but to say it’s not constant? Subscribe to any security list and you’ll see the constant flow of Linux security holes and patches. Many of them in commonly used apps included with Linux ranging from Apache to PHP to SSH.

  But then most of the security flaws in Windows and Linux these days are in 3rd party apps and optional services, not the core OS. Get rid of the optional modules and third-party apps and your rate of security patches will decrease considerably on both platforms.

• 45 DirtySanchez // Feb 22, 2008 at 9:48 am

  Give it up dude. I have a picture in my mind of Dimitri kissing some old HAL9000 looking server telling it how truly great UNIX is. HARDWARE is one of the major reasons. You can put Linux on anything you spec out. End of story. Either you work for AT&T or were fired and now work at Chucky Cheese. I’m not name calling, but from your long entries you’ve got nothing but time………..
  While you are screaming about how much of a waste of time Linux, we’ll be getting work done.

• 46 Dimitri // Feb 22, 2008 at 10:00 am

  InTheTrenches:

  “Linux has saved us plenty, and put us back in the driver’s seat. Windows has it’s place, but the value of Linux has been proven in our shop.”

  That’s great, but you appear to have the wrong thread. This is a thread about Linux and UNIX. UNIX is not Windows, although I can see how some folks might get the two confused.

  I know I’m asking people to think a bit out of the box by asking a question that compares Linux and UNIX. I understand that confuses some people who will try to dodge the intellectual demands of that question by avoiding it, and instead reflexively jumping to the usual comparisons and flame wars regarding Linux and Windows.

  Let’s try to respond to the question asking for a comparison of Linux to UNIX. It’s a tough question that requires thought, I grant you, but considering how James launched this thread surely it should not be so difficult to find someone to answer it.

  By way of remedial assistance: the question actually was what technical capacity was done for Linux that could not also have been done with the same effort or less had it been applied to real UNIX. A hallmark of real UNIX was that it was, indeed, available on every hardware item imagineable. I realize a debate about historical progression requires some additional concentration to understand the difference between historical process and present results, but that’s part of the deal in answering such questions.

• 47 Bill // Feb 22, 2008 at 10:50 am

  @Chad

  “You are left with Windows and MacOSX and OSX is a Linux base.”

  Just a clarification: OSX is based on FreeBSD, not Linux. FreeBSD is based on BSD UNIX, a version of UNIX.

  Not trying to pick sides on this one. Just wanted to clarify that point since OSX can actually be seen as an example of Dimitri’s concept of making UNIX better.

• 48 Chad // Feb 22, 2008 at 11:10 am

  @Bill: Doh, you are right.. I forgot about that. My bad.

• 49 AnonyMousey // Feb 22, 2008 at 12:53 pm

  Several commenters seem to be talking around what can be done with GeoServer with PostGIS, but not mentioning it by name.

  http://geoserver.org/

• 50 Dimitri // Feb 22, 2008 at 1:18 pm

  Good point Bill , especially so since OSX, an example of making UNIX better, is also the one and only example of an alternative OS gaining against Microsoft in the hearts and minds of the masses.

• 51 InTheTrenches // Feb 22, 2008 at 1:24 pm

  Dimitri - we were a Solaris shop for a long time. It was great, but now that our OS is de-coupled from our hardware and our costs have dropped dramatically. If only ESRI had ported WorkStation ArcInfo to Linux I could dump the last sun workstation.

  I know Linux and Unix are different. I was just responding to the notion that Linux wasn’t adding value to the equation for GIS users. For us it has in terms of less overhead and better reliability. I use them all, including OS X (at home) and each has their place.

  Sure you can do all that I described using a Windows servers and a bit of ssh. I’m installing ssh on the last remaining win servers that I run to do just that. But I tell you, the Linux boxes are just easier to manage. I set them up, they run with little to no downtime ever. I don’t need to update the OS weekly because I’ve got them locked down except for the few essential services that need to run. It works for me.

• 52 Cellulose // Feb 22, 2008 at 1:53 pm

  I was referring to Windows Scripting Host, not Secure SHell… They’re not related.

  Considering the amount of flack Microsoft gets for its “Super Patch Tuesday” practice, I think your perception of OS update frequency is a bit exagerated…

  But then I guess if you only check for patches once a month, then you only need to apply patches once a month so it’s a push.

  Just don’t blame the productfor your lack-of-experience with the product.

• 53 bender // Feb 22, 2008 at 1:53 pm

  i think it’s the perfect time now to start a good ‘ole editor war: here are my 5 cent: emacs is much better than vi, man!!

• 54 Cellulose // Feb 22, 2008 at 2:01 pm

  Bender–

  Being that emacs is basically its own operating system, are you opening a new front in the OS-wars?

  (BTW, some of us whimpy developers like the comforts and luxuries of vim, which I’ve learned from vi-zealots vi != vim).

• 55 Dimitri // Feb 22, 2008 at 2:05 pm

  Bill - “now that our OS is de-coupled from our hardware”

  I know you write that sincerely, but where do people get this idea that UNIX is somehow tied to hardware and Linux is not? They are both the same in that both are not significantly coupled to a single hardware architecture.

  Yeah, sure there are some binary applications execution issues peculiar to specific processor architectures, such as whether you are using Intel x86 or SPARC or whatever, but that’s the case with Linux as well. For decades x86 UNIX versions have been binary applications compatible.

  If anything, UNIX has always been most famously decoupled from hardware. Back in the day when there were more than just a handful of processors in use, you couldn’t turn around in any direction, to find any processor or any hardware of any sort and not have UNIX on it. Starting with the same porting base people ran UNIX on everything from microcontrollers inside coffee pots to massively parallel supercomputer clusters. And in the majority of such cases you could sit down to a console just about anywhere using your flavor of UNIX and regardless of the underlying hardware it would appear the same. Very useful.

  Now, if you are complaining about SUN’s famous desire to proprietarize UNIX so that whatever you did ran only on SUN hardware, well, blame SUN for that, not UNIX!

  Speaking of proprietarization I want to make it clear that in no way do I oppose Red Hat’s selling and commercialization of Linux. I’ve received some private email asking me why I am against charging for Linux (which seems an inattentive thing to ask given my comments in this thread). I’m not. I think what Red Hat is doing is great.

  They are doing a first rate job, and, frankly I think they are getting a bad rap in some circles for being too proprietary when in fact, like AT&T used to do, they are doing an outstanding job of running a commercial operation while at all times doing a pretty darned good job of upholding the public trust.

  Yes, they have some proprietary stuff and yes, it is true that sometimes those proprietarizations stand in the way of a totally anarchic freedom that some would like. But Red Hat has to run a business and they are not taking advantage of anyone while doing that.

  For example, all versions of Red Hat depend upon proprietary code (vendor supplied, usually) for things like 3D graphics. But to circumvent that Red Hat does deliver a very non-proprietary base set of 2D graphics stuff. They are making an honest effort to walk the tightrope and should be commended for it.

  I think a better example of where the balance is difficult to resolve is the tension between Red Hat selling support in a way that removes some of the benefit of open source.

  There is the obvious issue of losing your freedom to get support for your paid-for installation of RedHat Linux, when executing your other freedom of being able to modify the source code and recompile your system modules. The latter thing is of course being implicitly used by RedHat when they market themselves to the masses, and thus there is a situation when one of the strengths of the product they market (being able to alter the system) eliminates another strength (being able to get technical support).

  For all that I think Red Hat does a super job of resolving the tension, trying to keep things free as possible while at the same time doing a very good job of selling lots of product in commercial transactions for a high price. I did choose Red Hat pricing as an example of a paid Linux because, well, it is breathtakingly high and helped me make a point, but in all fairness to Red Hat they earn what they charge and they are not remotely close to being the most proprietary of the various commercial Linux distributions.

  I have nothing but praise for Red Hat and would encourage anyone seeking a commercial Linux distribution to buy their Linux product.

• 56 Bill // Feb 22, 2008 at 2:15 pm

  @Dimitri

  I think you incorrectly attributed that statement to me at the top of your last comment.

  I would agree that OSX does seem to have the “mindshare” at the moment.

• 57 Dimitri // Feb 22, 2008 at 3:48 pm

  Bill, you are correct and my apologies for that. I had meant to cite InTheTrenches.

  For everyone else:

  Bill is right about OSX. The interesting thing is that the path to OSX began with Apple’s acquisition of Next, mainly to get the Next OS (based on BSD UNIX) and also to get Steve Jobs back into management at Apple.

  Nothing but Steve’s famous antipathy to Intel and to PCs prevented Next from using mass market platforms to have rolled out the Next OS for PC clones in 1990. They could have pre-empted Windows by years and have become the operating system platform for the masses. For that matter, Steve could have done that at Apple had he embraced x86 at an earlier date.

  I’m telling you, I hear so many conspiracy theories about the rise of Microsoft but such folks rarely admit their own missed opportunities along the way, or for that matter, give enough credit to the very hard work in a come-from-behind mode by Microsoft.

  Lenin once remarked that when the Bolsheviks launched their October Revolution that “they had found power lying in the streets, and they just picked it up.” Looking back I see that same thing happening in the transition from UNIX to Windows, where the UNIX community had overwhelming advantages, indisputably superior technology, windowing GUIs of astonishing elegance, the refinement of applications like the Next suite, all that before 1990 and yet somehow frittered it all away, just left all that lying in the street for someone more focussed in purpose to pick up and use.

  Part of that frittering away came about when religious zealotry got substituted for common sense. When the chips were down, people clung to elitism over leveraging commercial methods. At a time when it would have made more sense to leverage the power of commercial systems, people went on a monkish snit about killing off AT&T and other irrelevancies.

  I see that same phenomenon happening again today with open source GIS. For goodness’ sake, it is not an either/or situation. Nothing whatsoever about commercial GIS products prevents you from deploying open source code in conjunction with commercial GIS products.

  Now is the time to take advantage of the proliferation of widespread, inexpensive commercial GIS platforms to expand the influence upon the masses of open source GIS tools as well. That is the way to expand your influence, to get your way, to allow your experience and your intelligence to shape the GIS expectations of the masses. Take what you find useful from the commercial products and contribute what you think best of your own.

  Look, I respect those folks who can R and GDAL and Proj 4 and GRASS their way into whatever they need to do in GIS. But you have learned nothing from the debacle of UNIX being replaced by Windows if you think that approach will be palatable to the masses. Instead, I strongly urge you to use commercial tools when useful to inject your views on algorithms, on process and on openness into the hearts and minds of the mainstream.

  No sensible commercial vendor will resist that. Every sensible commercial vendor will welcome you and will be eager to learn what they can do better, if they are not already doing enough, to make it possible for you to do whatever you want.

  I realize commercial products have a price, but as Gretch so aptly put it, why would you begrudge them that if the cost is very low and commensurate with the honest value they give you? It is like someone volunteering for Habitat for Humanity begrudging the $15 cost of a hammer. If you want to contribute, do not begrudge the small cost of a tool that expands your freedom to build new things and, if you like, to give them away as you see fit.

• 58 Bill // Feb 22, 2008 at 7:35 pm

  @Dimitri

  Thanks, I had forgotten about Next being based on BSD UNIX. That makes more sense. I had always thought the selection of FreeBSD as the base of OSX had more to do with the freer nature of the BSD license versus the GPL (which it may have) but the direct lineage via Next seems a stronger connection.

  BTW, I think I may officially be guilty of “tangicide” with regard to this topic.

• 59 Tim Bowden // Feb 24, 2008 at 5:16 am

  @ Cellulose, We settled the vi(m) v emacs v everything else war at our local lug the easy way; laser tag. Winning team could claim it’s favorite editor was the best without further justification. If the losers wanted to recommend any editor apart from the winning one, they were required to state “…however it has been proven that (winning_editor) is superior” for a period of twelve months. Worked great. Of course, vim won. Only natural though.

  Next up is paintball to settle the distro wars. We can only have two teams there, so it looks like it will probably be .deb v .rpm. The gentoo freaks will have a hard decision to make.

• 60 bender // Feb 26, 2008 at 8:54 am

  comment no. 60. let’s head against the century break. new topic is: google is still playground, “os” is in adolescense and esri is enjoing indian summer. popcorn, please!

• 61 Dimitri // Feb 28, 2008 at 5:19 pm

  Just a reminder that here we are at post 61 and for all that my challenge to the folks who found my quote “bizarre” to launch this thread still has gone unanswered.

  So let’s stop dodging the tough question. James slammed me for daring to say that re-inventing the UNIX wheel was a waste, except for taking AT&T out of the ownership picture. If it was not a waste, give me an example, a real, nontrivial example, of some material capability made possible by Linux that could not have been done with the same effort or less with real UNIX.

  I realize that part of the art of running a blog like this is expertly poking at a bee’s nest to get a swarm of content out. James is very good at that, the perfect ringmaster. I’m a good sport about the process and appreciate the opportunity he provides for a soap box.

  But, having had my say I will rest my case by noting the above challenge goes unanswered, which apparently has proven that my observation was not so bizarre as far as the community that reads this blog is concerned.

  Some friends have asked privately if feel unfairly treated. I do not: As a current candidate in the London mayoral race remarked when asked if he felt ill-used by the press, “That would be like a fish complaining about the sea.”

• 62