4 Days to Eliminate Lock-in

The ultimate lock-in is having software in ROM on a particular device. MAME (Multiple Arcade Machine Emulator) is a project to allow many thousands of arcade games embedded in old machines to run on modern PCs. To the extent that porting such an application is a big job, it is noteworthy that Google was able to do that in only 4 days to the Native Client of their Chrome browser.

“The port of MAME was relatively challenging; combined with figuring out how to port SDL-based games and load resources in Native Client, the overall effort took us about 4 days to complete.”

Shades of M$’s fear of Netscape providing a competitive platform… Google not only can do that through their browser but also through Android/Linux and web applications. The locks in the lock-in are being pried open one after another. How many applications, industries, businesses, individuals and PCs will escape M$’s prison in 2012? It could be hundreds of millions of PCs shifted from XP to GNU/Linux and a similar number of new PCs coming with nothing but a browser to access web applications.

In 2011, 12.5% of PCs shifted from XP to something (by death or OS change) but only 14.4% took on “7” while 24% of the world’s PCs were newly produced. 61% of new PCs got “7”ened. 39% did not. That’s a lot of potential breakage for the monopoly. 5% of new PCs got MacOS. Some replacement PCs got XP, but not many because XP’s share is declining rapidly. The slow decline of that other OS’s installed base hides a rapid decline in share of shipments.

Assuming web share=actual share, let’s do the accounting:

  • opening installed base of PCs = 1231 million
  • new PCs + 361million
  • scrapped PCs – Y
  • MacOS – 16.7million
  • “7” – 200million
  • GNU/Linux – Z
  • closing installed base of PCs – 1281

Using W3Schools’ numbers, (Open x 31.1% + 200)/close = 0.455 and (Open x 7.8% +16.7)/close = 0.088. We can solve these two independent equations for the opening and closing installed base of PCs. We assume that all 200million licences went to new PCs although some may have gone onto older PCs, the best possible situation for M$… Doing high school maths gives the opening installed base as 1231 millions and the closing installed base as 1281 millions. That leaves scrapped old PCs + new GNU/Linux PCs as 1281 – 1231 -361 +200 +16.7 = 95.3 million. There will also be old PCs converted to GNU/Linux. We can estimate how many PCs die of old age from the fact that even an old PC has some value and recyclers usually get a crack at them before they are scrapped. The recycler of working but old PCs here is Computers for Schools in major part. They distribute 8 year old PCs free of charge. PCs less than 8 years old will be repurposed rather than scrapped. 8 years ago was 2004 when the world produced about 200 million PCs per annum. Clearly, a lot (~100 million) of older PCs are being recycled with GNU/Linux and not destroyed. Any way you count them, a lot of PCs went to GNU/Linux in 2011 and more will go in 2012.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology. Bookmark the permalink.

107 Responses to 4 Days to Eliminate Lock-in

  1. Dr Loser says:

    @Robert:

    As an addendum to that “bog-standard Windows box” thing, here’s my experience of X in the ’90s at Visa.

    Now, Visa is not now and never has been a M$ shop. They are, and were, open to pretty much any platform that did the job (with something of a bias towards IBM mainframes, which is quite understandable given their early origins).

    What did they choose to use as the “thin client” for their HP-UX mega-system dealing with the card authorization and reconciliation system world-wide? Well, they had to deal with this stinking pile of offal called X. So did they buy X hardware?

    Hell no. Each and every one of us had a Windows 95 computer with Hummingbird on it. Even if we never did anything else with that Windows computer, it was still far cheaper to dole these things out than it was to buy actual X “small smart thingies.”

    Why? Because the genius of X is that it makes “small smart thingies” impossible.

    Hummingbird was a piece of crap, btw. These days I would replace it with either X-Mingw or X-Cygwin. But I still think I’d get enough residual value out of the rest of the Wintel box to make that extra $100 for the license worthwhile: X is just a minor annoyance on the side.

  2. Dr Loser says:

    @oiaohm:

    “Also you don’t get it either that the same binary possibly could be used in google nacl.”

    SimpleFact(TM): No, I don’t.

    Linux is not exactly famed for ABI compatibility.

    And since when is that particularly important, anyway? I don’t care what the checksum on my downloaded binary is. I don’t care whether Robert built it on his Beast in under thirty seconds, or whether it was churned out overnight by Google on a large-scale multi-platform build.

    You know why I don’t care? Because I’m not a bit-head. I recognise that there are no possible benefits to me of using the same binary all over the place. This stuff is supposed to be invisible to the user.

  3. Dr Loser says:

    @Robert:

    If you truly believe that X is “genius-level” then you have a far lower bar for “genius” than I do.

    Let’s leave aside the implementation (see Jamie Zawinski, hardly an anti-FOSS bigot, on the pain involved in rebuilding it from scratch). That clearly is, at best, dwarves standing on the shoulders of geniuses.

    The whole design is a monumental cock-up. The inversion of standard client-server terminology is no accident: it’s baked in to the thing. What you think of as a “thin client” is in fact a “fat server” … a reality that was pounced upon with glee by 1980s and 1990s *nix hardware manufacturers, since it allowed them to sell vastly overpriced remote terminals simply because no “thin client” was possible. In many cases the terminal hardware cost more than a bog-standard Windows box, which any rational person would regard as an absurdity.

    The security holes in X are legion. The instability of X remains to this day: have you ever seen an X session recover itself cleanly? And in order to target any application (which will reside on what X calls a “client,” which is in fact a hugely powerful server), you basically have to rely on something like Qt to abstract away all the *tif nastiness and the “mechanism, not policy” gibberish.

    I have a lot of sympathy with oiaohm on this one. If I truly believed his Utopian vision of Wayland wiping X off the face of the map, and Qt/LLVM/gcc/whatever making the port of applications from *tifiness to HTML5 an entirely painless process, then I would feel a lot more hopeful about the future of the Linux desktop.

    Unfortunately, and to borrow another commenter’s term, it’s all bushwah.

  4. oiaohm says:

    Dr Loser really you don’t get it. GTK and QT toolkit conversions the applications work on wayland X11 and html5 from the same binary.

    You don’t need to rebuild them. The number of applications effected is close to 20 000.

    Also you don’t get it either that the same binary possibly could be used in google nacl.

    http://devworks.thinkdigit.com/Features/Qt-5-Coming-for-Googles-NaCl-as_8287.html

    Also using a compiler to build the same program in javascript Emscripten project is to put a javascript backend on llvm and gcc so c and c++ can be built to javascript.

    So gtk and qt application goes in out the otherside comes a pure html5 application or a html5-nacl application or html5 remote application or X11 application or a wayland application.

    One source base 5 platforms.

    Reason for html5 is remote desktop. Something every client these days comes installed with by default.

    Gnome has always been LGPL so no requirement to give up source code for a closed source application built on top oldman. If you built you application to only KDE libraries under KDE you did not have to give up your source code either. This is the way it has always been.

    QT has been like annoying roadblock that people have been able to hold up as some phantom shadow not to release applications on Linux. That is gone these days.

    Basically closed source has been welcome for years oldman. Just they are not taking up the invite.

  5. oldman says:

    “ou are committed to stripping FLOSS of every achievement it has ever made, every gift it has ever given to you and the rest of us. .”

    The only thing I am committed to is answering the bushwah that Pog repeatedly posts on his blog. I and others remind him and twits like yourself that there is another side to the story.

    Besides to be blunt. My conversation is with Robert Pogson, not you a$$wipe.

    And robert Pogson, the last time I looked, could take care of himself in a discussion. And I he fgets tired of me or feels that I am too far out of line, he a can ban me, and then you will be rid of me.

    Nothing that I say is going to take away what you and Pog believe in, just like all the nasty cracks you can make are going to stop what I say.

    Frfom this point on, if I feel that you have a point worth answering, I will do so. But any other crap that you choose to direct in my direction is going to be answered with the 4 letters that they deserve.

    STFU

  6. oldman says:

    “Insults too strong for this blog are too good for you sir. If my words upset you then sit down with a Microsoft EULA and read it. I’m sure it will make you feel better in the end.”

    STFU

  7. Kozmcrae says:

    “But the other shoe to drop will need to be that closed source commercial desktop software will need to be welcome on terms other than source code surrender.”

    I use closed source software with my distribution of GNU/Linux. There’s no problem with using closed source software with GNU/Linux. No one has to “surrender” their rights.

    The FUD that is spread about “losing rights” has nothing to do with rights. I has to do with complying with the GPL. People want to lock up the GPLed code. They don’t have the right to lock up other people’s code. Those are the “rights” they are talking about “losing”.

    But you know all this and you’ve been told all this many, many times. You don’t give rat’s ass sir. You just keep on playing dumb. Over and over again, repeating the same crap. You are committed to stripping FLOSS of every achievement it has ever made, every gift it has ever given to you and the rest of us.

    Insults too strong for this blog are too good for you sir. If my words upset you then sit down with a Microsoft EULA and read it. I’m sure it will make you feel better in the end.

  8. Dr Loser wrote, “X was doomed in the first place. It’s one of the great weaknesses in the *nix world-view that it was even allowed”

    Absolute crap! X window system was a genius-level development of *nix. It eliminated the need for Citrix to even exist in the *nix world. That other OS was absolutely limp on a network compared to *nix. X came to be about 1984 when M$ was still wearing diapers and had no concept of a network at all and *nix could do the GUI over the network. M$ did not get RDP going until NT and it waited years before it got to XP. Lose ‘9x had nothing like it so *nix was a decade ahead of that other OS. M$ did not even become serious about thin clients until a few years ago while *nix has been cool with them from the beginning of the GUI.

  9. oldman says:

    “You are still attached to the old FUD @ldman. You are not even keeping pace with ways the Microsofties are attacking FLOSS. No one is associating GNU/Linux with geeks anymore. Do you still have a disco ball in your living room? That question would be in line with your remark about geeks.”

    Look jerkwad I am well aware of the non geek user that have come into the community. In fact they have been coming since the late 90’s They are the reason that we HAVE a desktop like KDE. the problem is that its still the geeks that write the code and controlled the show. If it is indeed the case as oiaohm says that the techno geek worlds is coming to an end on the desktop, then we have some new possibilities.

    But the other shoe to drop will need to be that closed source commercial desktop software will need to be welcome on terms other than source code surrender.

  10. Dr Loser says:

    @oiaohm:

    A back-end is not a cross-compiler, young man.

    If you’re going to convert the 3,000 or so Linux apps from X11 to HTML5 without benefit of such a thing (I specifically mentioned a resource cross-compiler, which I think would do 80% right and involve a bit of fiddling with the remaining 20%, as always), then I wish you well.

    It’s a thoroughly doomed project unless somebody with an awful lot of money spends an awful lot of time on it, though, isn’t it?

    For reference, X was doomed in the first place. It’s one of the great weaknesses in the *nix world-view that it was even allowed head-space in the first place.

  11. Dr Loser says:

    @oiaohm:

    A back-end is not a cross-compiler, young man.

    If you’re going to convert the 3,000 or so Linux apps from X11 to HTML5 without benefit of such a thing (I specifically mentioned a resource cross-compiler, which I thing would do 80% right and involve a bit of fiddling with the remaining 20%, as always), then I wish you well.

    It’s a thoroughly doomed project unless somebody with an awful lot of money spends an awful lot of time on it, though, isn’t it?

    For reference, X was doomed in the first place. It’s one of the great weaknesses in the *nix world-view that it was even allowed head-space in the first place.

  12. oldman says:

    “Linux could have one desktop and the usability still be crap. Solaris proved that was possible with cde.”

    So instead you have babel.

    Bushwah.

  13. oldman says:

    “The major dispute in the battle was the means to make commercial applications without paying. Not really something pure FOSS people should be worrying about.”

    Fair enough. Then I look forward to the unified desktop framework that will soon be standardize the linux desktop. and the raft of closed source commercial desktop applications that should follow.

    Let me know when it comes.

  14. Dr Loser wrote,“Your “thousands” don’t mean diddly-squat to me. I’m afraid I flat-out don’t believe you. You can introduce a mere dozen people to any given GUI, and at least three of them are going to have problems of some sort.”

    At Easterville I sprung GNU/Linux on 500 students and 40 teachers. Only one needed hand-holding, an old guy who likely needed a secretary to do anything with a PC, a completely non-technical geek. In 2010, I sprung it on 14 teachers. Not one asked how to point and click. It’s a GUI. It’s meant to work the way kids use their fingers, touching everything.

  15. oiaohm says:

    JairJy of course you are not aware that most of the alterations done for that test are now standard in KDE.

    “including removing useless applications for the test, renaming and organizing the other apps for easy localization, adjusting the file manager for easy access to system locations”

    Funny enough they did not rename one application on the Linux desktop. They changed the application descriptions. Anyone who used freedesktop environment for a while knows this as a difference they like.

    The idea that the kde 5.0 equal to Windows 8 alpha is not out now is kinda wrong as well. That is kde 4.8 would match up to Windows 8 alpha. So yes users can look at both now.

    Really I did read the article. JairJy its the last case that someone did a proper usability study.

    Yes the only study to back up all you claims of bad usability is 2003. You should want something more recent. Lets see if there has been change for better or worse. Only way to know is a proper study.

    Simple fact is I don’t know if KDE 5.0 will be better than Windows 8. Really doing the study will at least make comments about usability based on research not guess work.

    If people got out there and did reviews based on Usability reports more pressure would be on the 7 desktops to fix them.

    Linux could have one desktop and the usability still be crap. Solaris proved that was possible with cde.

    Number of desktops is not really a issue. How people talk about and review them is a major issue. That 2003 document is really what a good interface review should look like.

  16. oiaohm says:

    Dr Loser cut and paste is the fact X11 created 4 different ways todo it.

    “So, if I understand you correctly (always a crap-shoot, but I try), your belief is that HTML5 is going to replace X11. That’s an, ah, interesting and unusual viewpoint.

    Let’s hope they have a really, really good resource compiler to convert between Lesstif and HTML, shall we?”
    Lesstif at this stage it does not have a HTML backend. Might never get one its that rarely used.

    QT and GTK already do have html5 backend options with no X11 server. The Qt prototype http://www.meegoexperts.com/2011/12/anatomy-app-qt-developer-days-qml-app-html5-web-app/
    Production release GTK with HTML 5. http://www.webupd8.org/2011/09/gtk-32-released-with-html5-allows.html GTK will cover wxwidgets.

    Opengl support will take a little longer that needs the same requirements as wayland so you can process locally and send video stream so the opengl does not lag out badly. Its also the same support you need to run multi video cards not crossfired or equal and move rendering programs between screens.

    But yes its only a matter of time before you will be able to put a Linux like desktop as a html5 website. With the tool-kits rendering straight to html5. Of course items like lesstif apps would not run at this stage or would have to be wayland wrapped.

    X11 is going away. Wayland will be local. Wayland is what vnc and rdp feeds and other things most likely will come from. html5 will be straight from tool-kits.

    X11 has officially been announced as going to die by x.org lead maintainer in fact.

    Yes X11 is on death row the question is how long before we can put it in the chair and throw the switch on it. Wayland is x11 replacement for local and wayland does not have a networking protocol of it own.

    Scary part about wayland is there is no more frame-buffer terminal vs graphical server. Wayland can command it all. This is why when wayland hits chroot backwards compatibility of Linux can go up massively.

    Set the chroot contained X11 server to render to framebuffer and have framebuffer place that in a window on the desktop by wayland. So old binary are restricted to software opengl but work perfectly with the X11 server they know and love. That takes you back to 1995 with almost perfect support.

    Modern day X11 servers can be running contained in wayland with full opengl support. Rendering to a window even that they think they are rendering full screen.

    Yes wayland also depends on the in kernel memory management of video cards and other things. So a lot more like Windows way.

    Yes qml for graphical design in qt assists in making qt more html5 friendly.

  17. JairJy says:

    Oiaohm, did you read the usability esay you send me? It’s a joke! Not only Windows XP won all but one of the test, but the test designers did have to configure the Linux interface a lot to make the desktop environment more usable (including removing useless applications for the test, renaming and organizing the other apps for easy localization, adjusting the file manager for easy access to system locations) and even with those changes the tester where more comfortable with XP.

    And now you think KDE 5 will be better than Windows 8? At least users can test Windows 8 right now. KDE is a great DE, very flexible, but lacks usability. While the Linux developers spend time and effort maintaining 7 destop environments, Microsoft, Apple and Google are spending money on simple, universal, and usable interfaces.

  18. oiaohm says:

    oldman the problem with the desktop war on Linux is not the number of desktops.

    It the fact you change desktops it don’t work. Application from different desktops have different themes on the same screen. Different locale settings different password storage….. list goes on and on of differences causing problems. This is the section of the battle that is ending.

    Besides Windows 8 is shipping with two desktop environments so it is dead???

    Really in the end kde and gnome will be very much like Windows classic and current modes. Users will be free to choose yet not have application issues mixing and matching.

    Multi WMs is not a problem as long as everything works right. Current issue has been that change WM and everything could go south.

    Oldman there has been a basic peace treaty signed to end the war in the background ABI/API of the desktop.

    So the geek warfare you are talking about is over.

    Basically by your standards no version of windows from windows 9x on should have sold because they all had the option of two different windows managers.

    Your statement is insane oldman.

    I guess you missed the putting down of arms by the backend developers. QT converting to LGPL did this. Now GTK and QT are the same license. License Nazis had no more reason to say hey cannot integrate with QT due to license. Also the reverse is true as well. So no more case to keep libraries independent.

    Yes there was code in KDE libraries not up-streamed to QT due to its license as well.

    Asking for the wrong things oldman. You did not want 1 WM. You wanted 1 license for each section of the ABI/API making up the desktop core ABI/API. So sharing is practical and not require sending to legal department to find out if a particular area can be shared.

    Mess on Linux Desktop has been coming from a licensing mess in background that died a few years ago. What was MS advantage not a single wm because if you do look close enough that is not true. That the core API/ABI is under a single license.

    Almost everything you want will come true due to licensing change oldman. Something good nokia will be remembered for changing the qt to LGPL so making the licensing common between the two major toolkits on Linux. Yes Nokia is the one that made the peace treaty on the Linux desktop.

    gagl support of qt is also because of licensing change of qt as well.

    Thing is the rationalization that is possible will take time to play out. I expect fairly large sections of gnome and kde back-end code to end up deprecated. Now that the road block to common standards is gone.

    Fragmentation of the Linux desktop had a cause.
    http://en.wikipedia.org/wiki/List_of_widget_toolkits

    Really read the high level widgets oldman. Only one of the high level widget toolkits have anything license nazis worry about. QT has not been updated to LGPL on the first list is in the last list.

    Juce is the last of the high level with a license that can cause incompatibles with coder requirements. There is no windows manager for Linux using it.

    Yes the day of call the lawyers to find out if they can share code is over.

    The major dispute in the battle was the means to make commercial applications without paying. Not really something pure FOSS people should be worrying about.

  19. Kozmcrae says:

    “the people who you are going after do not care about the IMHO geek internecine warfare that is behind the different window manager.”

    They are not geeks dipstick. They are people like you, me and Robert. People in Enterprise and the whole nine yards. For you to associate window managers with geeks at this time reveals you to be either completely out of touch with the reality or to be a crusty old fool who is unable to change.

    You are still attached to the old FUD @ldman. You are not even keeping pace with ways the Microsofties are attacking FLOSS. No one is associating GNU/Linux with geeks anymore. Do you still have a disco ball in your living room? That question would be in line with your remark about geeks.

  20. Dr Loser says:

    @oiaohm:

    So, if I understand you correctly (always a crap-shoot, but I try), your belief is that HTML5 is going to replace X11. That’s an, ah, interesting and unusual viewpoint.

    Let’s hope they have a really, really good resource compiler to convert between Lesstif and HTML, shall we?

    Or is this going to be another of those daft chroot-style ideas where X never really goes away; it just gets encapsulated in an HTML5 carrier?

  21. Dr Loser says:

    @Robert:

    That’s just the problem: the Linux desktop is not a GUI, at least not in a modern sense. It’s a thin facade over a collection of command-line tools, and it shows. Windows and Snow Leopard are actual GUIs, talking directly to the system libraries.

    Even people who understand this thin facade stuff have difficulty dealing with the consequences (random cut’n’paste errors; an invisible bit of underlying Python going away when I reinstall a package for my own use; massive failure of the CLI for an audio app resulting in a pop-up window with frenetically scrolling error messages; these are just my own experiences).

    Your “thousands” don’t mean diddly-squat to me. I’m afraid I flat-out don’t believe you. You can introduce a mere dozen people to any given GUI, and at least three of them are going to have problems of some sort.

    For the record, incidentally, I believe the desktop Windows manager has been more or less the same, and across all the (I agree with you on this) stupid Home and Premium and Wazzock packages, since about Windows 2000 if not before. The skins tend to look different, and there’s the occasional extra like the Start/Search For box in 7 (or was it Vista?), but we’re not talking Gnome and KDE insanity here.

    Windows 8 is shaping up to be a bit of a disaster in this area, however, so you’ll at least have something to crow about in a year’s time or so.

  22. oldman wrote, “To the non Linux computer user versed in windows, two desktops is one two many.”

    That’s bogus, oldman, M$ constantly tweaks its desktop so they can sell the “new” version. I think Lose 3.1 had one or two versions. Now we have “7” in 7 shades of yuck and “phoney 7” and Vista debugged, and vapourware “8”. How many different window managers has it been?

    I have introduced thousands to GNU/Linux and none has had any problem using KDE, GNOME, XFCE4 desktop environments and several window and display managers. It’s a GUI, for Pity’s sake. It’s designed and intended to be used with little or no instruction. Just point and click.

  23. oldman says:

    “Of course they are not going to get rid of gnome and kde completely.”

    And as far as I am concerned that in the end is their downfall. the people who you are going after do not care about the IMHO geek internecine warfare that is behind the different window manager. To the non Linux computer user versed in windows, two desktops is one two many.

  24. oiaohm says:

    This is part of KDE 5.0 oldman
    http://community.kde.org/KDE_Core/Platform_11/Eliminating_Duplication_With_Qt

    Yes lots of the stupid KDE only stuff disappearing. With KDE 5.0

    Wayland sees all the X11 stuff other than for remote work disappear with html5 coming in all major toolkits X11 for remote might be pointless machines have a web-browser why install a X11 server as well. So X11 gone you are most likely going to get your wish oldman with X11 out of there.

    Really there is not a single application development framework on windows either oldman.

    Reducing down to KDE and Gnome down to GTK and QT that have to get along will be an improvement. Since this reduced the number of theme elements in existence to set.

    With services like KDE Wallet and Gnome Keyring for password storage. Are merging Secret Service stored in dbus.

    Yes they are pooling there resources one service at time. oldman we both would love the process faster.

    So yes one API to talk to all background desktop services exists. The issue is the old problem how to disconnect from the old and connect to the new without causing a stack of applications to fail and without losing features.

    Common control ABI to system wide servers is a little bit more of a debate. upstart vs systemd both contain a dbus interface. Ubuntu backing upstart everyone else backing systemd. Even better systemd and upstart interfaces only appear based on what one is running in dbus. So max of 2 dbus probes to find out what the service management system is of course if you are allowed access to it.

    We are also seeing like gagl more frameworks made without a toolkit preference.

    Basically oldman they are doing what you want just they cannot turn instantly. The old issue of backwards compatibility. Of course they are not going to get rid of gnome and kde completely. You still need competition to try out new features.

  25. oldman says:

    “Really it would be good if someone does a KDE 5.0 vs Windows 8 like report.”

    What would be really good would be if we the community cut the geek bushwah and pooled all their resources to create a single desktop framework With stable and common API’s for all services – bye bye Gnome and KDE and all the X crap, hello Linux standard Desktop.

  26. oiaohm says:

    JairJy those Gnome, Canonical and KDE usability teams are still there.

    Remember KDE usability team was sane enough to get outside assistance. Leading to Open usability.

    Read the gnome guidelines 1.0 you will see it has improved. They wrote those without and advice from design people. Coders design usability you are stuffed.

    Gnome has a major on going battle over it. KDE what I mostly use is on the right track. Pick application do assessment on it and improve its interface.

    JairJy I did not say it was enough. Openusablity has what you call a backlog of applications to cope with.

    2003 KDE was wanting an outside review of how good there usability teams were. Basically got told they suxed. Gnome still needs to put self up to a proper outside review.

    Canonical do I outside review I would be shocked.

    Basically openusablity is there not to be bias.

    2004 was a start to attempt to turn around over 20 years of bad gui design idea in open source.

    Yes just because KDE started openusablity Gnome guys were reluctant at first to deal with it. Usablity Gnome is lagging.

    Yes that KDE vs Gnome war does have its downsides.

    JairJy really the 2003 report kde vs XP is still a good read to show methods. http://www.relevantive.de/linux_usability_report_en.pdf

    Really it would be good if someone does a KDE 5.0 vs Windows 8 like report.

  27. JairJy says:

    Oiahom, Open Usability is an improvement to the poor usability managenent on the usual Linux applications, good job. It’s not enough. GNOME, Canonical, even KDE has their own usability teams long time ago, but even with those teams they made serious usability and design mistakes. Why? Because the eternal battle between programmers and designers, and because poor usability guidelines (just read the GNOME usability guidelines and you will get the idea).

  28. JairJy says:

    Canonical cann’t make any money if Shuttleworth spends $10 to make $1. C’mon Mr. Pogson, is basic economics!

  29. ch says:

    Please tell me how this is a lie: “Cannonical makes no money.” when Mr Shuttleworth says himself: “I’m confident that we have a reasonable CHANCE of GETTING Ubuntu to the point where it funds itself for ongoing growth.” (Emphasis mine)

  30. Phenom lied, “Cannonical makes no money.”

    Shuttleworth’s money would have been gone long ago if Canonical was not making money.

    “We have some initial revenues from services related to Ubuntu. We have been contracted to produce customised distributions, and we participate in large-scale tenders for big Linux deployments, usually in partnership with local companies, where our job is to provide escalation support. In addition to widespread adoption in developing countries, Ubuntu may well be running all over NASA’s Moffett Field soon… So we have the foundations of a sustainable project, and I’m confident that we have a reasonable chance of getting Ubuntu to the point where it funds itself for ongoing growth.”

    That was written by Mark Shuttleworth on his blog in 2011. I know they have salesmen. I talked with one a few years ago when I was implementing the system at Easterville. The question was how to make the system sustainable and one option was to hire Canonical to provide some level of support, particularly for servers as there was little expertise except me in-house. There was no money to hire Canonical so I provided training and documentation for my successor. I needed to provide some advice once or twice thereafter. Canonical could have done the same.

  31. oiaohm says:

    Basically gimp is migrating from a model that could never compete with photoshop to a model that does threaten to unite the complete FOSS graphical development world.

    Complete FOSS graphical development world could serous-ally kick photoshops but. Filters from gagl might even find there way into blender and video editing.

    This change can effect everything that handles images moving or not moving in the FOSS world. So they all can produce filters and extensions they all use.

  32. oiaohm says:

    “million-programmer-army” can work but there is a catch.

    Apache and jackaudio are examples of the million-programmer-army model that work.

    You design a system where small segments can link up to produce a greater object than any one alone is.

    You are correct about the complexity bit. Even the early Linux kernel with its modules was design that the million programmer army could work.

    So early Linux is million programmer army model. Goes through a time of BKL infection. Now it is returning to a model of a million programmer army but now they care company funded.

    Audio and graphics are in fact largely suitable for the million programmer army model.

    Think your vst filters they are million programmer army model. Any particular reason why vst and effects processing systems cannot be done the million programmer army model. They can be broken down into small simple parts.

    http://gegl.org/ this is the key alteration happening in gimp. Result is a universal graphic filter and modification provide engine that will be shareable with all applications.

    Of course small picture you compare Photoshop to gimp failing to see the bigger picture is altering.

    Result is gimp, krita, digikam and the rest of the open source graphical processing systems will be able to share tools even share and be working on the same image at the same time. After this point a developer no longer has to produce a super tool that does everything. The developer can now produce a small module that does something very great. If that module is beaten by someone else you can just swap it out.

    gegl has been specially designed without any toolkit dependance.

    This is why I don’t call Cubase anything more than a toy you cannot simply pull parts out and use better options. This is where you don’t get jackaudio effect when fully configured and connected. If you look at each jackaudio application alone you find weaknesses. Problem is each of those weaknesses don’t have to be used. You use the strengths of the applications you have.

    Apache is the big great program that people know that was designed by million programmer army. Method.

    Yes image processing is just starting down the path of being million programmer army compatible. The change coming to the Linux image processing world is as big as when someone decided to create vst for audio processing effects if not bigger.

    Audio effects work and Web services on Linux is million programmer army compatible.

    Cpan extensions to perl is another example. Dynamic libraries to break the work load down.

    Basically there are many areas where the million programmer army works.

    Design is key basically. If it one huge blob design it is not million programmer army compatible.

    Office suite and Webbrowser they can be a little hard to break down.

    Android if you look closer most of it code base is in fact not made by google or anyone else working on android. It is another million programer army base with a more restricted top.

  33. oiaohm says:

    ch not true example.
    “KDE 4.0. It had to be rushed out because the developers only wanted to do the fun stuff (writing new code for 4.0), so nobody maintained 3.x.”

    Maintainer-ship of 3.5 crew worked past the 4.0 release. The big disaster was a disconnect between developers and marketing teams in kde 4.0.

    As trinity that has tried to maintain KDE 3.5 and keep it going found out the internals of QT3 have bad limits. This effectively rendered the KDE 3.5 code base a death sentence. Only way to fix KDE 3.5 was to port to QT4 to fix the bugs equal rewrite all the applications.

    Yes KDE 3.5 was technically not repairable and still is technically not repairable.

    This information made marketing presume the developers with kde 4.0 would be rapid change over.

    In fact the lead of kde 4.0 was still doing repair work on 3.5 at time of kde 4.0 release.

    KDE 3.5 maintenance team was not pulled until end of 2008 26 August 2008. That was after the july when KDE 4.1 was released.

    Also after when Timothy Pearson had decided to make trinity. Basically if he wanted maintenance of it he could have it.

    Maintenance of KDE 3.5 was not stopped until a new maintainer had stepped up.

    As Timothy Pearson is being forced to port section by section of trinity to qt4 so it can keep on working. KDE 3.5 was dieing the death by millions of cuts.

    In the end trinity will end up not compatible with old KDE 3.5 applications.

    The reality is KDE 3.5 was a zombie even before the idea of KDE 4.0 was formed with a growing list of non repairable bugs coming from the design of QT 3.

    KDE developers with KDE 4.0 were hoping for a chance to start over and have time to design a new solid base. Due to people wanting to use it quickly they did not get that chance. KDE 4.1-4 had to be rushed adding features. Due to bad release handling.

    Looks like there goal of a good sold base will at long last be KDE 5.0.

    Basically KDE 4.0 is not a failure people to maintain. It is a example how not to handle a case when you have by a error made a code base that simply never can be fixed in any suitable way.

  34. ch says:

    “ch wrote some crap about GNU/Linux desktops.”

    Mr Pogson, just because you don’t understand business models doesn’t mean they are crap. Obviously you prefer the “million-programmer-army” model of development, what I classified as “hobbyware”. Yes, that’s a good model for producing tools up to a certain level of complexity, like that password manager you mention (mostly done by two guys, it seems). However, the available evidence suggests to me that the MPA approach doesn’t work well for more complex stuff like office suites, image manipulation and other areas. Look at the FLOSS success stories: Linux the kernel, Android, OOo/LO and Firefox are all essentially done by companies (with some input by volunteers). JairJy already explained the reason for this. Example: KDE 4.0. It had to be rushed out because the developers only wanted to do the fun stuff (writing new code for 4.0), so nobody maintained 3.x.

  35. ch says:

    “Sorry Cubase is a toy.”

    Ok, that’s it. Keep your illusions, I’m done with this discussion.

  36. oldman says:

    “Software development for Linux IBM does as well. If it service and it relates to Linux IBM does it.”

    IBM has always had a large service organization for a price. They also have no problems selling and supporting their customers on closed source commercial software as well. Remember Mr. Microsoft VAR, the customer is always right.

  37. oiaohm says:

    Phenom really out of all Novell departments the only one that was making money was the SUSE support contracts.

    Without SUSE Linux Novell would have died years ago.

    Cannonical in fact does make some money problem is they are a private company so are not require to report so much.

    Redhat has crossed the 1 billion dollars a year. Suse most likely would be in the same boat without Novell miss management.

    Boy you don’t know IBM. You have the IBM support line for Linux. What is a 24/7 call service for advice for you business with option of onsite staff been sent out. SLA so IBM makes profit there.

    Software development for Linux IBM does as well. If it service and it relates to Linux IBM does it.

    http://www.ibm.com/linux/ Really Ibm does very well out of Linux thank you very much.

    What cannot accept the fact that there is money to be made from Linux. Phenom HP and Dell also run Linux departments. If you are selling servers you Linux support services are one of your more profitable departments.

  38. Phenom says:

    Pogs, IBM made money out of their hardware and DB/2, plus a bunch of other enterprise apps. Linux is just something they got cheap after they screwed up OS/2 in more ways than I can count.

    Cannonical makes no money. It only spends the money of Shuttleworth. RedHat can’t even compare to the big guys, and Novell’s affairs are far from brilliant.

  39. oiaohm says:

    NT JERKFACE really redhat does share the patches upstream.
    “Even Red Hat, paragon of open source intentionally makes it difficult to incorporate their security patches. What the hell does that tell you? Sharing your intellectual work with your competitors is not a good idea.”

    Redhat has made it harder to take the patches from the downstream side. Not impossible but to work out what patches redhat has applied you have to compare the fused to what has been sent upstream.

    So yes redhat still shares everything with it competitors include its prototype patches.

    Result redhat has not made it any harder to incorporate their patches. Anything already sent up stream is not displayed independent. So this means you cannot go to redhat and snoop. Instead have to up to the upstream source and snoop.

    Redhat by forcing everyone upstream they are getting to read reviews of the redhat patches by other parties. That tell you that some of the redhat patches are not exactly helpful and are only addressing redhat particular distribution issues.

    So basically redhat got stick of people just coping and doing no work and no research.

    NT JERKFACE did you miss the memo over MS Office. MS Office Starter is given away for free due to heavy competition.

    The Gimp has taken out a lot of applications. Just not photoshop. There were a lot of web designer targeted image editors most have gone.

    NT JERKFACE something important here the community of open source around a project defines what comes out. For a longtime The Gimp was infested with web developers. There are a lot of things web developers don’t need that people printing stuff do. So that stuff was not important. Who needs more than 8 bit per channel color no monitor displays more than that.

    Digikam that was people wanting to print and manage photos you can see way different outcome on features.

    There is more focus today than ever before by libregraphics projects and other items to bring gimp and others up to pro grade for all classes of users.

    NT JERKFACE its shocking right up until about 2009 there very little care about winning for print. For web based work gimp and photoshop are about equal for results.

    This is where the problem comes from. Web developers someone asks for a replacement for photoshop they get told gimp what is the correct answer if you are a web developers. Problem here gimp was not setup for physical printing so hello splat. Photoshop complete defeats gimp in that area. There has always been better programs for physical printing than gimp on Linux.

    Time is changing NT JERKFACE I would not be surprised if we don’t see more free versions of adobe products like MS office as done to try to slow the change.

  40. NT Jerface wrote, “You guys *want to believe* that a company like Adobe can embrace open source and money will somehow magically flow to them. Sun tried that strategy and look at how far it got them.”

    IBM and Canonical and RedHat and Novell embraced GNU/Linux and made tons of money from it.

  41. oiaohm says:

    JairJy head in sand much.
    “The problem is that everyone wants to do the
    fun part of the job, and in those 10 hours of work
    nobody is looking for usability and design issues.”
    http://www.openusability.org/

    This project started in 2004. These are decanted interface designers and researches working on FOSS most had a commercial life before being aquired. I do mean aquired they are paid for. They are the ones that design the layout of gimp new single window interface. The coders you are counting only implemented it. Most of the people who like coding admit they cannot design a interface to save themselves. Then most of the people who are good interface designers cannot code to save themselves.

    So cause of mess in open source applications is being addressed.

    Yep some of the blame for the single window mode on gimp delay “the final design spec is also missing at the time”. Is that the final design for the coders to implement had not come back from openusablity as quick as expected either. Lot more study was required to design a interface to implement for the long term with the required expandability and not turn back into another dog breakfast.

    Gimp developers have not been designing the new interface. Also the 2.5 developer are gimp core that GUI and filters is prime. This is not the team working on the new backend gegl.org. The team working on gimp is larger than the gimp team for full-time developers.

    Yes full time lead developers assist a project no end. Gimp has a major issue of not having enough full time leads to run all the projects required at full speed.

    If you look at most FOSS applications the GUI side has improved since 2004. This is related to direct work by openusability and there indirect work. Openusability makes sure there are people at every Linux conference to talk to developers about interface design. This is indirect work.

    Of course some applications with poor interfaces there developers really should be sent to openusability for training or assistance.

    Openusablity started at a KDE conference 2003 when a commercial interface designer was asked to directly compare KDE and XP for usability. The KDE developers were kinda shocked how bad they really were at interface design. Lot of what the interface designer was pointing out was fort at first. Some of good usability goes against what the code base looks like under it.

    Issue is that coders will try to make the interface look like how the code base is under the interface because this makes it simpler to find section in code to fix at times to the suffering of the poor user because stuff is now in hard to get at locations. This is why interface designs and coders are rare to find as one person that is good at both.

    Interface design is a fun part to people with they attributes to it. Those people also normally will hate coding because logic requirements of computers will force them to layout code that does not match the attributes that make them a good interface designer.

    Key thing here is have interface designers able todo what they like. Also have coders able to do what they like.

    Its taken a fair while for coders to start accepting in the FOSS world they have to take some direction from interface designers. These days its basically status normal.

  42. NT JERKFACE says:

    Robert Pogson wrote some crap about some shit
    So, creating the GNU/Linux desktop pays handsomely and everyone gets to share. The fact that there are so many varieties of GNU/Linux desktop proves that.

    Linux Mint is the #1 distro and they only get enough in donations to pay Clem, and last I read about their budget he was being paid less than the average UK developer.

    One major problem with software is that you can’t divide the work infinitely. The Linux kernel is not developed by a million people working 10 hours a week, it’s mainly a corporate project worked on by paid developers in cubicles at companies like Red Hat and IBM. They have an incentive to work on the kernel but not the desktop as a cohesive package and this has been apparent in the quality of each for years.

    If even half of what you guys believed was true Adobe would be bankrupt because of The Gimp and M$ would have stopped charging money for Office long ago. You guys *want to believe* that a company like Adobe can embrace open source and money will somehow magically flow to them. Sun tried that strategy and look at how far it got them.

    Even Red Hat, paragon of open source intentionally makes it difficult to incorporate their security patches. What the hell does that tell you? Sharing your intellectual work with your competitors is not a good idea.

  43. JairJy says:

    “The problem is that everyone wants to do the
    fun part of the job, and in those 10 hours of work
    nobody is looking for usability and design issues.”*

    Never write a comment while sitting on a mexican bus.

  44. JairJy says:

    So, creating the GNU/Linux desktop pays
    handsomely and everyone gets to share. The
    fact that there are so many varieties of GNU/
    Linux desktop proves that.

    The problem is that everyone wants to do the fu. parf of tbe job, and in those 10 of work nobody is looking usability and design issues. That’s the main reason ahy we have more than 300 distros, 80% of them are the same but with a different wallpaper and theme; but only there are 2 main developers working on GIMP in their free time.

    Instead, if you hire a programmer, you will have someone who will work 48 hours a week on a software, and you call tell him to do the dirty work on debugging and stuff. You can hire a designer to improve the software and hire someone to do usability research. And of curse, now that you are paying them, you can force them to work together.

    Quality over quantity. Android couldn’t have been done with free handwork, and Google have money to hire because their propietary search engine and their business model.

  45. Thanks for the link, I repaired it. Supply and demand ebb and flow. Same with software. M$ would easily be able to have partners print another 50 million licences per quarter but their’s no such demand. M$ could lay off half their employees with no damage.

  46. oldman says:

    “Gimp is a large project. I count 216 contributors over 14 years. They are alive and well.”

    http://www.omgubuntu.co.uk/2010/12/single-window-gimp-release-delayed-but-why

  47. Phenom wrote, “the GIMP team. Just recently they complained of being short of hands.”

    A large project can always use more hands. M$ has been complaining they cannot get enough visa-holders to take US jobs.

    Gimp is a large project. I count 216 contributors over 14 years. They are alive and well.

  48. Phenom says:

    Pogs, you can say that to the GIMP team. Just recently they complained of being short of hands.

  49. NT JERKFACE wrote, “You guys don’t have a viable business model for the most common type of software and yet continually drone on about how open source is going to take over the world.”

    The most common stuff is supported by users. Users are in the millions and many are developers/programmers. Consider a widely used application that’s got lots of competition and it not a platform, say, Figaro’s Password Manager. It’s a pure commodity piece of software yet many people contribute to it because it’s useful.
    pogson@beast:~/Downloads/fpm2-0.79$ grep Copyright src/*.c
    src/blowfish.c: * Copyright (C) 2000 John Conneely except for portions taken from GnuPG.
    src/blowfish.c: * Copyright (C) 1998 Free Software Foundation, Inc.
    src/callbacks.c: * Copyright (C) 2000 John Conneely
    src/callbacks.c: * Copyright (C) 2008-2010 Aleš Koval
    src/fpm.c: * Copyright (C) 2000 John Conneely
    src/fpm.c: * Copyright (C) 2008-2010 Aleš Koval
    src/fpm_clist.c: * Copyright (C) 2000 John Conneely
    src/fpm_clist.c: * Copyright (C) 2008-2010 Aleš Koval
    src/fpm_crypt.c: * Copyright (C) 2000 John Conneely
    src/fpm_crypt.c: * Copyright (C) 2009 Aleš Koval
    src/fpm_file.c: * Copyright (C) 2000 John Conneely
    src/fpm_file.c: * Copyright (C) 2009,2010 Aleš Koval
    src/fpm_gpw.c: * Copyright (C) 2000 John Conneely
    src/fpm_launcher.c: * Copyright (C) 2000 John Conneely
    src/fpm_launcher.c: * Copyright (C) 2009 Aleš Koval
    src/main.c: * Copyright (C) 2000 John Conneely
    src/main.c: * Copyright (C) 2008,2009 Aleš Koval
    src/rijndael.c: * Copyright (C) 2000, 2001 Free Software Foundation, Inc.
    src/sha256.c: * Copyright (C) 2009-2010 Aleš Koval
    src/sha256.c: * Copyright (c) 2008 Damien Bergamini

  50. oiaohm says:

    Ch http://www.renoise.com http://www.energy-xt.com/ This is your problem.

    Ardour is no the primary DAW in that kit. Its one of 3 DAW’s in the kit

    2 are commercials. Very powerful commercials.

    Yes 100 dollars are for 3 DAW that could be used as primaries.

    You have made a fatal presume that I was talking about FOSS alone. What makes Jackaudio so powerful is the means to use applications in combinations. So design of DAW can be selected for task at hand. So you don’t have to use a round peg in a square hole.

    The 100 system can be used for anything from simple stage control to full blowing music production and you can change almost instantly. Simple fact jackaudio allows you to store application configuration templates and reload that configuration when required.

    There is a catch they are not talking about standard Ardour they way they talk about Ardour with the commercial extensions.

    Sorry Cubase is a toy. Items like I am referring to are not.

    Sum of the parts. Each part alone is a lot weaker that the whole. The power in the combination is not each application alone. It the power of everything.

    I guess this is most likely shock horror to you that VST is not only Windows and OS X. There are native Linux VST effects that are not for windows or OS X.

    Linux system I can use Linux VST and Windows VST. Currently developer of interface does not have a apple machine to make OS X wrapper. As well as the other native Linux effects options.

    Of course you want to pick one part out alone. This is weaker. Take on the full 100 dollars amount the cut back cubase is not even in the ball park.

    Sorry to say you have not worked with the best stuff out there. ch you only think you have.

    Remember I have used cubase. I know what it can do so don’t try a foolish bluff.

  51. ch wrote some crap about GNU/Linux desktops. The desktop is easy to pay back. Almost everyone uses it so many people are willing to contribute so that they can use a good desktop. It’s the purest FLOSS. As Linus said, “Imagine ten people putting in 1 hour each every day on the project. They put in one hour of work, but because they share the end results they get nine hours of “other peoples work” for free. It sounds unfair: get nine hours of work for doing one hour. But it obviously is not.”

    So, creating the GNU/Linux desktop pays handsomely and everyone gets to share. The fact that there are so many varieties of GNU/Linux desktop proves that.

  52. ch says:

    This is what the makers of Ardour have to say about their product:
    “As Ardour is getting more powerful and usable, we attract more and more users who expect the same feature set and product polish as they’ll find in a commercial product such as DigiDesign’s ProTools or Steinberg’s Nuendo. This isn’t the right way to think about Ardour at this time.

    Not that we don’t want to get there, you understand, but it’s simply not a reasonable comparison. DigiDesign has spent who knows how many man-hours worth of development on ProTools and can spend a lot on getting good documentation written, new features, debugging, installation process made smooth and generally polish the thing till it shines. In comparison, Ardour development is driven primarily by the interests of just a few people. Development is a full time job for the lead developer, who also raises a three kids, fixes up his house, has friends and even a relationship with a gorgeous woman.”
    http://www.ardour.org/support_expectations

    This is what your link wants us to believe:
    “Rivaling applications such as Pro Tools and Nuendo, Ardour is a brilliant yet elegant full featured recording application.”

    Yeah, sure.

    Re ReWire: It was developed specifically to let Reason plug into Cubase. That it does fine, and AFAIK some other apps support the interface, too. The main interface for plugging stiff together is VST (or AU on the Mac), and that works. And for €100 I can get the small version of Cubase that would suit my needs today. (Way back when I still had to buy the ful version at 600 Deutschmarks, but do you know what? It was worth it.) So enjoy your half-assed audio apps on Linux, I have worked with some of the best stuff out there.

  53. oiaohm says:

    ch jackaudio is really a patch table replacement. Networking it can join many computers or devices. Allowing digital mic feeds. Yes you can fit quite a few mics down a single bit of 10 G cat cable. Nice bit it resistant to outside noise sources once in cat. Yes and reduced cabling to set up.

    Also you did not allow on a fact I could be fairly lazy. http://indamixx.com/applications/audio-applications.html Mix of commercial and open-source configured to work with each other using jackaudio. Basically 100 dollars on Linux audio software goes a very long way. Remember cubase price tag is the same price I pay for a configured laptop already setup to go.

    If we have both exactly the same budget to spend on software ch using the linux solutions I am going to get more bang for my buck. More VST and AU filters. Basically same money I will be the one with the better tools.

    Yes the limited version is $49 for a usb stick version that the software can directly run from.

    Yes the Ardour you had is the limited version. Most people are not aware it has a commercial extension. Also not aware of reniose.

    “ReWire, an inter-app communications API for Windows and MacOS from PropellerHeads and Steinberg, ReWire is similar in that it provides inter-application audio routing, but does not allow for fully independent processes, and has silly restrictions (“up to 64 channels”, etc). JACK also comes without silly license restrictions. ”
    ReWire has a design restriction that makes it hazard.

    Does not allow fully independent processes this is critical. I am running live something fails on a feed I only lose that feed the other feeds make it with jackaudio. Something happens in ReWire and its house down. Live performance forget using what you are using basically.

    Also ReWire is not even close to jackaudio no network support to join up computers or devices. Instead you have todo horrid spif hacks if you want to join two computers running ReWire.

    Basically you should be pushing for jackaudio support in windows applications is a better quality solution than ReWire you are putting up with. Yes the FOSS solution is running rings around what the closed world dreamed up so far in audio streaming.

    I am looking forward to under 40 devices like rasberypi the will be powerful enough to encode a mic and send it into network also powerful enough to feed out to a speaker system.

    Yes the no commercial applications bit is crap ch. Audio applications Linux has tones of different applications. Some Linux only some cross-platform. There is in fact a lot of commercial out there for Linux that are Linux only for audio work. Also due to the long history of the jackaudio framework on Linux almost every single one supports it perfectly compare to rewire that is hit and miss.

    Gnome developers are making a stupid mistake with Dbus. “Most systems implement a privileged system channel, plus a private channel for each logged-in user.” The catch with d-bus is the private channel is optional to be free of policy-kit control. Only optional not to be controlled. Its even option that the private channel exists. Yes if I set in system wide policykit no private channel on dbus result is no private channel on dbus stiff briskets applications. Yes I expect one day to be able to turn gnome into a non working heap just by tweaking the policykit. KDE developers are more sane and their solutions will keep on working even against the worst policykit settings.

    So yes the user is fully inside there right to control everything in the private channel with D-bus. It is a secure bus design. This is what makes it major different to active-x and the prior com stuff. Even using the private channel on d-bus writing into another applications screen space directly may see you terminated by selinux for secuirty breach.

    D-bus is very restricted data transfer compared to kparts or active-x. Bonobo to D-bus will not work in all cases. dcop in kde was moved to D-bus. Kparts has been left outside D-bus for the particular reason Kparts does expect to be able to draw in the application window of the application calling it.

    So yes comparing Kparts to active-x is valid. D-bus to active-x is not valid they are major-ally different beasts due to one being a secure bus and one being a insecure bus. D-bus is kinda unique because it is a secure bus design.

    By comparing active-x to d-bus you get to fail to check windows for a secure bus that is currently still missing. D-bus design is also very much old Unix IPC with secuirty placed on top. Yes old Unix IPC is able todo lot of the things that com can. Even that it predates it by over 10 years.

    Its more than COM is designed off basic ideas of Unix IPC that D-bus appears kinda COM like. Closer inspection the Unix IPC nature shows out clearly. Processes in Unix IPC remain in completely split address spaces.

  54. ch says:

    Let’s talk business models. How can you develop FLOSS applications for the desktop and still pay the rent? I see the following possible models:

    – Hobbyware: You spend some spare time and develop some program, then give it away while your day job pays the rent. That’s nice and more power to you, but even the best developer can only do so much in his spare time (and few SW developers are good UI designers etc.). In a nutshell, that’s why Linux has a gazillion media players and window managers but no equivalents to Photoshop etc.

    – Supportware: You write the program and give it away for free, then you make a living from support. There are three problems with this model: Few people want to use desktop programs that need support they have to pay for. Few developers want to do hotline service all day. And finally, everyone else could do the same support – only cheaper because they wouldn’t have to recoup your initial cost for developing the program.
    This model seems to work for some server applications, but I don’t see much success on the desktop.

    – Consultantware: You write something as complex as SAP and charge for getting it to work in the environment of the customer / customizing it. Should work for very specialized business apps.

    – Add-on-ware: You write the program as something you can give away with the stuff that people have to pay for (e.g. hardware). The reverse model kinda works too: You give away your almost-Exchange-like SW and charge for the Outlook connector.

    – Sponsorware: You find a sugar daddy like Google to pay for your services and write SW with that money. Works as long as you can find a sugar daddy in the first place.

    – Outsourcingware: Several film studios put the money they previously spent for their in-house-developed tools into one pot that pays a team developing tools they can all use. Since this means the tools are no longer a competitive advantage, they may as well give the results away.

    – Donorware: You hope that users will give you some donations. Anybody heard any success stories here?

    OpenOffice/LibreOffice is in it’s own categorie, I believe. In the Sun days, it seemed to me like “Spiteware”: “We do this just to annoy M$.” Now LO is “Needware”: Ubuntu et al. NEED to support it because without a somewhat usable office suite, even the last person with a working brain will realise that Linux on the desktop is dead (except for geeks who don’t do office stuff).

    Did I miss any possible business models ?

  55. ch says:

    @oiaohm:
    When I asked around the last time (a little while ago, before Ardour added MIDI) what equivalent to Cubase I could use on Linux, the answer was “Ardour Jacked-together with Rosegarden”. If you do need to stitch together apps just for the basic functionality, then yes, I can see Jack being useful. However, here on Windows (or OS/X, for that matter) I just use any of several great DAWs (eg Cubase) and add stuff like effects and instruments by VST or even ReWire. Works fine.

    I just don’t see what improvements Jack could offer me, but even if it were every bit as good as you say, the problem would still remain that good apps are sorely missing on Linux, and I don’t see that changing anytime soon. At least the developers of Ardour are honest enough to admit that they can’t compete with the big guys.

    Re ActiveX and D-Bus:
    WP says: “The GNOME release has officially deprecated Bonobo sometime since GNOME 2.4,[1] and developers have been advised to use or switch to an alternative such as D-Bus.[2]”
    Note the word “alternative”.

    And: “Most systems implement a privileged system channel, plus a private channel for each logged-in user. The private channels are required because the system channel has access restrictions. The purpose of the private channel is to provide unrestricted communication among any applications of the user.”

    “unrestricted communication among any applications” does sound like ActiveX.

    However, the main point I made is that ActiveX allows for exactly that collaboration between applications and parts that you claimed would not be possible with commercial SW. MS developed it, and KParts and Bonobo copied it because it is so useful. On Windows, there is quite a market for (earlier) VBX and (now) AvtiveX controls you can use in programs you write to add all kinds of functionality. Collaboration that works.

  56. oiaohm says:

    NT JERKFACE the for Sony Vegas replacement you are talking lightworks. Editshare is behind that.

    Editshare business is making specialist hardware for video storeage for TV channels.

    The model for Editshare is software to sell hardware. Hardware is worthless without the right software and software is worthless without hardware to run it on.

  57. NT JERKFACE says:

    @Pogson

    Mozilla uses the clusterfudge (censored for Pogson’s naughty filter) codebase open source model. The FF codebase is a clusterfudge to wade through which discourages forking. They also have an established name so trying to beat them with their own source becomes a risky venture.

    FF also is used for browsing the internet which makes it ideal for internet ad partnerships since they are unnoticeable to the user. So they have a working business model that sustains development. But for local programs that model does not work.

    If you released an open source Sony Vegas alternative with ads they would simply be ripped out. Your FooMovie5000 that you invested 20 million in would become GNUMovieAdFree within a week. That type of software also can’t use the Red Hat support model so what is left? You guys don’t have a viable business model for the most common type of software and yet continually drone on about how open source is going to take over the world. Sorry but there aren’t enough bored millionaires to fund it all.

  58. oiaohm says:

    ch I missed the big exact different that shows dbus is different.

    You call a dbus object. If you treat it as com you will think you can pass section of screen and have the dbus object draw on it. Sorry you cannot dbus forbids this. All data to be displayed to user has to be displayed by the user side program. The programs inside the dbus cannot send data straight to the users screen. And user side programs if user is not privileged as per normal cannot see inside the dbus run part.

    Yes d-bus is part a joke. Desktop bus yet in fact nothing it directly running can draw on the desktop.

    Is this com. Answer no. Could you make a vb based off d-bus nop. Because you would not be able todo gui objects using d-bus tech. This is why people who think d-bus is com start coding on Linux end up completely screwed. Yes because d-bus is design not todo particular things. If you need gui objects you have to use different system. So yes d-bus is not active-x either.

  59. oiaohm says:

    ch VST parts I get to use with Jackaudio under Linux http://www.anticore.org/jucetice most of them in fact work.

    Yes jackaudio is a nice piece of work. If there are items that don’t work under Linux. Like AU (on OS X) or some VST plugins it is possible to run Jackaudio on Windows or OS X so allowing me access to everything they have as well into a single platform.

    Basically if there is something somewhere on any platform that does audio work that does what I want I can integrate it into my work system.

    ch
    “Coupled with applications that offer all basic functions that works quite nice.”
    What is basically crap because if you asked them to use a LADSPA or DSSI they cannot . As the windows ones to use AU you are up the creek. They are all platform locked solutions and all limiting.

    Jackaudio on Linux give the the most when I have to choose single platform. Jackaudio gives me everything when I can choose multi platform. FOSS provides some great platform unifying applications that here is nothing equal to in the Windows or OS X closed source software worlds.

    ch
    “That’s what MS invented DDE/OLE/COM/ActiveX for. And it’s so good that it was copied in KDE and Gnome – today it’s part of the d-bus.”
    This shows how big of a idiot you are. dbus and DDE/OLE/COM/ActiceX have nothing that far in common.

    gobject is the gtk equal DDE/OLE/COM. QT also has it own but can also interface with gobject and you have kpart and QObject(these two will most likely merge). These are all technologically in the DDE, OLE and COM.

    d-bus is called desktop bus. There is a d-bus server its a Inter Process Communication class item yes but that is where the releationship between the two divides. d-bus chat method is based on a older Unix method. What you can access on d-bus is controlled by policykit so an application not authorized to access a particular d-bus part may not even see it as existing. Action application can send to the d-bus part can be exactly defined by policykit anything not that will be rejected.

    d-bus was design to solve a issue. How to give programs access to privileged sections of the system without putting the system at too much risk.

    A privilege code runs in fully independent user and memory space in dbus.

    Dbus closest relations of all things is in-fact sudo and unix IPC. There is really nothing in common with com, ole, dde or active X.

    There are other options as well. Universal Network Objects and xpcom from mozilla to do com like stuff.

    There are a lot of application developers who cause themselves tears on Linux because they are using d-bus for non security requiring items. So fighting with the secuirty of d-bus. D-bus was invented to solve a secuirty problem with hotplug and other items. User needs to mount a device they don’t have enough privilege but we don’t want to give them privilege either.

    Windows lacks a secuirty controlled com system. Applications on windows work around this by installing many services each add possibly another weakness to the OS. Linux is head of the game here with d-bus. Linux did have too much infighting with the basic non secure com system. Mostly that is dieing down to gobject and QObject as the winners. Most likely there will be a C++ based winner as well.

    Basically Windows needs a dbus. With a dbus a lot of applications could drop there horrible custom privilege control system services. Leading to secuirty nightmares. There are many third party software that can cause a privilege escape due to services installed on Windows.

  60. Ray says:

    And besides, MAME on javascript allows for better benchmarking 😉

  61. Ray says:

    You know, for NaCl, I think it’s simpler to make HTML5 and Javascript faster, since they can be used to make web apps through them, instead of having to convince Mozilla, Apple, Opera, and Microsoft to incorporate them in the browser, especially Mozilla and Microsoft.

  62. JairJy says:

    Developers need to eat, they need money to buy food and feed their families, also to maintain other necessary goods.

    Firefox will be what is it today if their main developers haven’t gain any salary for work on Mozilla? Of curse not, the same is with the Linux Kernel, yes, there are a lot of free programmers, but IMB, for example, pay to some of them to work full time on the kernel, and I bet you than those coders are the better on maintaining Linux.

    Netscape made Navigator to get money, they couldn’t get it so they made Mozilla/Firefox and use a practically new business model on that time, and use the Free Software movement to get free handwork. Canonical, IBM, Google and Oracle do the same.

  63. JairJy wrote, “Mozilla wouldn’t have been created an awesome product if they couldn’t get money.”

    The sourcecode from NetScape was made available to all. Anyone could have used it with or without any concern for making money. The chief motivation was to maintain and to improve good software, something that inspires a lot of programmers. It was decided to abandon most of the NetScape code eventually and FireFox emerged. The Netscape code may not have lived on but it served as a lesson for developers how not to make software. Interestingly, AOL sued M$ and won but M$ paid them partially in royalty-free use of IE… which took away any incentive AOL had to make money from Netscape. Obviously Mozilla had ways and means of making money from FireFox but money was not the initial objective.

    “”Mozilla” is the every-day name for the free and open-source software project that was founded in 1998 in order to create a next-generation Internet suite for Netscape. On 15 July 2003, the organization was formally registered as a non-profit organization,” see Wikipedia

  64. Clarence Moon wrote, without reference to anything, “The way I read the Red Hat contract, if I have a subscription on any computer using Red Hat, I have to have a subscription on every computer using Red Hat or else I have to uninstall it. “

    Nonsense. Here’s the relevant legalese:
    “Distributing the Software or any portion of the Subscription Services to a third party or using any of the Subscription Services for the benefit of a third party is a material breach of the Agreement even though the open source license applicable to individual software packages may give you the right to distribute those packages (and this Appendix is not intended to interfere with your rights under those individual licenses).”

    There’s nothing about uninstalling the software there. The subscription for support may not be used after termination but you can use the software as long as you want according to the licences for individual packages. Individual packages distributed under the GPL, for instance may be used with or without paying a subscription fee to RedHat. If you want their support, however, you have to play by their rules and report all instance for accounting. Access to RHN etc. are restricted to subscribers. If you can support the software on your own and you don’t use RedHat’s non-FLOSS stuff, go ahead. That’s hard work and most people taking subscriptions will play by the rules because it’s easier and it works. The Linux kernel and the GNU stuff for sure you can continue to use. I don’t know about every single package in RedHat, however.

  65. JairJy says:

    “Why does any hunter need an
    expensive rifle? You don’t need
    more than bow and arrow or a
    good spear to kill anything!”
    Thank you ch, you encapsulate the whole debate of why often comercial software is better than free software.

    Comercial software gets money by a bussiness model, its very simple. Mozilla wouldn’t have been created an awesome product if they couldn’t get money. Thanks to the money, companies try to satisfy their customers. That’s why Ubuntu and Linux Mint exist, and are the most used distros. Thats why it takes 15 years to some guys to create a free clone of Warcraft 2, but in 2 years a small team created Bastion and sold 500,000 copies on dicember. Thats why Windows has better apps, and why Windows is the most used OS. And that’s why design professionals spends hundreds of dollars on Photoshop and Macs.

  66. Clarence Moon says:

    “go through the Sound Generators particularly and notice how many I would be missing out on if I went windows so making the tool very limited”

    I went through some of them and it looks to me like the big ones use Windows and the ones that do not are the one-man, obscurities that haven’t had much, if any, development for years. In any case, it is a minuscule usage case in terms of applications. Next to nobody cares about Jack Audio. I just thought that it had a cute name to apply to your general lack of understanding.

    I get a chuckle out of your absurd responses even if no one else does.

  67. Clarence Moon says:

    “How many incidents per annum does that other OS have? A lot more than three, I suspect.”

    I have been using Windows for almost 20 years now and I have never called Microsoft for anything. I called Dell once because it had a bad video card as it turned out. Have you ever called Microsoft for support? What happened?

    The way I read the Red Hat contract, if I have a subscription on any computer using Red Hat, I have to have a subscription on every computer using Red Hat or else I have to uninstall it. Doesn’t that seem rather harsh?

  68. Printing [s a small niche and a lot of printing is done in b&w and cartoon colours.

    I do serious editing of images every time wife pulls her camera out. I am trying to get her to use a database instead of folders but she’s set in her ways and retraining will be difficult.

  69. ch says:

    @oiaohm:
    “Basically its a collection of applications that can work as one huge unit.”

    That’s what VST (on Windows) or AU (on OS X) is for. Coupled with applications that offer all basic functions that works quite nice.

    “Most of the audio production software in existence is at my finger tips on Linux. Were on windows only a small percentage is.”

    Most of the SW in existence? Maybe. Most of the good SW in existence? No way. Once again, you just don’t know what you are talking about.

    “Basically mind refrain from telling me I need to leave Linux ch.”

    Where exactly did I tell you what to do or not to do?

    “Closed source infighting does not create a lot of these great groups of applications. Basically its not what the single application can do it what the group of applications will do.”

    That’s what MS invented DDE/OLE/COM/ActiveX for. And it’s so good that it was copied in KDE and Gnome – today it’s part of the d-bus.

    “VS is just another IDE nothing special.”
    And once again you don’t know what you’re talking about.

  70. Clarence Moon says:

    http://www.computerworld.com/s/article/9222980/Google_to_pay_Mozilla_300M_yearly_in_new_search_deal_says_report

    I got to wondering who gets the moola that the Mozilla Foundation is now taking in courtesy of Google’s largess.

    http://www.sec.gov/cgi-bin/browse-edgar?company=Mozilla&match=&CIK=&filenum=&State=&Country=&SIC=&owner=exclude&Find=Find+Companies&action=getcompany

    They don’t seem to have any SEC filings to show officer compensation, so I still wonder. I did find this old blog:

    http://thetruthaboutmozilla.wordpress.com/2007/10/22/financials-be-damned/

    Are thee FLOSSers being taken for a ride?

  71. oiaohm says:

    ch http://libregraphicsworld.org/ project understands what you are talking about.

    They are running plans to bring CMYK and more to gimp and other open source graphics applications. Problem s getting everything in shape will take a bit of time.

    Gimp engine fix up to a point that it can start supporting CMYK properly and larger bit depths is almost done.

    http://libregraphicsworld.org/blog/entry/what-hasnt-happened-to-gimp-2.8 Yes there is a formal plan under way to address the problems you are complaining about ch. Also go past it and give gimp some true performance exploiting video card.

  72. oiaohm says:

    Clarence Moon In fact you don’t know jack and you just proved it.

    It the suite of applications that work with Jack to product a unified solution.

    Windows jack server is a paper weight without the applications to go with it. Clarence Moon sum of parts. No single part is worth anything alone that much other than a few exceptions.

    http://jackaudio.org/applications go through the Sound Generators particularly and notice how many I would be missing out on if I went windows so making the tool very limited.

    Also this http://bristol.sourceforge.net/ I have a feel for some oldhardware I cannot afford even if I could it most likely would not work. Basically a few 10 thousand dollars of hardware replaced by 1 bit of software.

  73. ch says:

    “Anyone can change a bit any way they like in their own images with many dozens of apps.”

    If that was all image manipulation is about today, people could just do that with Windows Paint. Please, let’s be reasonable here. There are professionals who do photos and designs to be printed in just the right colors, and then you need a lot more like good CMYK support.

    “What does “fully equivalent” mean, that a FLOSS app is identical to a non-free app?”

    Of course not, it’s about functionality and ease-of-use.

    “Why would I think of doing IT the PS way/the wrong way?”

    YOU of course wouldn’t because you don’t do serious image manipulation, but someone who actually needs the functionality surely would – or rather, they actually do it day in, day out.

    Frankly, your points amount to: “Why does any hunter need an expensive rifle? You don’t need more than bow and arrow or a good spear to kill anything!” Could we please continue this discussion on a more mature level ?

  74. oiaohm says:

    http://www.xaraxtreme.org/ oldman is where it hiding.

    But due to xara fairly much leaving it alone for 2 years a new developer team is getting assembled. http://code.google.com/p/xarino

    So this is effect stuff does not just die if it good it keeps on coming back in the open source world.

    This is the difference with FOSS when they do abandoned and the FOSS version if it can stand on it merits it will come back to life this time outside its first makers control. In fact it can be the worst mistake the closed source company can make because something base off there own code base can end up competing with them.

    Freeware just dies.

  75. Clarence Moon wrote, “why not say that it is an honest day’s work for Microsoft to do much the same, only at a much lower price per capita?”

    Many organizations budget $1000+ per pc per annum to prop up that wrotten pile of spaghetti that is that other OS. Even M$ cannot afford to support that other OS for free, charging $60 per incident in some cases. How many incidents per annum does that other OS have? A lot more than three, I suspect.

    When I set up the system for the school in Easterville, we had zero requests for support except from a tiny number of really-set-in-their-ways users. No hardware or software failures in the clients at all. The labour to keep it running was a few minutes per day just to check that everything was running smoothly for the whole system.

  76. oldman says:

    “Like Netscape giving away its browser which became FireFox which now gets $300million per annum for mentioning Google…”

    So what!

    One company was in the right place at the right time.

    Plenty of companies drank the FOSS Kool-aid and died or had to salvage a business by doing exactly what I described.

  77. oldman says:

    “FLOSS works, oldman. Get used to it.”

    Not everyone is going to do the technological equivalent of taking a welding torch to a 23 year old roto tiller to keep it going. FOSS is not enough.

    Get used to that!

  78. Clarence Moon says:

    I cannot control myself, Mr. Oiaohm! This is just too much!

    “There is nothing in audio tools on Windows that is even in the same game.”

    Well, if you look at:

    http://jackaudio.org/download

    you can see that the Jack audio suite is available for Windows, so it must be true that you literally don’t know Jack.

  79. oldman wrote, “Frankly I would not be surprised if this lack of publicity for any linux version turned out to be deliberate. There have been a number of cases of companies that started out with totally free products and then shifted the major portion of their effort to supporting the commercial version and all but abandoned the pure freeware version.”

    Like Netscape giving away its browser which became FireFox which now gets $300million per annum for mentioning Google… Oh, wait, that’s going the other way in a big way…

    Here’s the source code.

    Essentially the developers give away the product for $0 but are paid by advertising Google. That works for them.

    Then there’s pure FLOSS, like VLC which started as a student’s project and grew into a global mediaplayer.

    FLOSS works, oldman. Get used to it.

  80. Clarence Moon says:

    “The world can take care of its own poor huddled masses. They don’t need M$ or Apple to lead the way.”

    I guess you are a Ron Paul fan, Mr. Pogson! He had much the same sort of answer for people who had severe medical problems and didn’t have any insurance. “Let the churches and charities take care of them!”, he said in one of the debates. “It is not a job for the government.”

    I don’t think that the poor, huddled masses using the internet can necessarily take care of themselves. You still read about old ladies taken in with phishing schemes or who send money to the Nigerian colonel’s widow in anticipation of a huge commission check.

    If Red Hat can honestly sell “support” for Linux and charge $179 per year, why not say that it is an honest day’s work for Microsoft to do much the same, only at a much lower price per capita? Many people are more than willing to give up some money and the freedom to pore through source listings in order to just use their computers, I believe.

    People such as yourself who can hunt their own meat and fix their own OS do not need that service, I agree, but there are quite a few who do and providing that sort of help to them is a fair go.

  81. oiaohm says:

    ch full Acrobat there is pdfedit and libreoffice both can do the job you require.

    Visio there are quite a few ways around that.

    Cubase runs quite well in wine quite handy when I wish to use it with Jackaudio. Note jackaudio does not particular like windows.

    Dentist to create inlays would not happen to be CadBlu Dental by any chance. That is Windows Linux and OS X. Has some nice milling and 3d printing tech to produce the item on the spot quite quickly.

    Most of the good Dental software is tri-platform. Same with the good jewelery design.

    There might be no FLOSS equivalent but there is not always a requirement to spend on Windows to use them.

    Ch try finding something on windows to replace this http://jackaudio.org/applications Basically its a collection of applications that can work as one huge unit. There is nothing in audio tools on Windows that is even in the same game.

    Yes jackaudio on the right setup is fully real-time.

    Remember I can still use some windows part by wine with jackaudio. Most of the audio production software in existence is at my finger tips on Linux. Were on windows only a small percentage is.

    Basically mind refrain from telling me I need to leave Linux ch. There are applications that are great in some cases it set of applications that work as one to be great on Linux. Closed source infighting does not create a lot of these great groups of applications.

    Basically its not what the single application can do it what the group of applications will do.

    Midi feed from jack can be controlling blender so 3d models are moving in perfect sync with the music playing exploiting blenders game engine. Live events can synced real time.

    Sum of the whole. Each part alone is basically worthless.

    I don’t miss photoshop. Photoshop cannot handle huge images. vips can. VS is just another IDE nothing special. QT IDE is quite nice these days.

  82. ch wrote, “There still are no fully equivalent FLOSS apps for those commercial apps”.

    Nonsense. It’s a bit-mapped image we are talking about here. Anyone can change a bit any way they like in their own images with many dozens of apps. “Commercial” means what, that licences are sold instead of given away? That’s quite irrelevant to the utility of the software to the end user. What does “fully equivalent” mean, that a FLOSS app is identical to a non-free app? That’s copyright violation, most likely. Even commercial apps are not “fully equivalent” so why should a FLOSS app be that way?

    for f in *.JPG;do convert $f -resize 500x450 ${f%JPG}png;done is a lot easier, cheaper and faster than firing up PS. Why would I think of doing IT the PS way/the wrong way?

    Further, I can do just about anything without even firing up a gui:
    dpkg -L imagemagick|grep bin
    /usr/bin
    /usr/bin/compare
    /usr/bin/animate
    /usr/bin/convert
    /usr/bin/composite
    /usr/bin/conjure
    /usr/bin/import
    /usr/bin/identify
    /usr/bin/stream
    /usr/bin/display
    /usr/bin/montage
    /usr/bin/mogrify

  83. oldman says:

    “Xara xtreme on Linux costs nothing windows find money.”

    Xara Photo and graphics designer (its name now) Seems to be one of those interesting niche products that do exist and the price seems right.

    Unfortunately, I cant find any reference to a linux support at all let alone a free version on the xara.com website. Perhaps you can assist me in pointing it out Mr. oiaohm.

    Frankly I would not be surprised if this lack of publicity for any linux version turned out to be deliberate. There have been a number of cases of companies that started out with totally free products and then shifted the major portion of their effort to supporting the commercial version and all but abandoned the pure freeware version.

    In short, nice try, but all you have proven is that some companies are willing to support desktop linux applications. But it is interesting to note that none of the companies do so exclusively.

    They may contribute to the community, but they all keep their day jobs by selling the windows and OS X versions where the money is.

  84. Clarence Moon wrote, “What useful program of any note is available on Linux and not on Windows?”

    Good software is portable. That is not a feature of an OS. M$ has gone out of its way to encourage the world to write non-portable software. I recommend avoiding M$ for that reason and others.

  85. ch says:

    “I have only used it a few times in my life and Gimp and ImageMagick work for me.”

    Nice for you. That’s why I wrote that your demands are rather small. Other people need PS – or Visio or Cubase or VS or that stuff my dentist used to create my inlays or the full Acrobat or …

    There still are no fully equivalent FLOSS apps for those commercial apps, so what do I do if I need any of them ? On the other hand, what Linux app that cuts it isn’t available for Windows ?

    If Linux and the SW available for it are enough for your needs, that’s fine. But please refrain from telling me what I need.

  86. Clarence Moon wrote, “is it not fair to have a massive hand-holder sort of company, as I suggested is Microsoft’s core competency, as an entity to service the immediate needs of this class of user? “

    The world can take care of its own poor huddled masses. They don’t need M$ or Apple to lead the way. M$’s core competency is lock-in, not hand-holding. Everyone I know who uses that other OS sees re-re-reboots and the fix-it shop as the hand-holding services. I don’t know anyone who has ever phoned M$ for help but many have phoned me, even users of that other OS. M$’s attitude for years has been that once the sucker pays for a licence, they are dead wood until the next licence is available. M$ relies on slave labour by “partners” to help end users.

  87. oiaohm says:

    Clarence Moon
    “I cannot think of any application that runs on Linux and is used by the general public that does not also have a version for Windows.”
    Its not quite the black or white.

    xara xtreme. Same package is able to be got for windows but there is a difference. Clarence Moon cost. Xara xtreme on Linux costs nothing windows find money.

    isomaster is the same camp as Xara xtreme.

    There are a lot of applications to a Linux user that are free cheep and effective to a windows users are big dollars.

    Yes there is a really confusing divide at times you really do need to look at what the user requires sometimes it can be a few thousand dollars in license cost putting the user on Linux yet you have not changed the program they are using. Worse is that the application performs better.

    K3B what is a really nice burning software.

    I do audio work at times requiring synths. There are particular synths that are unique to Linux.

    Linux menu is more newbie compatible. Applications are grouped by common sense systems instead of application makers name and other random messes.

    So really the Linux is the one with the easy to use basic interface. But its when you go to change a battery in the scooter does the user find themselves in trouble.

    If Linux could get it advanced interfaces new person friendly it would be a winner all round.

    Issue why Linux advance interfaces have suxed is a historic one. X11 has suxed was not dependable so you designed no administrator interface that depended on it. X11 poor design issue should hopefully end this year as wayland becomes more dominate.

    Rule one of administrator interfaces they must work. X11 was breaching rule 1. It could die in way the terminal interface to the machine was killed. So administrator interfaces were designed to work remote.

    One issue cause of one huge weakness.

  88. Clarence Moon says:

    A philosophical thought arises on re-reading the posts here, Mr. Pogson.

    It would seem that you class those who follow the mainstream, institutionalized thoughts about IT as “twits” and suggest that your own enlightened view would suffice for anyone’s needs. You often offer what you think is proof of that and it is an open and shut case the way that it is presented.

    I really cannot argue against that position for anyone who is competent in enough disciplines to be able to do it onself, but it seems to me you ignore the plain fact that very few people could, without additional training and relevant experience, do the same. Most likely there are many who could never do that because they are not motivated to put out the effort or who do not have the time to spend on such learning.

    But is it not fair to have a massive hand-holder sort of company, as I suggested is Microsoft’s core competency, as an entity to service the immediate needs of this class of user? A paraplegic could get around on his own with a skateboard and leather gloves to protect his fingers if he spent enough time developing the upper body strength to do so, but he could opt for an electric scooter with a joystick as a more expensive but easier to obtain alternative. That is the same sort of economic tradeoff that seems to be the case with FLOSS versus proprietary software, I think.

  89. Clarence Moon says:

    I cannot think of any application that runs on Linux and is used by the general public that does not also have a version for Windows. That extends to things used in software development itself.

    Strictly for my own edification, I am embarking on a hobby effort to be able to write useful programs for my Kindle Fire.

    I have set up an Eclipse environment for it so far and have successfully run a number of samples successfully. I have even opened a developer account at the Amazon cloud and store. All that with Windows 7 on my Dell laptop.

    What useful program of any note is available on Linux and not on Windows?

  90. ch wrote, irrelevantly, “much less runs on Linux”.

    Are you sure of that? Source? There are hundreds of thousands of apps that run on Linux these days and literally millions of developers create for Linux. If I am driving a lightly loaded truck with one tire of an 18-wheeler on dual axle flat, I can easily keep going to a place where I can fix it. Lack of a useless app similarly does not stop IT. An app is useless to me if I have another app ready and able to do the job. Why should I care about the useless apps?

    Consider the much-vaunted PhotoShop. I have only used it a few times in my life and Gimp and ImageMagick work for me. To the extent that I don’t need PS, it is a useless app and I would be foolish to alter my life and workspace just to acquire PS. OTOH, if I want the best price/performance from IT, I will use FLOSS because of its low cost and reliability and performance.

  91. oldman wrote, “If you worked in an institution where you were required to conform to institutional standards, you WOULD be using the software that you were requested to use to complete the task.”

    All institutions have standards but they are not all the same. Where I last worked, XP was the standard and I persuaded the institution to change to GNU/Linux. Everyone there saw the benefits when the software was run side by side on the same hardware. I have moved on from places that would not accept Free Software, mostly because IT was extremely unreliable and FLOSS was the only thing that allowed me to work. In one place no one could print except me and my students (even with M$-certified IT guys), yet the institution felt FLOSS was a threat of some kind. Go figure… In another, FLOSS was able to upgrade a lab of old PCs (some 8-10 years old machines) to 24 working machines when XP could only be kept running on 14 machines reliably for $0 capital cost. I took wrecks from storage to complete the lab. Still the institution felt FLOSS was “not ready”. Twits can be found everywhere, even running institutions.

  92. oiaohm says:

    ch there are reverse cases as well. There are applications that don’t work on windows that are truly great at what they do. 99.9% claim is a bit out there.

    Sony Vegas how many seats in a business need it NT JERKFACE. Areas of weakness like this is known reason for the porting of lightworks.

    Cenon and many other programs don’t exist for windows at all. Exist for Linux Unix and OS X.

    Basically is a fairly clean split most free and great applications are on Linux and OS X. Most commercial pay for applications that are great are on Windows.

    Running Linux where it suits can in fact improve the applications your users have access to.

    Like lets say you are developing android applications. Android compilers run faster on Linux. Yep locking your developers on Windows would have reduced there performance to producing final product.

  93. oldman says:

    “I usually ignored the request and used FLOSS or software I created. It makes more sense. It costs less and works better.”

    In the real world however you don’t just get to ignore a request. If you worked in an institution where you were required to conform to institutional standards, you WOULD be using the software that you were requested to use to complete the task.

    This does not even address that reality that you average computer user has neither the skill nor the inclination to write software themselves. And even when they do, no person that I know is going to set aside a tool that gets a job done as needed, to use a FOSS tool to do the same job, let alone incurr more effort rolling your own.

  94. ch says:

    “One created an .exe to emulate a browser session on a web page…”

    Wooha, now that’s really a weird use-case. WTF?

    “So many apps for that other OS, so many useless ones, too.”

    Well, the old adage that 90% of everything is §$%& applies to both FLOSS and commercial SW, I suppose. But the important issue here is: ~99.9% of all useful SW runs on Windows, much less runs on Linux. Fortunately for you, your needs are so modest that you wil never notice, but it sure does matter for others.

  95. NT JERKFACE says:

    I once was asked for an alternative to Sony Vegas and I whipped one up with Javascript and PHP.

    I also used open source to travel back in time and shoot Franz Ferdinand.

    Proprietary applications are never needed. You can trust me, I’m from the internet.

  96. JairJy wrote, “Windows has more apps.”

    Does anyone really care if that is the case if the apps are unnecessary. Only a few times in my life has anyone ever asked me to use a certain app to get the job done. I usually ignored the request and used FLOSS or software I created. It makes more sense. It costs less and works better.

    For example, one time I was taking a course and was asked to run two proprietary apps. One created an .exe to emulate a browser session on a web page… I asked the instructor if I could set up a server and give him the URI. He said yes and it was done without using that other OS at all. In another similar case, I was asked to use some crazy application that produced fancy graphics. It would have taken more time to learn to use the app than the software tools in GNU/Linux so again, I was ahead of my classmates by not following the dictates of the course curriculum. I got high marks. So many apps for that other OS, so many useless ones, too.

  97. lpbbear says:

    “No, not special at all. Just possessed of a sense of humour.”

    You’re claiming you are “possessed of a sense of humour”?

    Possessed maybe, humor, not even a little.

    (although you do have the extra special talent of typing to excess and saying nothing….worth reading)

  98. Dr Loser says:

    @Koz:

    No, not special at all. Just possessed of a sense of humour.

    Wasted on you, obviously. Happy New Year!

  99. Dr Loser says:

    Just as an addendum to JairJy’s point:

    MAME.

    I’ve taken the liberty of picking the 0.144 version for “all platforms.”

    Now, I’m too lazy to contact the relevant Google engineers, Robert, but perhaps you can do this for me. Could you ask them why they are wasting four days duplicating somebody else’s effort?

    And, while you’re at it, could you kindly request them to make the source code of Android FLOSS?

  100. Kozmcrae says:

    “I’m even prepared to give them four months for that one.”

    Wow, aren’t you special.

  101. Dr Loser says:

    @oiaohm:

    “You are forgetting the existence of Google Chrome Frame and the fact you can install Nacl as a firefox plugin.”

    You are entirely correct, and I am diminished by that omission.

    I will now go and roll in the snow and, afterwards, I will whip myself mercilessly with birch twigs.

    Not because of what you said. Merely because it gives me a perverse thrill.

  102. oiaohm says:

    Dr Loser you are forgetting the existence of Google Chrome Frame and the fact you can install Nacl as a firefox plugin.

    So Nacl conversion is fairly neutral. The competitor to Nacl is native code to javascript compiler that is still in development.

    Nacl version is as fast as windows version JairJy. Of course Mame port was bring a interface layer SDL to to Nacl. There is QT for Nacl and I think there is GTK for Nacl. Basically in time number of applications that will be able to run Nacl will increase.

    Dr Loser there is another possibility why you cannot see the 100~ million. Of course it requires doing homework.

    Ubuntu and other Linux distributions can be installed in chroot setups on android devices. That is not removing the android OS. That is android and a Linux distribution running side by side.

    Population of devices that can run Desktop Linux has increased. Market share should increase accordingly.

    So basically you might not be able to see the 1 million because a large percentage is walking past you hidden as a Android mobile device.

    This is my problem there are too many places for Linux to hide and very poor counting.

  103. Dr Loser says:

    Clearly ~100 million?

    You need to talk to oiaohm, Robert. He is a registered fantasist, and we all enjoy his brick walls of text, without necessarily believing a word he says.

    Look out! There’s a Gruffalo behind you! Maybe even 100 million of them!

  104. Dr Loser says:

    But, back to reality.

    Do you know what I would do if I lived in the Third World and had a dollar to spend on a kilowatt-hour, Robert?

    I’d sodding cook a moose with it. That’s what I’d do.

    Do you know what these guys earn? $10 a day is wealth. I wouldn’t be pratting around with some nifty little small thingie. I’d be slow-roasting a chicken or two to feed my family.

  105. Dr Loser says:

    And I beg to differ with your hypothesised statistics.

    Let’s assume a model (in this case, yours). A model is worthless without at least a back-of-the envelope cross-check (to stop your planetary explorer crashing into Mars, for example).

    OK, your model hypothesises “~100 million” PCs being re-purposed as GNU/Linux (desktops, I assume). Here’s the back of my envelope: I live in a developed Western country with roughly 1% of the world’s population, and it’s not unreasonable to assume that this evens out across the world for your alleged conversion.

    Which means that I should be seeing about 1 million GNU/Linux desktops around here right now.

    Well, I saw three being used as emulators for embedded software at work a couple of years ago. I’ll just go and hunt for the other 999,997, shall I?

  106. Dr Loser says:

    I’m sorry, but if “four days” is “seriously challenging,” then these people need to pick a proper challenge. Like making Android completely FOSS, for example? I’m even prepared to give them four months for that one.

    Which bit is the relevant challenge? Porting the assembly code? Building a rather trivial cross-assembler? Writing the converter from screen to HTML without tripping over yourself and targeting the far more standards-compliant IE9 instead of Chrome?

    Or just reading the manuals and learning how MAME works?

    Nice job those game-jockeys have down in Mountain View, isn’t it?

  107. JairJy says:

    From a game emulator to Windows is Doomed… okeeey.

    But Google is too late, Windows’ users have been using MAME a decade before, with much better performance and flexibility. No to mention all the other emulators from other consoles; Windows has more and the best ones, and Windows’ users have more freedom because, again, Windows has more apps.

Leave a Reply