M$’s Desktop Destiny

Gartner is taking credit for advising M$ to shift away from that other OS, NT version, to the new stuff in order to stay relevant. Gartner estimates in ten years almost no one will be running Lose 32 applications. Instead it will all be about Lose RT…

I think Gartner is right about that other OS sinking into oblivion in a decade but I don’t see M$ being able to dictate the shape of IT beyond the horizon. People are tired of Wintel and M$ whether on x86/amd64 or ARM. No one is in love with M$ sufficiently to follow wherever M$ leads. If the apps go, so will the sheeple. If people have to make an effort to migrate to RT, they might as well migrate to GNU/Linux, developers, IT people, users, everyone. The new applications will be on the web or LAN anyway so there’s no need for particular clients from M$ or anyone else. People will choose the cheapest clients that work and those are */Linux for the time being. If M$ becomes irrelevant, why send them money?

Windows desktop destined for long slide to oblivion, says Gartner.

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.

50 Responses to M$’s Desktop Destiny

  1. oldman says:

    “Oldman please don’t give me credit for something that is not mine. Those theory’s are based on Sony white papers. Not in English”

    I assure you that I gave you no credit sir. I assumed that you had read it somewhere and were presenting (as usual) someone else s ideas as if you were the voice of authority.

    You are pathetic mate!

  2. oldman says:

    Oldman thing was there are two different common solutions to the same problem. With the IBM San volume controller you cannot run it virtual options if you are in a multi supply environment. Because what you have to use on parts from other makers will be incompatible. “This is the problem Oldman there were two solutions both valid and you were only considering one. It had to happen to you at some point. You had got your self a little to isolated to all the options. About time you get over it oldman.”

    No what we have here is the following:

    1) I made an observation that the abiquo cloud solution that you had practically shoved down my throat in one of our previous exchanges had been evaluated by one of out teams and washed out as a cloud solution because it did not support Fiberchannel SANS.

    2) Instead of interpreting my “not supported” as meaning that abiquo could not automatically provision storage from arrays on a fiberchannel san, You chose to interpret it in a way that you felt would allow you to go after me me once again as “incompetent”

    3) When I gave reminded you of the true meaning of “supported” in a cloud product. you then accused me of being behind the times because I did not know about Netapp wonderfulness and assured me that you were familiar with with the SVC and “Knew” that it was not up to the job.

    4) I then asked you for details about the environment that you were familiar with…

    And we received the vigorous hand-waving that you gave above, which is as close as we are going to get to an admission that you do not have a clue about the SVC and have never worked with one, and that you were blowing it out your a$$ in your zeal to attempt to prove “once again” that I am “incompetent”.

    Look genius, I am very well aware that the SVC is as of now a block storage ( Fiber channel, 10Gb iSCSI or 10Gb FCoE) solution. But I did give you a clue about how I solved this in production. Remember the IBM N7900G that I mentioned? Did you look at it? Well genius, This unit is a rebranded NetApp appliance that provides NAS services to upstream hosts but uses as its storage an external (i.e. non-netapp) array. As it turns out one of the storage array providers that it supported was (you guessed it!) the IBM SVC!. So (Surprise surprise) I had an all in one solution already in house in production.

    So you have been caught out as the bullship artist that you are.

    ROFLMAO!!!!!

  3. Clarence Moon says:

    the supports are down to one or two pillars like retail shelf-space and various other lock-ins

    Only retail space and “lock-ins” like customer expectations and desires to protect prior investments? The only thing wrong here is using “only” to describe what amounts to “every” reason to keep Windows and x86.

    BRIC countries, Malaysia, Kenya, much of South America and Europe are accepting GNU/Linux widely.

    The grass is always greener just over the hill, eh? Windows computers are the norm in those countries as well as over here in the main world, Mr. Pogson. Apparently they are the norm even in the wilds of Canada since you periodically complain of not finding Linux in your local emporiums.

  4. Clarence Moon, denying the possibility of change wrote, “You cannot turn back the clock on PC software OS, Mr. Pogson, the ship has sailed and you can only look to new opportunities elsewhere. “

    The world does not owe Wintel a living. They will have to work for it if they want to be part of the future. Wintel is a house of cards and the supports are down to one or two pillars like retail shelf-space and various other lock-ins. BRIC countries, Malaysia, Kenya, much of South America and Europe are accepting GNU/Linux widely. Even USA has large and increasing usage of FLOSS and */Linux. That ship has sailed and refuses to go back to port.

  5. Clarence Moon says:

    The price of anything in a free market is the lowest of…

    You make a lot of mistakes here, I think. For starters the “cost of production” is a negligible element of the accounting for software products and is just the cost of the box and disk at the worst and essentially zero for wares sold as downloads or as licenses to OEMs who have a master to copy. If you want to say that production costs include management and admin, research and development, and sales and marketing as well, you might be closer. Then do not forget “profit”.

    As you note, it takes a few billion to just get a product initially created for a major product market such as an OS. You cannot sell it to anyone for any amount of money until that product has been produced and that cost paid for from investment funds. Do you have $10B to risk on a new idea for an old market? You would be the first person crazy enough to risk it.

    But what if you did? Would you plan on recouping your investment via some razor-thin margin on unit sales, fighting others in the market to have the lowest price for a consumer to consider? I think that would be a bad idea and I bet you do, too.

    Of course Microsoft did not take billions and invest them and then go to market with some product. Rather they made a single sale to IBM that gave them a certain return for the initial effort and funded the purchase of the rights for the original DOS. They grew their business by re-investing their profits over a long period of time. That was possible because the market itself was tiny compared to today and did not need to be addressed on any widespread front. The market requirements were simply the product specs for DOS and they grew with DOS development.

    You cannot turn back the clock on PC software OS, Mr. Pogson, the ship has sailed and you can only look to new opportunities elsewhere. Google, Facebook, Twitter, or even Angry Birds are examples of new thinking. Linux is an old thought and even pre-dates Windows if you consider it was meant to be Unix on a PC.

    Linux today has most, if not all, of the parts necessary technically to match Windows as a mainline OS for PCs. But that is all that it has. It does not have the public recognition and immediate acceptance that comes along with the 30 years of evolution of the Microsoft OS products. The cost of obtaining that recognition is probably a lot greater than the cost of duplicating the functionality in code. There is no way to recoup and so there is never going to be an investment to make it happen.

  6. Clarence Moon wrote, “how is it that you think your ruminations about the market have the least bit of validity? “

    The price of anything in a free market is the lowest of

    • the cost of production, assuming the producer will sell cheaply to advertise or is generous, etc.,
    • the cost of local production, the potential buyer might decide to make his own,
    • the cost of similar acceptable products, and
    • some fraction of the value of goods/services that may be produced using the software (seeing the acquisition as an investment).

    Suppose M$ or anyone makes an OS that’s huge and complex and useful. We know from estimates that it costs $billions to do. That sets an upper limit to the price. Any one buyer would be crazy to pay more for it than it would cost to produce. They might pay more to save time, as an exception, but we also know it only takes a couple of years to produce something like GNU/Linux if you have many thousands of developers available. Now, if 400 million people get together to buy the software or to produce it, that huge cost gets shared and $billions becomes $hundreds. We have more than a billion users of PCs so the price of an OS per user should be way less than $100. If we produce a well-designed, modular OS like GNU/Linux instead of a mess like that other OS, the price may well be ~$10. If we consider the production as a continuous process instead of a one-time deal, $10 is a reasonable price. $100 and more is ten times too high. That’s the market distortion of which I write.

  7. oiaohm says:

    Oldman thing was there are two different common solutions to the same problem. With the IBM San volume controller you cannot run it virtual options if you are in a multi supply environment. Because what you have to use on parts from other makers will be incompatible. This is the problem Oldman there were two solutions both valid and you were only considering one. It had to happen to you at some point. You had got your self a little to isolated to all the options. About time you get over it oldman.

    Oldman there are many studies over this that were done to reduce factory floor accidents. Then those studies were found basically to apply everyone in every field. Problem is most course teaching how to manage programming teams complete do not take those studies into account. Remember those studies started at designing the item all the way through to final production. So covers everything. Also it needs someone who can read Japanese to get the information out of a lot of them.

    Oldman please don’t give me credit for something that is not mine. Those theory’s are based on Sony white papers. Not in English. Yes there is a issue with USA development firms being isolated from the research in other countries on how to get production out of there IT staff to be correct how to get production effective. I would say this is another case of you being isolated from the rest of the world. That you are saying its my theory not straight up seeing hey that Sony research paper based.

    Yonah there is a lot of research papers on how to effect peoples performance positively and negatively when it comes to sex for some reason it is a common topic.

    Like the bridge experiment. You will look more attractive to the opposite sex if you go to a high floor restaurant with a window view than what you will look on a ground floor restaurant with a window view. Yes a office in a high rise with a window view is more of a reward/distraction than most people think. Yes its a funny one people with window offices are more likely to have inter-office relationship and more likely to have sex in the office. It is not that they are physically better looking or the most paid at times. It is an effect on the human mind at play that makes them more attractive due to where they are. Location Location Location.

    Also sex drive is different from person to person. Some questions like foreplay time is unique to the person but what works in foreplay is has a set of rules. There are some communality that can be exploited to your benefit if you know it. Mind you it also prevents you from making a complete ass out yourself on the first date.

    Yonah asked for an example. A person basic nature is the basic nature. Does not matter what job title you give the person the thing does not go away. Fighting a persons basic nature in any form of production increase errors in production. Some of the basic nature things don’t change between people. Some are important to be aware of particularly as a manager when it will effect there performance.

  8. Yonah says:

    Oiaham, “Clarence Moon. You don’t understand what I just said.”

    Does anyone? I think one would have to be unemployed to sit down and try to digest each one of your really long comments that drone on and on. I think that’s a sign you don’t have a lot of people in your real life to chat with. Cheer up man!

    “Human bio-chemistry evil point you have been annoyed the next few days those chemicals are still in your brain effecting your operations.”

    Seems like you would have much to atone for around the office, he he. Dr. Oiaham, please fill my beggar’s cup with your vast knowledge of the universe. How much time should be spent on foreplay to ensure a satisfied woman? I’m sure, like all of us here, you know some things about human reproductive behavior that we don’t. Please share with us.

  9. Oldman says:

    And now we have dr hams theories on. Human behavior as it relates to code development.

    Of course such bloviation is far easier than coming up with the proof that he actually isn’t blowing it out his a$$ about his experience with the IBM San volume controller in a production environment.

  10. Clarence Moon says:

    Software is a commodity and should and does sell/license at low rates in competitive markets…M$ has distorted the market.

    Microsoft has defined the market, Mr. Pogson. You consider it distorted, but it is what it is and that is the reality.

    Are you a PC hardware or software vendor, Mr. Pogson? Of course you are not. Are you a software supplier? Of course not. So how is it that you think your ruminations about the market have the least bit of validity? Hating Microsoft will not make them go away or even cause any of their existing customers to abandon them as a supplier.

    You can root for emerging nations to select Linux due to a perceived lower purchase cost, but even that does not seem to be born out by usage statistics from those regions. When faced with contrary statistics you argue that they must be invalid and biased against your position.

  11. oiaohm says:

    Clarence Moon. You don’t understand what I just said.

    Because what I said is psychology. Lot of PC production businesses don’t take psychology into account so leading to fighting with there development teams todo what they need todo.

    This is where they end up in big problems. And the psychology I am talking about there is general humans and appears in every profession.

    Yes its a laugh sitting in courses teaching how to manage software development. Most of them are written by people who have no clue of psychology. So they take the point of view that humans are creatures that can be made do anything(yes not reality dream world thinking this is).

    Reality there are distinct groups of humans. You know the ones that disassemble everything to see how it works and those who love building new item.

    This nature is very heavily in grained. Cannot be got out of a person they are either one way or the other. The disassemble everything to see how it works people love QA and reworking code for higher performance its playing into there nature. Those who like building new things love creating new features. Again playing into the person nature.

    Either todo the other requires a big stick.

    Most mature trades give them clear different titles. Like a person who does demolitions you don’t get building stuff or the reverse in most cases.

    Information Technology is still young and its titles are not clearly split to match what humans are. Instead businesses try to drive round pegs into square holes with IT staff and wonder why they are battling with there IT personal. That one todo quality control programmers have to be threaten as you mentioned Clarence moon is a classic case of not understanding the human psychology of what is going on and why they are not doing it happily. So you are making them unhappy this does have bad effects on productively the next few days after. Human bio-chemistry evil point you have been annoyed the next few days those chemicals are still in your brain effecting your operations. Yes an increased error rate when forced todo something this is human. A person has a lower error rate when they do it willingly even lower error rate when they love doing it. This is not IT particularly.

    Understanding human psychology you can have a very happy and productive programming teams. No need to be forcing people todo items they don’t like doing in fact you have particular people lining up willingly to audit the code and to write new features. Yes the ones who like auditing will be asking managers what needs auditing next they will even do this in there spare time. They see it as fun.

    In fact having the code audited by different person than who wrote it detects more errors including missing documentation. Does not suffering from I thought I wrote documentation problems.

    Yes the funny part splitting the two jobs creates better resulting code and happier staff.

  12. oiaohm says:

    Clarence Moon
    “You have a very weak grasp on the PC software business and should not be making such a fool of yourself.”
    You don’t understand what I just said.

    Because what I said is psychology. Lot of PC production businesses don’t take psychology into account so leading to fighting with there development teams todo what they need todo.

    This is where they end up in big problems. And the psychology I am talking about there is general humans and appears in every profession.

    Yes its a laugh sitting in courses teaching how to manage software development. Most of them are written by people who have no clue of psychology. So they take the point of view that humans are creatures that can be made do anything(yes not reality dream world thinking this is).

    Reality there are distinct groups of humans. You know the ones that disassemble everything to see how it works and those who love building new item.

    This nature is very heavily in grained. Cannot be got out of a person they are either one way or the other. The disassemble everything to see how it works people love QA and reworking code for higher performance its playing into there nature. Those who like building new things love creating new features. Again playing into the person nature.

    Either todo the other requires a big stick.

    Most mature trades give them clear different titles. Like a person who does demolitions you don’t get building stuff or the reverse in most cases.

    Information Technology is still young and its titles are not clearly split to match what humans are. Instead businesses try to drive round pegs into square holes with IT staff and wonder why they are battling with there IT personal. That one todo quality control programmers have to be threaten as you mentioned Clarence moon is a classic case of not understanding the human psychology of what is going on and why they are not doing it happily. So you are making them unhappy this does have bad effects on productively the next few days after. Human bio-chemistry evil point you have been annoyed the next few days those chemicals are still in your brain effecting your operations. Yes an increased error rate when forced todo something this is human. A person has a lower error rate when they do it willingly even lower error rate when they love doing it. This is not IT particularly.

    Understanding human psychology you can have a very happy and productive programming teams. No need to be forcing people todo items they don’t like doing in fact you have particular people lining up willingly to audit the code and to write new features. Yes the ones who like auditing will be asking managers what needs auditing next they will even do this in there spare time. They see it as fun.

    In fact having the code audited by different person than who wrote it detects more errors including missing documentation. Does not suffering from I thought I wrote documentation problems.

    Yes the funny part splitting the two jobs creates better resulting code and happier staff.

  13. Clarence Moon wrote, “You have a very weak grasp on the PC software business and should not be making such a fool of yourself.”

    People who live in glass houses should not throw stones. Software is a commodity and should and does sell/license at low rates in competitive markets. Equating Wintel and partners with a free market makes Clarence Moon seem like a fool.

    M$ went out of its way to lock developers and users into using Wintel-only stuff so M$ could rake in huge profits per PC. That’s leveraging other people’s work for M$’s profit. M$ does not make PCs, most applications nor run the Internet yet they take the biggest slice of the pie per PC. In a free market the lowest priced/best OS would be plentiful. M$ has distorted the market.

  14. Oldman says:

    You have to excuse our Aussie expert
    Mr. Moon. Being a legend in your own mind is
    A tough burden.

    Then of course there is the strain of having tO come up with that description of the IBM San volume controller environment that he worked with.

  15. Clarence Moon says:

    I said it elsewhere, Mr. Oiaohm, but I’ll repeat it here. You have a very weak grasp on the PC software business and should not be making such a fool of yourself.

  16. oiaohm says:

    Clarence Moon that is another example script writers and editors.

    Both can have equal understanding of English plot and so on. Script writers like creating new scripts and editors read and correct scripts.

    Some cases the editors can turn a weak script into a fairly decent one by cutting out crud and reordering. Exactly like QA forms of programmers do they can take a unstable program and make it decent.

    Both can have great talent. In every field there is the people who like creating new and those who like fixing the existing.

    Clarence Moon the problem is simple when you think about it. Its like asking a book author to edit script or a editor of books to write a script. They might be the best editor or author in there field but ask them todo the other job they will resist and can be a complete disaster.

    Companies have the problem of not seeing the the two different groups in there coders. Like you would not create a newspaper with just authors with no editor and you would not create a newspaper with just editors and no authors. Yet this is exactly what companies do when they write software. They have either all editors and no authors(Stable but have to battle all the time to get the editors to add features) or all authors and no editors (Unstable full of features can be a lot of features you never need and have to battle all the time to have quality control). Either mix is not production. Moderation is the key here. You need the correct mix of QA and feature adders. Yes this is why it can be for small companies to pay places like IBM or Redhat or Oracle todo there coding work for them.

    So as soon as someones says they need to beat there coders with a big stick to do QA its a straight up sign of missing a class of staff. Same with having to beat coders with a big stick to add features. Both are a sign that your coding staff pool is not balanced correctly. Having to beat them with a big stick to do QA means you need to add people who love QA. Now having trouble adding features you need to add the type that like adding features.

    Moderation not too much of one class of coder is very hard to achieve. No open source project is a perfect mix.

    Microsoft is horid mix because they employ a lot of people based on if they will add more features and don’t employ enough of the QA class.

    The art of moderation is not simple. Moderation is key to a lot of things. Clarence Moon I put yours in this case down to lack of experience and have not seen enough coding teams fail yet.

    Sooner or latter you would see a all editors event then wake up there are two ways coding teams completely fail to be productive. While in moderation they are productive and need less management force to get the required work done. This leaves management more time to plan what features should be worked on first and process feedback from users.

    Both failures equals fighting that eats up the all important development time.

  17. Clarence Moon says:

    I guess those are not big projects and they don’t have a clue.

    Script writers they are, Mr. Pogson, mere script writers. Certainly there are those among them who have great talent, but that talent is not in computer science but rather in whatever business logic that the script is about. Don’t confuse the two.

  18. oiaohm says:

    Clarence Moon developers and software engineers types of programmers.

    You have the Quality Assurance programmers and engineers as well. They are the bug fix class and find quality issues ones.

    This is the problem Quality Assurance Programmers and Engineers are different beasts to who you want designing new features for the applications. Quality Assurance people are not the risk takers.

    Yes the two different human natures. You need both to make quality and productive software. Ones who will take risks who will get bored out of there brain doing quality control. And the ones who love quality control and are after perfection but ask them to take a risk and they think you are insane. Yes QA Programmers are also the ones who do code performance optimisations because they are happier to read the code and understand it even if they did not write it.

    Some of companies have the big bad problem of employing developers and software engineers who show creative ideas resulting in a failure to employ those who love Quality Assurance so have having to beat there developers with a big stick todo it.

    Yes section of employment need to be checking if they like debugging or not. Yes those who love reverse engineering as well make good QA in a lot of cases. Its a particular minds. One like making new things. One likes dissecting things to see how they work. In any field you normally don’t find that in the same person.

    Surgeon vs Corner in medical for example. Surgeon like fixing and making. Corner likes working out why someone failed to stay alive. Engineer vs Forensic specialist again both can almost have the same skill set. Its the way they look at the problem.

    Failure to understand humans means companies don’t higher the right mix of stuff. You think about it asking a normal Surgeon to go be a Corner they not going to go willingly. Same with the reverse even if the are equally skilled.

  19. Clarence Moon wrote, “Another note here is that the people who work on the big projects do not use the term “programmer” to describe themselves.”

    Hmmm. If you say so, it must be true. These guys must be out to lunch…
    IBM which has been in the business since the stored-programme computer was developed don’t know anything about IT (sarcasm):
    “The application programmer is responsible for developing and maintaining application programs. That is, the programmer builds, tests, and delivers the application programs that run on the mainframe for the end users. Based on the application designer’s specifications, the programmer constructs an application program using a variety of tools. The build process includes many iterations of code changes and compiles, application builds, and unit testing.”

    see Application programming on z/OS

    I guess those are not big projects and they don’t have a clue.

  20. oiaohm says:

    Chris Weig
    http://scan.coverity.com/
    USA government contracted to check open source and closed source products to everyone else but open to the USA government. They publish yearly reports on quality issue.

    So some customers the source code is very important. Like armies.

    There are other parties outside distributions who pay check is to check the quality of code developers provide.

    Then you have developers working on https://sparse.wiki.kernel.org for the Linux kernel. This is coded to find and detect common coding errors.

    This has evolved to http://coccinelle.lip6.fr/

    With KDE there is another evolution. https://github.com/KDAB/Qt4to5 we are starting to get to the point that changing libraries will not need coders to alter each program. One programmer writes tool that does the transformation.

    Machine is freeing up coders todo other tasks and leading to less coding errors.

    Mind you overview of the Linux kernel is http://www.makelinux.net/kernel_map/ There are automated tools to make a map from software as well.

    Then there is work to make it even more automated http://klee.llvm.org/TestingCoreutils.html

    Chris Weig goal is to be able to review the code every single alteration. So a human review is basically not possible. Even not require a high speciality todo it.

    So with source code it is getting simpler to perform your own review if you wish.

    http://phpmd.org/ and http://rips-scanner.sourceforge.net/ for PHP developers.

    This is the thing code reviewing is not a days and days thing.

    Fabien Sanglard stuff I would not call reviews. I would call code documentation and explanation. Some open source projects are better for this than others.

    Does what Fabien Sanglard do help you when you apply a patch to know that the resulting program will work right. Answer it does nothing to make sure resulting program will work right. Just tells you what it looked like at one point in time.

    Performing defect scanning does not require that much coding skill to locate the defect today. Can require study to work out how to fix them.

    Chris Weig really your argument shows a base of lack of understand of the users of FOSS.

    Even the gcc complier over time gets more picky. Some of the reason why you cannot build old program source is that defects the complier did not detect it now does and refuses to build the code until it fixed.

    Yes death of old code in FOSS is caused by the machine auditing evolving. So its not just human eyes auditing. This is very cyborg the many eyes is part human part machine and its not perfect but the cyborg machine parts are getting better so as long as the code is being rebuilt with newer compliers more defects will be removed.

    This is where closed source where it stops being rebuilt after its released comes a problem. The machine checking process has also stopped.

    proprietary before normally are not documented.

    Items like samba has huge amounts of documentation http://www.samba.org/samba/docs/ that are in the same class as that review you point to. Yes you can order them in dead tree.

    There is lots and lots of documentation out there on the older more mature open source projects explaining there internals and how they should work in theory.

    Yes the BSP Marathon events from debian show the existence of the other kind of developer. One who sees bug hunting as fun and as a social event.

    Yes Clarence Moon there is a problem with some companies programmer recruitment they fail to higher some bug hunters who love that task so have trouble doing code auditing and reviews.

  21. Clarence Moon says:

    marginal interest to programmers.

    Another note here is that the people who work on the big projects do not use the term “programmer” to describe themselves. They are either “developers” or, more commonly, “software engineers”. A “programmer” is a low-level non-professional, $25/hr type with limited horizons and next to no say in what goes on. Frequently an amateur wannabe sort such as those who create the projects at Source Forge and such.

  22. Mats Hagglund says:

    “Where are the many eyes?”

    To find the answer to that question we have to know the numbers of “eyes” of Microsoft and Apple and then compare it to those of Linux communities. However Linux communities seems to fix bugs much faster than Microsoft. So i guess there are either more eyes than in Redmond or those eyes might see bugs sharply. And actually – does Microsoft care too much of those Windows bugs?

  23. Chris Weig wrote, “Where are the many eyes?”

    Debian GNU/Linux is an organization of ~1K developers/package maintainers. They regularly pour over bug reports and fix things. Some fix Debian installer or configurations/scripts and some pass upstream to the original source or use the latest source release to fix.

    see a list

    Not only hard-core developers contribute to bug squashing. Ordinary users who want things fixed get involved testing stuff. For example, an ordinary user may have a particular piece of hardware or a particular set of packages installed that the developers don’t. Even ordinary users may examine the init-scripts of Debian to see how things work. It’s text.

    Debian is only one of many organizations that have many eyes.

  24. Chris Weig says:

    Okay, let me rephrase that: the availability of FLOSS code is of no interest to users, and it is only of marginal interest to programmers.

    But now that you’ve mentioned it, I can’t indeed remember that I’ve ever stumbled about such a thing as a public code review for FLOSS code. Except for the kernel where hopeful wannabe patch writers are shot down on the mailing lists. I can remember code reviews for open sourced code that was proprietary before. For example, there’s this guy who has published code “reviews” for the open sourced id software stuff. Just recently he reviewed Doom 3.

    Where are the many eyes?

  25. oiaohm says:

    Clarence Moon
    “Only the dullest of the breed have any desire to read existing code. The innovative and creative developers are always off working on their own new inventions.”

    Nice way to put how companies stuff up. The higher who they see as the most inventive but don’t look for a particular class. They type that you need in lots of companies get though of as dumb what is incorrect.

    There are different classes of developers. Some do truly love debugging and code repairing. Companies like Redhat and IBM became aware of this. People like this normally end up being called distribution package maintainers.

    The good ones of this class are highly inventive and exists to the group who loves doing code audits. Why they have loves of creating tools to find defects automated. To write those tools you have to read code study code. Work out why you program false positives and false negatives. The advantage of the good ones they go away for 2 weeks your coder teams work will be automatically checked by the build system to reduce the number of errors going in.

    “They can be beaten into doing it during mandatory code review periods set prior to new release points, but they do not do it willingly.”

    This should be rolling code reviewing automated. With a specialist who job is to study the code and make the automated tested better strong and more and more effective so is always reading over the code for bugs to see what can be added to the automated system.

    This is because you don’t have any of the class of defect detection tool construction personal that you have to beat the new ideas for clients programmers over the head todo code review that can put them off there game. Some ways the ones you need are the vampires of coders that the other coders hate and don’t bring profit other than dependability of results.

    Yes those who love defect detection tool construction most likely will not have the interest in the rest of the project you are working on. Highly good at making sure code is quality but producing anything else they can be next to worthless since there creativity does not work that way. Creative at thinking how the code might malfunction they are. You would call them the world half empty of coders.

    The type that move on to new projects all the time are the half full of coders. Yes it is the same personality trait you can partly sort them by what they call a cup half full/empty of liquid. FOSS gets a mixture of both.

    Something to be aware of the class that love defect detection can be completely hated by there peers for being a nit picker.

    Lot of companies make this mistake thinking the people who are coder audits are not inventive people. Yes it does help having the automated system since this can also create competition for the most number of prevented defects.

    Beating them to code review is bad management and sign of lacking some of the right staff.

  26. Clarence Moon says:

    Lower costs, particularly when it’s a choice of IT or no IT is a huge practical, not theoretical, benefit to everyone.

    Doubtless the freeloaders of the world are enthusiastic fans of not paying for software, that is what makes the software piracy world go around. But it is not much use to those who want to sell their talents in order to eat. In isolated instances, some companies make money from providing support or utility services to the freeloaders using Linux and some, such as Red Hat, even make a very comfortable existence from that, but, by and large, the big money and the resultant developer salaries, are in commercial companies such as Microsoft, Oracle, Adobe, Intuit, Symantec, etc., that have made it to the front of their markets and are swimming in cash.

  27. Clarence Moon says:

    The availability of source code is interesting for programmers

    Not as far as I can see, Mr. Weig. Around here, developers shun reading other people’s code like the vampires avoid the sunlight. They can be beaten into doing it during mandatory code review periods set prior to new release points, but they do not do it willingly.

    Only the dullest of the breed have any desire to read existing code. The innovative and creative developers are always off working on their own new inventions.

  28. Chris Weig wrote, “So for most users the advantage of FLOSS is purely theoretical.”

    That may actually be true except that even modestly sized organizations may well do some coding and share it with the world. The Linux kernel is a prime example of that, but so is LibreOffice. SUN bought StarOffice for less than one round of licensing and hardware for PCs. Their employees were already using a UNIX OS and were using PCs just for word-processing. That purchase was eventually shared with the world and hundreds of millions benefit from it. Apache web server has also been supported by a lot of businesses and shared with others large and small. FLOSS does work and for many situation is a better way to do IT.

    Chris Weig and others need to think outside the box Wintel has put them inside. GNU/Linux and other FLOSS is a cooperative project of the world and it is globally a net benefit regardless of particular situations. Lower costs, particularly when it’s a choice of IT or no IT is a huge practical, not theoretical, benefit to everyone. For example there are still schools that struggle with 10 students per PC with that other OS. They can easily move on to 3 students per PC for a similar cost. That makes the difference between IT that is barely accessible to IT that is always accessible. Where I last worked, students had access to 40 PCs but only half were running. When I left they had 80 all running. The increase in PCs only cost them freight and very low maintenance, and FLOSS which made it all possible.

  29. oiaohm says:

    Chris Weig
    “So they have to hire someone to do it for them. Is that a realistic option when it comes to consumer software?”

    Narrow minded twit here. Part of support contracts for Redhat and other distributions is so many hours of programmer time as part of the support contract. So amount of spend required is very minor. Yes these hours are usable even for internal projects that no one will ever see.

    So if you are using Linux full time the cost is quite small in a lot of cases normally due to support contracts here and there you have hours of developer time that are going to expire not used.

    “With FLOSS they have the code but lack the capability to change it.”

    Really they could learn. python that the tool you were talking about is written in is a learning to program language. Its a macro language in many programs.

    So its not realistic for end users to write macros????. This is effectively what you are saying with Quod Libet. Really python coders are really common. Python is standard for 3d animation work.

    About time you look at were you are standing. Quod Libet is completely written in a novice language to program in.

    FOSS they never lack the capability to change it. Question is will they put the effort in you see kick starter and other times around the places pool up money between many users to pay for a feature to be implemented they need.

    So if the people wanting ID3v2.3 got unified and kicked in 10 dollars each they most likely would have enough money to pay a coder to make the plugin.

    Lack of cooperation between end users is the problem. Also lack of willingness to pay for features is another problem.

    Blender lot of its features have been got by people chipping in 10 dollars here 20 dollars there until enough money was stock piled to get the feature coded.

    Yes less expensive than paying for a closed source program license to get the feature.

    Chris Weig you are just the normal idiot expecting something for nothing. Not thinking you need to cooperate to get anywhere also not understand for us Linux users we have coder hours to spend every year.

    This is why it suxs to be using Microsoft you don’t get coder hours.

  30. Chris Weig says:

    Chris Weig there is no such thing as a free lunch. Even in the FOSS world you have to pay.

    Sure, you can go out in the nature where nothing’s yet anybody’s property (those spots may be getting tougher to find) and eat what you find there. It’s free — in a monetary sense.

    Be that as it may, you’ve just confirmed something. For the majority of users, who happen to be no programmers, there’s basically no difference between FLOSS and CSS (closed-source software). With FLOSS they have the code but lack the capability to change it. So they have to hire someone to do it for them. Is that a realistic option when it comes to consumer software? No. Is that probably more expensive than buying a software license? Yes. So for most users the advantage of FLOSS is purely theoretical. The availability of source code is interesting for programmers, but not for individual users.

  31. oiaohm says:

    Chris Weig
    “There’s no need to put software out, if you’re not prepared to deal with users’ wishes.”
    Who says the wish for ID3v2.3 is the majority of users. You?? Pure arrogance. Proper support of ID3v2.4 will be important to other users.

    Wishes are at times conflicting. Developers don’t have the time to look at every option just the option that was put forwards. The option of a plugin using a different library todo ID3v2.3 on that project has never been put forwards. So that might be a yes or no answer. More likely yes since it will not cause breaking of ID3v2.4 support.

    “The developer has always more power, because he’s the one who can code.”

    And anyone can commission IBM or Redhat or many other developer time seller to have a developer work on for them doing what they want.

    This is the important point are you paying the Developer a wage or paying something to have the developer do what you want. All those cases no. Are there ways to pay a developer to work on that project to do what you want yes. Even pay a developer to fork the project.

    Basically since you are not willing to spend money you don’t have control. You need to consider the other users as well who don’t own devices that need ID3v2.3 to work. So need ID3v2.4 that is perfect.

    Really is very reasonable thinking that most people for mp3 use phones or other devices not to support ID3v2.3 as default.

    Chris Weig there is no such thing as a free lunch. Even in the FOSS world you have to pay.

  32. Chris Weig says:

    Failure in developer meeting end user is not one sided.

    There’s no need to put software out, if you’re not prepared to deal with users’ wishes. Even if your software is FLOSS, there’s a kind of informal contract with the user. Availability of source code doesn’t exempt you from reacting to users’ wishes in a meaningful way. Lambasting them is not meaningful. The developer has always more power, because he’s the one who can code. The user in almost all cases cannot. Therefore the developer holds a greater responsibility. And this responsibility is all too often defined away in FLOSS by referring to the source code.

  33. oiaohm says:

    oldman
    “And you my dear sir have just given the primary reason why I prefer closed source. When a developer/company has some skin in the game ( to quote my fellow “troll”) they tend to listen better and respond to their “customers””

    In fact I have had the other. Where closed source does not listen. At least with FOSS I have the option of commissioning a coder to do as I direct without question.

    When you are not paying you have to be considerate and listen. This include not being tunnel vision and remembering the old saying there is more than one way to skin a cat.

    oldman
    “Demanding that software that you don’t pay for actually work is totally unreasonable.”
    This is not exactly correct. When you are not paying you have to accept your limited role.

    If you are paying for the development or developers working on the project you have more of a guiding force.

    Nothing is 100 percent free.

    Also oldman you are rude as hell when it comes to your treatment of open source. People wonder why they have problems. Yes being willing to spend cash and/or trade is part of FOSS.

    You have the point of view that closed and foss are magically massively different. The big question is always are you paying for control of the development or just have to take what is dished up.

    Paying does not have to be just in cash. To make a game I wanted work in wine to work. I gave my time test casing related applications for the developer.

    Lot of people who have problems with FOSS products are not paying in either cash or trade. They expect something for nothing. Something for nothing is rude.

    Most common and the worst mistake is thinking just because you put forwards a idea that you don’t have to listen and take on board why its rejected and try again this time with something more suitable to the project objectives.

    The FOSS programs have to work for the many not just the few.

    ch I have run 20 year old stuff recently on Linux. Its not that hard. In fact some stuff I run on Linux is pre Linux Unix stuff for access archives. Chroot idea is a old idea of Unix to allow running very old binaries. Very few modifications are required. In fact seccomp filter will make it simpler to run older binaries by providing a way to userspace trap system calls so allowing support back to first version Linux and some pre Linux UNIX system that use the same syscall addresses.

    ch year of the desktop you did not have to look far each time to see that the items were not in place.

    The year of the multi-seat and enterprise desktop is closer. Than the general desktop.

    Java has evolved lot of java ideas of gui design are inside android.

  34. oldman says:

    “Really your example shows rudeness on the part of end users. ”

    Quite true actually. Demanding that software that you don’t pay for actually work is totally unreasonable.

    And you my dear sir have just given the primary reason why I prefer closed source. When a developer/company has some skin in the game ( to quote my fellow “troll”) they tend to listen better and respond to their “customers”

  35. oiaohm says:

    Chris Weig Quod Libet can be forced todo ID3v2.3 tags. Add an addon using one of the other taging libraries. http://wiki.python.org/moin/UsefulModules#ID3_Handling

    ID3v2.3 tags aren’t supported is not exactly true. Reading ID3v2.3 is supported its the writing bit that is missing from the default back end.

    pytagger could be used for the legacy crap that is ID3v2.3. As a add on for those users who have legacy players.

    And as per the FOSS world if there is enough demand a project not doing the will of its end users should be forked.

    The fault here is sometimes end users tunnel vision. Demanding a feature by default instead of putting for an idea for a addon. Yes letting ID3v2.3 die out is a good thing in the default setup. Adding a extension so those who need ID3v2.3 are not harmed is still a valid option.

    Chris Weig requests have been wrongly formatted to get around problem. Listening is require on both sides. ID3v2.3 contains bits that don’t conform to the ID3v2.4 standard.

    Really your example shows rudeness on the part of end users. Developers says no to doing something instead of going hmm how could we comprise this without head butting the developers. Add extra back end for tagging until ID3v2.3 dies out. This way the developers are not stuck with that crud forever.

    Failure in developer meeting end user is not one sided.

    “That may be true from a certain point of view, but for the user who simply wants to use his MP3 player it’s irrelevant.”

    See complete disregard for developers wishes by end user then end user wonders why there wish is not granted. Respect is required. End user has not asked why developer does not want to merge that into where asked. If you look at the requests they are requesting that the feature be added to Mutagen. Yes I would say no. Some formats Mutagen supports adding tags to ID2v2.3 tag on them is in fact invalid. Not that Quod Libet adds a library for doing legacy tagging.

    I have a prehistory mp3 player that only support ID3v1 tags. There is no upgrade for it either.

    So yes adding multi tagging options would open a slippy slope.

  36. ch says:

    “They invested $billions in Lose 32.”

    And made – and still make – many $billions from Win32 application.

    “ISVs in great numbers will port to something more durable than Ballmer’s whims… like Java, or GNU/Linux or PHP.”

    As far as durable is concerned, I mentioned that I installed Win8 RP on my netbook. Well, I can still run some Win16 games on it – about 20 years after they were released. OK, Win16 isn’t supported on 64-bit versions of Windows, but I think that’s acceptable. Linux is about 20years old, too, but don’t even think of trying to run some 20-year old SW on a current Linux version. Oh, and Java and PHP both date back to 1995, so they are actually younger than Win32 – not to mention that for desktop-side applications, Java is pretty much dead and PHP not even an option.

    “M$ killing Lose 32”

    Why should MS do that? Win32 will still be around for years.

    “They will gladly migrate to */Linux”

    Yeah, the YEAR OF LINUX ON THE DESKTOP is coming … again … very soon … really … this time it will …

  37. Chris Weig wrote, “It only writes ID3v2.4 tags, ID3v2.3 tags aren’t supported. Unfortunately many MP3 players out there only properly support ID3v2.3.”

    With FLOSS, however, the user has a choice and can use other software. Wintel likes to take away such choices for key pieces of software.

    Debian:
    “Package: quodlibet (2.2.1-1)

    audio library manager and player for GTK+

    Quod Libet is a music management program. It provides several different ways to view your audio library, as well as support for Internet radio and audio feeds. It has extremely flexible metadata tag editing and searching capabilities, using the same interface as Ex Falso.

    Supported formats include MP3, Ogg Vorbis, FLAC, Musepack (MPC), WavPack, and MOD/XM/IT.”

    The thing probably plays files that it writes so the only functionality “lost” is exporting to other MP3 players. Of course exports could be done in other formats supported by other MP3 players. My little $20 gadget has played anything I have sent it.

    So, I call straw man, a pretty weak argument.

  38. Clarence Moon says:

    The difference is that ISVs are not going to rewrite everything to suit M$ this time

    From a developer’s point of view, your statement is nonsense, Mr. Pogson. The Win32API is not being totally deprecated by Microsoft overnight, rather it is Gartner’s prediction that the consumers will shift their preferences over time and “conventional” apps will fall out of favor, just as command line and text screen versions of things like WordPerfect or 123 went away. Or, for that matter, apps that required the Win16API long ago.

    Product code is mostly recoverable from Windows API version to version outside of presentation layers that naturally must adapt to any new paradigm.

    Companies like to hype things in order to gain some recognition, but the latest round of Windows development is still just an evolution of the past and is not any sort of revolution. Microsoft will endure regardless.

  39. oiaohm says:

    oldman “You wish the linux desktop was as easy as android!”

    Exactly why can it not be. This is the problem the joke is on you oldman. Android is laying foundations.

    Please beaware wayland works inside android in prototype. Android interface can be just another interface management system.

  40. oldman says:

    “Treating GNU/Linux like Android is more natural fit.”

    HAHAHAHAHAHAHAHAHAHAHAHA…

    You wish the linux desktop was as easy as android!

  41. Chris Weig says:

    Chris Weig yet it is possible to install third party apps on android devices directly.

    But who does this sort of thing? Not the majority of Android users.

  42. Chris Weig says:

    No. I’m not claiming that free (as in beer) software can’t be good. I’m claiming that many free (as in speech) software authors have a rather particular mindset which usually works to the detriment of the user. It’s basically the unchecked egotism, its pure form, that according to Linus Torvalds forms the backbone of free software.

    One example of the top of my head, which illustrates what I mean: there’s this nice, capable audio player called Quod Libet. It only writes ID3v2.4 tags, ID3v2.3 tags aren’t supported. Unfortunately many MP3 players out there only properly support ID3v2.3. But Quod Libet’s authors don’t want to bow to users’ wishes, instead stating that it’s the industry’s fault. That may be true from a certain point of view, but for the user who simply wants to use his MP3 player it’s irrelevant. That the source code is available, and that you could fix the problem yourself is the cozy excuse for such things. Ideology beats pragmatism.

  43. oiaohm says:

    Chris Weig yet it is possible to install third party apps on android devices directly.

    Chris Weig android is getting users use to the idea of a strictly controlled environment.

    Something had to pave the way.

    “FLOSS ain’t prepared for users” Mostly because users try to treat Linux like Windows and wonder why it bites them. Treating GNU/Linux like Android is more natural fit.

    Yes there has been a culture shock problem.

  44. Mats Hagglund says:

    “FLOSS ain’t prepared for users…”

    Did i hear a voice “there ain’t a free lunch…” there?

  45. Chris Weig says:

    FLOSS ain’t prepared for users. Not outside of a corporate or strictly controlled environment. That’s also why Android works. Because it is, after all, a strictly controlled environment.

  46. Mats Hagglund says:

    Another “market share” showing the decline of Windows both in mobiles and non-mobiles:

    http://stats.wikimedia.org/archive/squid_reports/2012-06/SquidReportOperatingSystems.htm

    “Windows…70.58% “. April 2009 the figure was 89,50%. Even non-mobiles are showing decline of Windows (89,50%—>~85%)

  47. oiaohm says:

    Clarence Moon really you failed to read product life cycles properly.

    When a product cycles to end of life. The replacement does not have to be from the same company.

    So android technically could displace Windows.

    Windows RT on tablets puts it head to head with Android.

    Are end users getting use to the Android interface yes they are. Windows RT is really Windows vs Android the victor goes everything. Problem is android still plays a trump card. Just like OS/2 could run Windows applications. Windows RT can run Android applications. So why as an application developer will I bother making Windows RT applications.

    History repeats and unless something changes MS is screwed. Everyone bar Microsoft with Windows RT was talking about including Android runtime out box. Problem is third party will also off this for MS surface devices.

    So the non apple tablets for 100 percent OS coverage with least number of apps made is make an Android application.

    Microsoft threatened a long time ago that they would kill Google. It looks like Google is going to kill Microsoft.

  48. Clarence Moon wrote, “If Microsoft supplies RT to OEMs and people get used to it, then RT is going to be what is bought in 2020, just like Windows was the GUI OS after DOS command lines went away for good, only moreso.”

    The difference is that ISVs are not going to rewrite everything to suit M$ this time. They invested $billions in Lose 32. M$ does not pay their salaries. ISVs in great numbers will port to something more durable than Ballmer’s whims… like Java, or GNU/Linux or PHP. Anything but something controlled by M$. They will produce web/cloud applications galore.

    DOS was a monopoly granted by IBM. ISVs naturally went along with that. When the GUI came along riding on DOS there was not a lot of resistance. M$ killing Lose 32 means $billions in software will become worthless within a few years. ISVs will want to cut M$ out of the equation this time. Consumers don’t owe M$ a living either. They will gladly migrate to */Linux if they can browse the web with it. They also have no loyalty to Intel. Munich had it right. Independence from M$ is valuable and worth the effort. Independence from Intel will save everyone a lot of cash.

  49. Clarence Moon says:

    I think Gartner is right about that other OS sinking into oblivion in a decade but I don’t see M$ being able to dictate the shape of IT beyond the horizon.

    You have a wrong view of the future, Mr. Pogson. Going from “conventional” Windows to RT is more akin to the shift from command line entry to GUI presentations. The underlying OS is going to be whatever people think it has to be. If Microsoft supplies RT to OEMs and people get used to it, then RT is going to be what is bought in 2020, just like Windows was the GUI OS after DOS command lines went away for good, only moreso. There is no OS/2 for Microsoft to contend with, simply the opportunity to move from Winforms to Metro. There is no product market change here, just a gradual migration of preferences for look and feel.

    You still haven’t read about Product Life Cycles, I take it?

  50. oiaohm says:

    Really the slide started with Munich and others with the first to Linux migrations.

    That have basically proven 80/20 split. 80 percent of uses don’t need MS products. 20 percent do. Most of those can be serviced by virtualization. Reducing operating costs.

    Basically 80 percent of Microsoft current business market could go by by no problems.

Leave a Reply