Announcement

Collapse
No announcement yet.

Matthew Garrett: How-To Drive Developers From OS X To Linux

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #81
    Originally posted by Luke View Post
    I would not own Apple hardware unless I found it in the dumpster. The main reason is Foxconn, whose moved their motherboard factory from Taiwan to mainland China for cheap labor. They are so abusive to their workers there that there have been persistant reports of worker suicides. Apple has traditionally used Foxconn motherboards. An iPhone factory in China reported similarily horrid conditions. Apple is now moving some production back to the US, but I suspect that will be assembly, not motherboard production.

    Apple will probably be doing what many of us do when we "build" a computer: assembling parts built elsewhere. Making silicon dies and computer motherboards with hundreds of traces far exceeds my fab capabilities, I would not be surprised if this also is too much for Apple. Hell the original iPod was mostly three major off the shelf parts from what I hear.

    I have actually built boards for analog hardware, prototype boards built entirely fron scratch and always optimized on the fly from a basic design. No two are ever alike. It can be delicate and demanding work, even at very large fab sizes they are vulnerable to solder bridges and other such defects. The fumes can be an issue, and until recently lead was the big issue in commercial work. Still is in prototyping, where non-lead solder can be much harder to work with in installing small discrete transistors that really don't like heat.

    Now imagine mindlessly stuffing the same parts in the same holes 6 or 7 days a week for barely enough money to eat-and a huge pile of "debt" subtracted from your pay keeping you in bondage to your boss. That's the magic of Apple.
    +1000

    Comment


    • #82
      Originally posted by sandy.martel View Post
      I feel like it's the 90s all over again :-)
      That would be because Apple is in the same position is was in the 90s: in a death spiral heading towards irrelevancy. The difference between now and the 90s is that Jobs is dead (and thus can't save them this time) and the only developing device categories that they don't already have an entry into either nobody wants (smartwatches), or nobody is really sure that they want (smart glasses). Further the creation of the iPhone 5Cheap clearly demonstrates that Tim Cook does not understand how Apple operates and bought into the delusion himself that they actually have a better product, and that people weren't buying iPhones because they were so expensive, when the reality is that they simply sell an inferior Hardware/Software set, and that what Apple is really selling is an image, not a phone or computer. As a result when they actually try to compete on the basis of a phone or computer of course they're going to fail.
      Last edited by Luke_Wolf; 21 May 2014, 04:38 AM.

      Comment


      • #83
        Originally posted by Luke_Wolf View Post
        The difference between now and the 90s is that Jobs is dead (and thus can't save them this time) and the only developing device categories that they don't already have an entry into either nobody wants (smartwatches), or nobody is really sure that they want (smart glasses).
        To be fair, nobody "wanted" GUI PCs, hard-drive based MP3 players, smartphones, or tablets, either. At least few people were buying them before Apple released their version. What Jobs was good at was realizing that people actually would want these things, then stealing the bits and pieces of what everyone else was doing that were good and combining into a package people actually wanted to buy.

        There are three major problems with this approach.

        The first, as you pointed out, is that Jobs is dead, and remains to be seen whether anyone else in the company has the same knack.

        The second problem, which we saw with PCs, MP3 players, smartphones, and now tablets, is that you can only do this at most once every ten or twenty years. The ideas have to already be there to steal, and they have to result in a really big change in the product relative to what is already there. And once it happens, other companies see what Apple is doing and catch up quick. They have at most a couple of years before everyone else catches on. After that, Apple loses its obvious advantage. Once that happens, what wins new customers are the small, incremental innovations. And Apple has never been very good at that. They have their big revolutions, but between those the changes are fairly minor. So they secure a lot of customers during the couple years after the big revolution, but after that their sales rapidly fall off since they can't keep up with the smaller innovations everyone else is making. The high price compared to seemingly similar devices also doesn't help. So they get a group of customers early that tend to be extremely loyal, but they have trouble getting new customers after the initial window.

        Just look at the iPhone. The only clearly major change new features they have had over the last few years are Siri, which was originally developed for android and was just bought by Apple, and a 64bit processor which was just an implementation of an existing spec. Every other "new" feature had been in Android for years.

        The third problem is that this approach only works if they are the only ones who can do it. If other companies figure out how to do this, and combine it with the incremental improvements Apple doesn't do, then Apple is in big trouble. I think companies like Google have the potential to do something similar to Apple, if the opportunity presents itself.

        The big test is smart watches. I think there is certainly a potential there, and Apple appears pretty likely to release their own soon. If it is really something new and much better than the upcoming Android smartwatches, I am watching (pun intended) the Motorola one particularly, then Apple may still have some of Jobs left in it. If it just a watch with notifications and Siri, then I don't think we should be looking to Apple for revolutionary new ideas anymore.

        Comment


        • #84
          Originally posted by bridgman View Post
          Sure, but you can murder and rob people and not get caught, doesn't mean you're going to do it or even that it happens very often.

          I don't know about other companies, but we work really hard to make sure that we do not violate GPL or other license terms.
          Indeed. All parts of the industry I have been in contact with is very much clear on avoiding GPL violations. I can only assume that Vim_User has no experience on how these things work.

          I do believe that the choice of permissive licensing of Xorg and Wayland gives food for thought though. Why was permissive licensing chosen, and how did it work out? Apple is now moving along with their own XQuartz fork, Android made their own Surface flinger, and Ubuntu opted for Mir with GPL and CLA. I guess that leaves us with supporting the BSDs as the only argument for dropping copy-left. At the same time the leading BSD, FreeBSD, is primarily marketing itself with zfs, a copy-left filesystem with a license specifically crafted to deliberately leave linux out in the cold. So exactly how important is it for BSD to have Wayland permissively licensed? At the end of the day, I believe it was a terrible error to not use GPL for Wayland. GPL has consistently proven to be the best vehicle for building communities. I do believe that Wayland would see much more community involvement had it chosen a copy-left approach.
          Oops, just occured to me that I forgot Sailfish. Not that it changes any conclusion. All in all permissive licensing has demonstrated its main weakness on the display server, namely the incentive for proprietary forking. It is costing us all dearly, I just pray and hope that it will not single-handedly postpone the viability of linux on the desktop (maybe it already has).

          Comment


          • #85
            Now imagine mindlessly stuffing the same parts in the same holes 6 or 7 days a week for barely enough money to eat-and a huge pile of "debt" subtracted from your pay keeping you in bondage to your boss. That's the magic of Apple.
            Hopefully the automation revolution can supplant this labor soon. We have had the technology for a decade, but somehow Chinese workers were cheaper than the up front investment in robotics. Some companies are moving factories back into the US as automated assemblers with a skeleton crew, finally. The automobile manufacturers post 2008 were really big on the idea.

            Comment


            • #86
              Originally posted by Del_ View Post
              I do believe that the choice of permissive licensing of Xorg and Wayland gives food for thought though. Why was permissive licensing chosen, and how did it work out?
              Permissive licensing is the norm when the priority is to establish a standard implementation with broad adoption. I think X has been pretty successful in that regard.

              Remember that X grew up in the days of proprietary Unix workstations, where binary distribution was the norm, and that the workstation vendors contributed (heavily AFAIK) to the development of X implementations. In order to get the broadest adoption (ie on both proprietary and open source target systems) the permissive licensing was a hard requirement.

              Same goes for things like Berkeley TCP/IP, where the licensing was permissive with the specific intent of driving adoption into proprietary OSes, in order to establish a standard implementation and make the internet work a bit better. Again, I think it's fair to say that was successful (even MS used it as far as I know).

              If the priority is encouraging individual (rather than corporate) volunteer contributions, then a copyleft license is generally considered more attractive, but it really depends on the specific circumstances rather than one licensing model just being "better" than the other.
              Last edited by bridgman; 21 May 2014, 09:48 AM.
              Test signature

              Comment


              • #87
                Originally posted by TheBlackCat View Post
                *snip*
                There's a few major problems with smartwatches:

                1). Cellphones killed normal watches for most people. As a result you have an entire generation of people who have never worn a watch and do not want to wear one, as well as a few generations of people who did wear watches but now just use their cellphones to keep track of time.
                1a). Watches are uncomfortable
                2). Smartwatch Screens are too large to be comfortable and too small to be useful
                3). As a result of 2 even smartwatch proponents have a hard time coming up with use cases for the devices beyond notification systems, and hooks into the voice command system.
                4). Smartwatches cost way way too much for what they're offering

                On the other hand smartglasses do have potential although the formfactor needs to change where you have full on lenses as the screens along with generally making them more aesthetically pleasing, and the price needs to come way down.

                Comment


                • #88
                  Originally posted by Del_ View Post
                  I guess that leaves us with supporting the BSDs as the only argument for dropping copy-left. At the same time the leading BSD, FreeBSD, is primarily marketing itself with zfs, a copy-left filesystem with a license specifically crafted to deliberately leave linux out in the cold. So exactly how important is it for BSD to have Wayland permissively licensed? At the end of the day, I believe it was a terrible error to not use GPL for Wayland. GPL has consistently proven to be the best vehicle for building communities. I do believe that Wayland would see much more community involvement had it chosen a copy-left approach.
                  First off... Big Whoop on FreeBSD using ZFS, Linux is just as guilty on the filesystem front with ext4 and Btrfs being under GPL and thus unusable by the BSDs, and further it's vindictive to exclude the BSDs on something that is kernel agnostic like a display server protocol. Additionally unless that display server is LGPL or looser you'd be tainting the license of anything that is to be run on it, which is actively hostile to application developers. Long term the BSDs are probably all going to use HAMMER2 anyway.

                  Second off you're presuming that
                  A). There are developers who are refusing to work on wayland because it's not copy-left
                  B). That there are individuals forking wayland and not making the source available
                  Please show evidence for either of these cases

                  Originally posted by Del_ View Post
                  Oops, just occured to me that I forgot Sailfish. Not that it changes any conclusion. All in all permissive licensing has demonstrated its main weakness on the display server, namely the incentive for proprietary forking. It is costing us all dearly, I just pray and hope that it will not single-handedly postpone the viability of linux on the desktop (maybe it already has).
                  Given that all of the UNIX-like and UNIX OSes other than OS X, use X speaks highly of permissive licensing. The problem however with X, which led to the creation of Wayland, Quartz and Surfaceflinger is that the protocol is both expansive and broken, and I would like to point out here that neither SurfaceFlinger or Quartz are forks, they are independent developments. This is not a problem of the permissive license but of a protocol that needed to be thrown out and rewritten from scratch, which has been being worked on and is known as Wayland. The primary reason we've been waiting so long on wayland is that the toolkits needed to develop support for it and the desktops have needed to refactor significantly to account for handling wayland. Further I can guarantee you with absolute certainty that wayland taking this long has had no effect and will have no effect on Linux's viability on the desktop.

                  The simple reality of the situation is that while the infrastructural stuff could be significantly better, it is Good Enough (tm), what is holding back the linux desktop is the lack of 2 things:
                  1). Applications, and in particular games
                  2). A company with enough clout and interest to bring Linux to the mainstream desktop.
                  Valve is the answer to #1 and likely the answer to #2. Wayland, systemd, btrfs, and friends are definitely all niceties but they are totally irrelevant to linux desktop viability.
                  Last edited by Luke_Wolf; 21 May 2014, 11:48 AM.

                  Comment


                  • #89
                    license

                    Originally posted by Luke_Wolf View Post
                    ... Linux is just as guilty on the filesystem front with ext4 and Btrfs being under GPL and thus unusable by the BSDs, and further it's vindictive to exclude the BSDs on something that is kernel agnostic like a display server protocol. ...
                    Given that all of the UNIX-like and UNIX OSes other than OS X, use X speaks highly of permissive licensing. ...
                    I agree.

                    Comment


                    • #90
                      Originally posted by wizard69 View Post
                      You obviously have no idea here. Windows is a broken platform stuck to far in the past to be even meaningful any more.
                      Yes, it's broken, but os x is even more broken. Windows offers much more software and better 3D performance.

                      Comment

                      Working...
                      X