On Mon, Apr 21, 2008 at 4:29 AM, Quim Gil <[EMAIL PROTECTED]> wrote: > You keep talking about the quite stable desktop architecture. The > relevant comparison for this discussion are Linux distros or whatever OS > installed in devices fitting in your pocket offering long term support > (say 3 years) in 2005 (or today). > Not really true. This is just a weak excuse for continuing current (bad) practice. Desktop hardware is changing just as fast, and just because a subsystem performs the same apparent function doesn't mean the same drivers will work, and in fact the opposite is true. Have you tried finding OS and app support for multi-core cpus? Even 64-bit support is scarce, and it's been around for several years.
> 1. From a strictly engineering point of view: look at the desktop > hardware in 2003 and now, look at the mobile hardware evolution in the > same time frame. The fact is that mobile hardware architectures are far > more unstable and this create extra hassle for platform development, > leave alone stable APIs. Sorry, but this simply isn't true. The majority of the mobile hardware has been around for a long time now, in the same way as desktop hardware. Making changes to desktops is easier not because the hardware is older, but because the hardware is better supported by its manufacturers than mobile hardware. Swapping out a desktop video card is easy because full-featured drivers come with the new card, not because the drivers are the same as for the old card. For example, GSM modems haven't changed any more than desktop subsystems, and neither have WiFi or bluetooth etc. Most of the changes have been in software and form factor. The real difference is that mobile hardware were allowed to get away with being more closed from the beginning, and are fighting tooth and nail to keep it that way. Fortunately, it looks like they may be beginning to lose their advantage. > > 2. Repeat exercise 1, now looking at top use cases in the desktop and in > mobile devices. Most computer use cases are same or similar, just deeper > and faster - requiring more memory or clock speed - but that's it. In > mobile devices the usage is quite different, getting all kinds of new > hardware pieces, APIs and performance issues. > See above. > 3. The Nokia tablets don't have planned obsolescence. The > online/multi/media context and the consumer expectations make them > obsolete faster, just like most mobile products. Online video, full AJAX > and long etc can't be easily fulfilled with old mobile hardware. You can > do some miracles on the software engineering side but at what cost and > in exchange of what. *ALL* consumer products have built-in planned obsolescence. The trick is in figuring out whether it's both necessary and acceptable due to changing needs, or artificial and premature due to the manufacturer's desire to force customers into otherwise unnecessary purchases. This is the case in the USA with digital TV. Consumers weren't buying new, extremely expensive hardware fast enough to suit the TV manufacturers, so they finally made analog TVs obsolete. HDTV is great for home theaters, but in most real-world situations it's extreme overkill. (Yeah, like I really need 1080p - or even 720i - for that 20", 12" or 7" TV, never mind hand-held devices...) The other aspect is that the standards deliberately further limit one's ability to view legally purchased content in one's own devices. It's all about making people pay many times for the same thing. > > 4. If what you have is working for you now then it should just be enough > to keep using your device like the day you bought it. However, what > happens in most cases is that a new device in a new category (like the > 770 was) gets new use cases from the people who bought them than > established products. > That's just naive. Like I said before, the N800 has the exact functionality that I want, and the N810 has little attraction for me. However, I'm speaking specifically of the hardware *possibilities*, which can only be realized with software, much of which does not ship with the device, and much of which still isn't even available from the community. The potential of the device has barely been tapped much less completely fulfilled. Yes, first-generation devices often are more limited than later generations, because you have to start *somewhere*, and you learn with each generation. Meanwhile, technology marches on and new hardware becomes available. That doesn't mean that the old hardware is no longer useful. > > > > The last time I looked at Maemo, I found that there were a lot of > > programs that I couldn't even think of adding to my N770. > > http://maemo.org/downloads/OS2006/ reports 249 third party applications > for the 770's original OS2006. Note that this is more than the number of > applications found there for the N800/N810. > The number of apps is not really relevant. What's relevant is the specific applications that individuals want or need. As long as certain older apps aren't being updated for the newer models, and newer apps aren't being backported for the older, you're comparing apples with oranges, and *nobody's* needs are really being met. > > > > Has the OS changed so much that backporting it to the 770 would be that > > hard? > > Yes, specially if you need to respond to the performance requirements of > a sales product. > It seems to me that that is as much a development environment issue as anything else. > > > If so, how do I know that if I buy an 800 or 810 that we won't > > be having this same conversation when a 820 or 900 comes out? > Unfortunately, you are guaranteed that sooner or later that will in fact be the case. That's the nature of capitalism. As with many things in life, though, you just have to decide if whatever you're getting involved with is worth the investment (time, money, whatever) to you. You're the only one who can make that decision. Mark _______________________________________________ maemo-users mailing list maemo-users@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-users
