All the code in the world is useless until I can actually change the software on my phone and make it do what I want and not what some phone company thinks I should and shouldn't be able to do.
Which is why the OpenMoko Neo FreeRunner is so good. Decent hardware for a phone (including touch screen, GPS, tri-band GSM, WiFi and Bluetooth) with almost all of the source for the phone being 100% open (and replaceable).
The only closed bits are the GSM stack (which runs on a seperate baseband processor and talks to the host CPU via 100% documented open standards, all the stuff you need to know to talk to the baseband is documented and open), the driver for the GPS chip (people are reverse engineering it and making an open source replacement) and some of the fancy stuff to do with the GPU.
And with regards to the GPU (which is aparently being dropped from the next model), the only closed thing is the official docs and specs provided to the OpenMoko team from the GPU vendor. The GPU vendor is quite happy for the OpenMoko team to produce and open source a driver for the GPU (and even a new set of specs for it), they just dont want any code or specs created by THEM being released publicly (having everything that goes public created by a 3rd party helps with legal issues I guess)
The hardware is as open as they can legally go too. For example, they have released the same CAD drawings for the case and such as they themselves used to produce the molds for it. So if you want to make a new case in a color (or material) they dont offer (such as a rubber case so it can survive being dropped on the ground), the info is there.
The thing is that's all well and good for a tech based community who want that, but at the end of the day a Mobile Phone is a fast moving consumer electronic product.
People buy for a lot of reasons, colour for example, or "shiny". Most people want something cheap, that does the job.
This is a spoiler for MS and Google.
I'm going to wait and see. I've been working in the phone industry for 7 years now and I've not seen anybody fix a new device issue that quickly. I've seen lots of failed launches and multi-month delays though.
If Apple haven't used AT&T for the field testing (and some contacts tell me that is the case) then they've a problem.
Most of the MS Smartphones out there are based on Quad band GSM. They'll work anywhere with a GSM service. In the last 3 weeks I've used mine in Taiwan, Hong Kong, China, England, Denmark and Sweden.
Interesting points; I'm not surprised to hear that about the RAZR (nobody in the industry likes it) but I am surprised by the Ericsson comment - I've had 3 SEMCs (P910i, T68, T610) and they all had amazing radio power management. They're pretty much the best in the business in my experience. Most of the Windows Mobile problems have been down to under estimating the effectiveness of modules versus ground up phones.
I've already bet some friend's outside of the industry about the reboot frequency. Even if the device works flawlessly on your test cases, I just don't hear that Apple have been running live network field trials and without those they're going to be in a world of hurt.
I still reserve the right to be unimpressed with the touchscreen. No amount of clever design will stop my ears being greasy (uck, I know) and if, as I suspect, they use a tactile feedback on the touchscreen then that will bleed power over time - not to mention the need for a back light to actually see anything on every occasion.
Why is it that much harder in the phone world?
I suppose "because it is" is a little glib?
Part of the problem is in the radio module and stack design and why celular networks are much more complex than Bluetooth and Wifi. They don't have to worry about constantly monitoring the radio landscape, looking at the cell sites and handling the hand off between different Base Stations all while handling data and/or voice traffic. I believe that Apple have tried to get around this by buying in a 3rd party radio module. While I've seen this work ok(ish) on data enabled devices, generally speaking the results have been less satisfactory for actual mobile phones.
The problems tend not to be obvious at first glance, but I suspect users will soon find that the IP stack will hang randomly and not notice that its not working until they try to re-use it. As you can't take the battery out, then they'll need some way of doing a reboot easily.
The other problem you see is in the communications between the UI and the radio - so you can have a lot of problems answering calls, handling 3-way calling and that sort of thing. It will be very interesting to see what these devices behave like on the real networks. I know for a fact that as recently as 3 months ago AT&T were not being allowed to do network field trials and were still in the lab.
That alone rings alarm bells for me, as typically, even with phone platforms built from the ground up to be a phone (like you get from Nokia and Sony Ericsson) they'll have at least 3 months in the field testing the radio performance.
That isn't to say Apple can't solve these problems but I'm prepared to bet its going to cause them huge pain and negative PR.
Regarding the battery life; the problem is down to the natural conflict between radio power management and applications. Another thing I've seen on radio module based devices is there isn't a huge amount of optimisation for power consumption on the radio i.e. turning off the frequency band sweeps when you're out of coverage, so in poor signal areas the phone just runs through the battery until its all gone rather than knowing to stop bothering until the user touches it.
All of this is orders of magnitude harder to get right than Wifi and Bluetooth.
Oh, and touchscreen only phones also generally suck too.
No, it probably won't, anymore than it has done for the billions of mobile phones already out there. Running out of charge on an iPod is a pain; people will view running out of power on a phone more critically.
Slashdot Mirror
All the code in the world is useless until I can actually change the software on my phone and make it do what I want and not what some phone company thinks I should and shouldn't be able to do.
Which is why the OpenMoko Neo FreeRunner is so good. Decent hardware for a phone (including touch screen, GPS, tri-band GSM, WiFi and Bluetooth) with almost all of the source for the phone being 100% open (and replaceable). The only closed bits are the GSM stack (which runs on a seperate baseband processor and talks to the host CPU via 100% documented open standards, all the stuff you need to know to talk to the baseband is documented and open), the driver for the GPS chip (people are reverse engineering it and making an open source replacement) and some of the fancy stuff to do with the GPU.
And with regards to the GPU (which is aparently being dropped from the next model), the only closed thing is the official docs and specs provided to the OpenMoko team from the GPU vendor. The GPU vendor is quite happy for the OpenMoko team to produce and open source a driver for the GPU (and even a new set of specs for it), they just dont want any code or specs created by THEM being released publicly (having everything that goes public created by a 3rd party helps with legal issues I guess)
The hardware is as open as they can legally go too. For example, they have released the same CAD drawings for the case and such as they themselves used to produce the molds for it. So if you want to make a new case in a color (or material) they dont offer (such as a rubber case so it can survive being dropped on the ground), the info is there.
The thing is that's all well and good for a tech based community who want that, but at the end of the day a Mobile Phone is a fast moving consumer electronic product. People buy for a lot of reasons, colour for example, or "shiny". Most people want something cheap, that does the job. This is a spoiler for MS and Google.I'm going to wait and see. I've been working in the phone industry for 7 years now and I've not seen anybody fix a new device issue that quickly. I've seen lots of failed launches and multi-month delays though. If Apple haven't used AT&T for the field testing (and some contacts tell me that is the case) then they've a problem.
Most of the MS Smartphones out there are based on Quad band GSM. They'll work anywhere with a GSM service. In the last 3 weeks I've used mine in Taiwan, Hong Kong, China, England, Denmark and Sweden.
Interesting points; I'm not surprised to hear that about the RAZR (nobody in the industry likes it) but I am surprised by the Ericsson comment - I've had 3 SEMCs (P910i, T68, T610) and they all had amazing radio power management. They're pretty much the best in the business in my experience. Most of the Windows Mobile problems have been down to under estimating the effectiveness of modules versus ground up phones.
I've already bet some friend's outside of the industry about the reboot frequency. Even if the device works flawlessly on your test cases, I just don't hear that Apple have been running live network field trials and without those they're going to be in a world of hurt.
I still reserve the right to be unimpressed with the touchscreen. No amount of clever design will stop my ears being greasy (uck, I know) and if, as I suspect, they use a tactile feedback on the touchscreen then that will bleed power over time - not to mention the need for a back light to actually see anything on every occasion.
Why is it that much harder in the phone world? I suppose "because it is" is a little glib? Part of the problem is in the radio module and stack design and why celular networks are much more complex than Bluetooth and Wifi. They don't have to worry about constantly monitoring the radio landscape, looking at the cell sites and handling the hand off between different Base Stations all while handling data and/or voice traffic. I believe that Apple have tried to get around this by buying in a 3rd party radio module. While I've seen this work ok(ish) on data enabled devices, generally speaking the results have been less satisfactory for actual mobile phones. The problems tend not to be obvious at first glance, but I suspect users will soon find that the IP stack will hang randomly and not notice that its not working until they try to re-use it. As you can't take the battery out, then they'll need some way of doing a reboot easily. The other problem you see is in the communications between the UI and the radio - so you can have a lot of problems answering calls, handling 3-way calling and that sort of thing. It will be very interesting to see what these devices behave like on the real networks. I know for a fact that as recently as 3 months ago AT&T were not being allowed to do network field trials and were still in the lab. That alone rings alarm bells for me, as typically, even with phone platforms built from the ground up to be a phone (like you get from Nokia and Sony Ericsson) they'll have at least 3 months in the field testing the radio performance. That isn't to say Apple can't solve these problems but I'm prepared to bet its going to cause them huge pain and negative PR. Regarding the battery life; the problem is down to the natural conflict between radio power management and applications. Another thing I've seen on radio module based devices is there isn't a huge amount of optimisation for power consumption on the radio i.e. turning off the frequency band sweeps when you're out of coverage, so in poor signal areas the phone just runs through the battery until its all gone rather than knowing to stop bothering until the user touches it. All of this is orders of magnitude harder to get right than Wifi and Bluetooth. Oh, and touchscreen only phones also generally suck too.
No, it probably won't, anymore than it has done for the billions of mobile phones already out there. Running out of charge on an iPod is a pain; people will view running out of power on a phone more critically.