Advertisement
Front, Kernel, Security, Distributions, Development. See your byline here on LWN.net.
| |||||||||||||
|
| |||||||||||||
Openmoko: looking forward to 2009Openmoko: looking forward to 2009Posted Jan 6, 2009 4:07 UTC (Tue) by dw (subscriber, #12017)Parent article: Openmoko: looking forward to 2009
We asked Raster to integrate this keypad into Om 2008 and extend it to make it more hacker friendly (i.e., usable from places like the terminal). After two months of more or less silence he showed us his own version, written from scratch. Spot the fallacy near "two months of more or less silence". So for 2 months the project manager never once thought to ask the developer the status of this task? Leave any geek alone in a room full of his favourite toys for 2 months, and only one thing will happen. That's what the cat herders^W^Wmanagers are for! If the geek knew a from-scratch rewrite really wasn't the best plan, he'd probably be earning better money working as a manager himself.
(Log in to post comments)
Openmoko: looking forward to 2009 Posted Jan 6, 2009 6:54 UTC (Tue) by BrucePerens (subscriber, #2510) [Link] Add to that the fact that they had Harald Welte on the team, and yet they have had significant delays connected with device drivers, and the thing just barely works as a phone today. Now, it is hard for me to believe that this is a problem with Harald, but I can believe that it has been a problem with the way that the company has utilized Harald.Bruce
Openmoko: looking forward to 2009 Posted Jan 6, 2009 7:08 UTC (Tue) by daniels (subscriber, #16193) [Link]
Harald's absolutely fantastic, but you do realise that it takes more than one developer to write the kernel support for an entire phone, right?
Openmoko: looking forward to 2009 Posted Jan 6, 2009 16:09 UTC (Tue) by BrucePerens (subscriber, #2510) [Link] To evaluate that we have to consider what parts were not written when they started. The GSM stack is in another processor and there's the usual port-hole to it using the Hayes modem protocol. The radio is on the GSM side of the device. So, we are left with the GPS chip, the accelerometer, the audio chip, the video chip, the wifi chip, power management, on-CPU peripherals, and a glue ASIC or two. Some of those had existing drivers of some kind.
Openmoko: looking forward to 2009 Posted Jan 7, 2009 6:54 UTC (Wed) by daniels (subscriber, #16193) [Link]
And all that can be written by one person, including the core SoC bits? No.
Openmoko: looking forward to 2009 Posted Jan 7, 2009 7:06 UTC (Wed) by BrucePerens (subscriber, #2510) [Link] Well, I would have expected that between such a really good kernel programmer and software architect who had previous experience withe cell phones (see OpenEZX), and a staff of folks in Taiwan who had primary responsibility for the job, the job would have gotten done. What boggles me is that they had Harald doing other stuff.
Openmoko: looking forward to 2009 Posted Jan 7, 2009 7:22 UTC (Wed) by tajyrink (subscriber, #2750) [Link]
Basically everything is now working pretty much as intended on the driver side. Power management is not an easy job to get completely right very quickly, but it's getting there. One of the biggest reasons for delays has been the unfortunate choice of the graphics chip, which would give griefs to any developer team.
Openmoko: looking forward to 2009 Posted Jan 7, 2009 7:58 UTC (Wed) by BrucePerens (subscriber, #2510) [Link] Yes, it looks like a terrible choice, but it must also be that this was compounded by not having a good contract in place before design-in acceptance for the hardware.
Openmoko: looking forward to 2009 Posted Jan 7, 2009 11:49 UTC (Wed) by daniels (subscriber, #16193) [Link]
Which mobile graphics chip would you recommend?
|
|
||||||||||||