|
Administrator H/PC Oracle Posts: | 17,986 |
Location: | United Kingdom | Status: | |
| Windows CE .net changed the face of the Handheld PC community, allowing what we consider the Handheld PC to continue to survive long after the demise of the Handheld PC as a platform.
With its modular roots, CE 4 has allowed developers to generate a new breed of innovative devices, far sooner than developers of platform based counterparts could ever have dreamed of. The SmartBook G138 from Book Digital Co. is one such device which on paper offers all the features you would expect from a Handheld PC, with full telephone support and a great specification.
We take you back to 2003 with a device that is still being manufactured in our latest Flashback review here on HPC:Factor, as we take the SmartBook for a test drive.
Read the review and have your say right here on the forums!
Review: Book Digital Co. SmartBook G138 |
|
| |
|
H/PC Sensei Posts: | 1,330 |
Location: | North of England | Status: | |
| Nice review Chris - check your email  . I'm really looking forward to this arriving now! I'll try to benchmark it against the Sig3 - I suspect the Sig3 may be better from the processing side of things, but the video question will be interesting given that the Sig uses an ATI chipset and the G138 is using the Hitachi chipset. I'll also benchmark the battery.
I'm particularly interested to see/know how they've implemented telephony as if it's using the standard WinCE telephony extension API then that really does open up some nice possibilities.
As mentioned in the email, what is provided in the way of contact management, etc? The Sig3 and Nexio vary widely in this respect, so it'd be nice to see what the G138 does. |
|
| |
|
| When you mention USB keyboards does this mean you suspect that the USB socket also has partial host support? |
|
| |
|
H/PC Oracle Posts: | 16,175 |
Location: | Budapest, Hungary | Status: | |
| no, chazco, it is not a host port, if the registry contains that stuff it doesn't mean the developers of the system didn't just leave it there accidentally.
chiark, it won't be as good as the sig3, sig3 can do 2.5 mbps, i highly doubt the smartbook can. but i'm sure it is on par with most new pda devices on the market except those high end ppc's that have very good video support in tcpmp, because the smartbook's chip won't be supported  |
|
| |
|
H/PC Sensei Posts: | 1,330 |
Location: | North of England | Status: | |
| Agreed cmonex...
it might be possible to get the hitachi chipset supported in tcpmp - after all, they hacked at a driver for the Nexio to get it working acceptably fast... But given that the Sig3 is the best performing 800x480 handheld I've seen by a long way I doubt that the smartbook'll get anywhere near it. If it's usably fast, that'd be great with that screen 
You never know - there may be USB host facilities in there... The Nexio makes do with one port as both the host and client port, and relies on a cable to be a host. I doubt it's there, but all the same it'll be interesting to try it. |
|
| |
|
| Quote You never know - there may be USB host facilities in there... The Nexio makes do with one port as both the host and client port, and relies on a cable to be a host. I doubt it's there, but all the same it'll be interesting to try it.
Well, i've got an adapter cable that would fit if its a simple matter of getting the right socket, but its probably not that simple If it is though... |
|
| |
|
Administrator H/PC Oracle Posts: | 17,986 |
Location: | United Kingdom | Status: | |
| I would be (and was) wary over concluding one way or the other over the USB host support functions. It is by far possible that the one port can support both modes, and there is more than one driver instance in the registry which alludes to USB conncetivity. Why I didn't conclude one way or the other was simply because I didn't have the cable combinations necessary to test it.
GAPI certainly supports the hitachi, in fact I was getting black screens on video in tcpmp without it - what Chiark is saying re native support. The device gave good results at stretched to fill resolutions, and very good results on 100% size playback on 320x240 source files. It does have the 2MB of 2d accelrated vram which helps.
Chiark, in your experience, how flexible do you think CoreCodec would be with regard to helping bolster performance? |
|
| |
|
H/PC Sensei Posts: | 1,330 |
Location: | North of England | Status: | |
| It really depends if there's a demand, but moreover whether one of the developers has enough interest in the machine to roll up their sleeves. I had a look at the source for TCPMP about a year and a bit ago, and it's a fairly big, complex application. I did find the parts which are customised per chipset and whilst there's not a huge amount of routines to write in order to create a driver, it would be necessary to know the full details of the chipset - I'll google to see if Hitachi have published them.
if WinCESoft.de have created a GAPI for the machine, it would suggest that the data is available. I wonder if Frank might share it?
Edited by chiark 2006-08-22 8:05 AM
|
|
| |
|
H/PC Sensei Posts: | 1,298 |
Location: | Florida | Status: | |
| Chris
Once again an excellent review.... of course, now that you've set the standard, we've come to expect nothing less from you. Always a pleasure to read your reviews.
BTW - one little glitch....the forum discussion link at the bottom of the review took me to the NuShield Screen Protectors Review forum thread -opps! |
|
| |
|
H/PC Sensei Posts: | 1,330 |
Location: | North of England | Status: | |
| Very excited now - it uses the same controller as the Simpad... Which means that the opensimpad project already has the necessary documentation to get things up and running for Linux.
Hmm! |
|
| |
|
| If anyone has already got one of these can you test the USB port with something simple, like a thumbstick or keyboard? |
|
| |
|
Administrator H/PC Oracle Posts: | 17,986 |
Location: | United Kingdom | Status: | |
| I know, I know  . It was an intentional oversight due to time constraints. Publishing these things is not a 30 second job unfortunately. I can't fix it until I get back to my office, that is what the plan was when I made the decision.
Thanks for everyones observations in that regard. (Particularly Chiark who was so aghast he phoned me  ) Seriously though one and all, I do appreciate it.
Chiark, you are in luck here I think. I believe that I have the unabridged specification for the chipset at work (chances are you already found the same documentation ). With what you said above, perhaps we have the information between us to carpe diem.
I need to export out the entire registry and have a look through it to see if I can map out the driver setup. |
|
| |
|
H/PC Sensei Posts: | 1,330 |
Location: | North of England | Status: | |
| I've got the same documentation (I guess ) from the opensimpad website. Page 146 details the blitter function from memory to screen which, realistically, is all that is needed. They might not even want to do that, as they may prefer to write directly to the frame buffer using the CPU, but if it's possible to blit from memory to screen then that'd be faster.
Will drop a mail to the controller of the source and see what he says  |
|
| |
|
Global Moderator H/PC Oracle Posts: | 12,669 |
Location: | Southern California | Status: | |
| Nice device. (Price is not bad either. ) |
|
| |
|
H/PC Oracle Posts: | 16,175 |
Location: | Budapest, Hungary | Status: | |
| Quote C:Amie - 2006-08-22 1:43 PM
I would be (and was) wary over concluding one way or the other over the USB host support functions. It is by far possible that the one port can support both modes, and there is more than one driver instance in the registry which alludes to USB conncetivity. Why I didn't conclude one way or the other was simply because I didn't have the cable combinations necessary to test it.
GAPI certainly supports the hitachi, in fact I was getting black screens on video in tcpmp without it - what Chiark is saying re native support. The device gave good results at stretched to fill resolutions, and very good results on 100% size playback on 320x240 source files. It does have the 2MB of 2d accelrated vram which helps.
Chiark, in your experience, how flexible do you think CoreCodec would be with regard to helping bolster performance?
picard once said they'd not support the mediaq... maybe he can be convinced?
gapi: naaah it can do video in GDI without gapi in my experience.
nice smartbooks i have here by the way but you've already found that out from my last line
320x240, i wouldn't watch that on such a wonderful 800x480 screen, especially as it can do 640x480 just about OK... about same performance in GDI as on the Qtek 9000. GAPI beats it at about 900 kbps (990 without sound ).
hmm, about the usb, i'd be surprised if it was a host capable port with the manufacturer never stating it in the specs. but who knows.. |
|
| |
