Who am I?
I've been poking my nose in mobile software development since 2000, starting on what was then the ER5 and Ericsson's R380. Since then I've been doing a bit of everything on the S60 Platform, from learning its bits and pieces to helping others find their way trough the maze.
Calendar
Recently...
- Qt for S60 "Tower" demo running on Nokia N97
- Qt for S60 "Tower" released
- Symbian Foundation promo videos
- Happy Birthday Symbian Signed!
- More Symbian Foundation beta invites
- Update your links to SDL
- Facelift at SymbianSigned.com
- Feeling inspired? (II)
- Can you write an antivirus?
- Carbide.ui Theme Edition registration
Categories
- ActivePerl [1]
- Browsing [0]
- Business Opportunities/Services [0]
- Carbide.c++ [7]
- Carbide.ui [1]
- CDMA [0]
- Connectivity [0]
- Enterprise [0]
- Entertainment [0]
- Event [2]
- Flash [0]
- Fun [1]
- Games [0]
- GCCE [2]
- General [9]
- Hack [4]
- Java [1]
- Location Based Services [0]
- Maemo [0]
- Messaging [0]
- Mobile [1]
- Mobile [1]
- N96 [1]
- Open C [1]
- Perl [1]
- Python [0]
- Qt for S60 [3]
- S60 [20]
- SDK [4]
- Series 40 [1]
- Series 80 [1]
- Symbian [6]
- Symbian C++ [14]
- Testing [0]
- UI [1]
- Vista [1]
- Web Run-Time (WRT) [1]
- Widget [0]
- Windows [1]
- Windows 7 [1]
- Windows Vista [0]
Lucian Tomuta's Forum Nokia Blog
The S60 SDK emulator: faster if you do not start it
ltomuta | 29 July, 2008 23:00
Just now I've run across a thread in the Forum Nokia's Discussion Boards promising an 1000% improvement of emulator's startup time.
Two things came to my mind seeing this thread: a paradox attributed to Zeno of Elea and ... the S60 3rd Edition SDK, supporting Feature Pack 2.
First, Zeno's paradox, adapted so that it fits the topic:
In order to satisfy the developers, the emulator would have to be fully up and running in about 10s. But, in order to do so, it would have to be in a "half way there" state first, which it cannot do unless it reaches half of that state to begin with. And so on, ... and so on, ad infinitum.
I have seen at least two conclusions of this paradox, pick your choice:
- The "logical" conclusion is that the emulator cannot actually ever start since it will have to pass through an infinite number of intermediary states in a finite time interval.
- The "logical" conclusion is that in order for the emulator to achieve the desired performance it is mandatory that the emulator is not even started.
Ok? No more complaints about the emulator please! 
Now the serious part:
The S60 3rd Edition SDK FP2 does implement two start modes for the emulator, a partial one, with only a few critical servers started, and the full slow start-up mode with which we are all accustomed by now.
You can switch between the two start-up modes by modifying a setting in emulator's Preferences dialog (through the Tools -> Preferences menu option).
The SDK release v1.1 also adds a graphical indicator to the emulator's window so that the developer knows all the time whether the emulator runs in partial or full startup mode. This is intended as a reminder likely to be useful should you experience some strange application error, which can be a side effect of the chosen startup mode.
For tweaking the emulator of the previous platform releases we are each on our own, but it should be quite safe to try ...
S60, SDK |
Next | 
Previous |
Comments (2) | 
Trackbacks (0)
Comments
ltomuta
ltomuta | 31/07/2008, 10:54
One old joke about the depreciation of the US Dollar said "No matter how low the dollar is, when you see one bend down and grab it"
I have the same approach to the emulator, and I will always use it, however slow it might be :)
Not starting the emulator at all...
mgroeber9110 | 31/07/2008, 10:31
It seems to me that this is a bit the approach that the Carbide.c++ v1.3 release has adopted: the best way to avoid problems with the emulator is to not start it at all, and to use On-Device Debugging instead... :-)