![]() |
2010-06-29
, 12:51
|
|
Posts: 122 |
Thanked: 135 times |
Joined on Dec 2009
@ Helsinki
|
#282
|
The .pro file is in the same directory as the Makefile, and since qmake generated the Makefile from the .pro file, it should get the path right on its own.
Essentially, it's prepending a huge roundabout directory path in front of files inside the project, completely unneccessarily. If I manually remove the lengthy ../../../Users/talvala/Repositories/tmp/qthello/ section from all paths in the Makefile, it works fine. But that gets tedious very quickly, and qmake shouldn't be doing such a thing in any case - why doesn't it just encode the local relative paths, instead of what looks like an attempt to round-trip through 5 layers of directories?
![]() |
2010-06-29
, 15:42
|
|
Posts: 122 |
Thanked: 135 times |
Joined on Dec 2009
@ Helsinki
|
#283
|
![]() |
2010-10-27
, 13:27
|
|
Posts: 59 |
Thanked: 42 times |
Joined on Nov 2009
@ Wageningen (NL)
|
#284
|
![]() |
2010-10-27
, 16:42
|
Posts: 1,048 |
Thanked: 979 times |
Joined on Mar 2008
@ SF Bay Area
|
#285
|
The Following User Says Thank You to uvatbc For This Useful Post: | ||
![]() |
2010-10-31
, 18:57
|
Posts: 393 |
Thanked: 67 times |
Joined on Feb 2010
|
#286
|
![]() |
2010-11-01
, 14:20
|
Posts: 393 |
Thanked: 67 times |
Joined on Feb 2010
|
#287
|
Hello,
I am still on PR 1.2 and I updated to the latest Mad Developer version 2.0 when it came up, it looks like that was a mistake. Now whenever I run Mad Developer I get the following error:
Code:
./mad-developer: symbol lookup error: ./mad-developer: undefined symbol: _ZN9QListData11detach_growEPii
When I try to get the versiion prior to 2.0 from http://maemo.org/packages/package_in...eveloper/1.94/ I cannot download it due to a server error. The Status field on that site says "Old version cleaned by repository management." But it is the old version that I need.
Can anyone please tell me where I can get the previous version of Mad Developer which worked perfectly in PR 1.2, such as any version prior to 2.0? Perhaps someone has the binary saved on their computer?
Thank you
The Following User Says Thank You to mail_e36 For This Useful Post: | ||
![]() |
2010-11-15
, 07:24
|
Posts: 2,006 |
Thanked: 3,351 times |
Joined on Jun 2010
@ N900: Battery low. N950: torx 4 re-used once and fine; SIM port torn apart
|
#288
|
![]() |
2010-11-18
, 21:16
|
Posts: 3,319 |
Thanked: 5,610 times |
Joined on Aug 2008
@ Finland
|
#289
|
The Following User Says Thank You to attila77 For This Useful Post: | ||
![]() |
2010-11-18
, 22:13
|
Posts: 2,006 |
Thanked: 3,351 times |
Joined on Jun 2010
@ N900: Battery low. N950: torx 4 re-used once and fine; SIM port torn apart
|
#290
|
The updating mechanism is platform agnostic, so you should use the updater plugin, just as anybody else to get the newest MADDE. IIRC there is a bug though that might prevent you from a proper update in older NQS versoins - if you do an update and seemingly nothing is offered/happens, just download&install the latest NQS over the old one.
it's been some time since I have discovered this issue and have been fiddling with the phone + notebook a bit. The original issue is described here:
http://talk.maemo.org/showpost.php?p...&postcount=214
The results of my experiments are following:
1. Yes, file storage mode and pc suite mode works both well, I have not noticed single trouble.
2. When I connect after loading g_ether, it takes a second longer, but the device crashes the same way.
3. I have tried other USB ports on the same machine - same restarts always.
4. I have tried different computer also with windows and it works, loading g_ether doesn't trigger restart. So there is probably nothing wrong with my N900.
5. I managed to corrupt the "Linux USB Ethernet Gadget/RNDIS" device created on my computer the way that windows could not load/enable the device properly when I connect the phone with g_ether loaded. In that case the device appears in the device list marked with yellow exclamation mark saying that the registry settings are corrupted. When done this N900 in half of the occasions does not restart however message pops-up on N900 side saying that the charging is not possible due to insufficient power level and in the other half of occasions the device restarts.
Update: I managed to connect via externally powered USB hub. So there must be something wrong with power levels in my computer's USB ports. It somehow interferes with the charging over USB and causes N900 to restart.
Update 2: OK, so the powered USB hub was just a coincidence. I finally figured out what is causing the trouble. Not a HW / USB ports problem .. it is my VPN software (AT&T network client) what is causing this. Whenever I am connected with this VPN software and try to connect N900 via USB to the computer, when I load g_ether it causes N900 to restart. So obviously there must be some kind of bug in g_ether which triggers the restart. When I load g_ether Windows is via plug&play enabling the RNDIS gadget device, but there is this virtual network adapter (AGN virtual network adapter) used by the VPN SW and if it is enabled at that time something goes nasty and probably the sequence of USB commands is not as expected by g_ether .. that's what I am thinking. The virtual adapter is some kind of a filter network device which sits in between the network adapter and the windows TCP/IP stack. When I am connected with g_ether and then start the VPN SW, this virtual adapter is becoming enabled and this causes immediate restart of N900.
I'm posting here since the original questions went here though it is not MADDE / MAD Developer related - that is obvious now. Any hints / suggestions to help understand the root cause are welcome :-)
Thanks for your hints!
Last edited by flyskyhigh; 2010-06-30 at 08:27.