Ticket #1832 (assigned defect)

Opened 6 years ago

Last modified 5 years ago

2008.8 cannot dial *123

Reported by: hedora Owned by: john_lee
Priority: normal Milestone: Om2008.10
Component: Qtopia Version: Om2008.8
Severity: normal Keywords:
Cc: tick@… Blocked By:
Blocking: Estimated Completion (week):
HasPatchForReview: no PatchReviewResult:
Reproducible:

Description

That is the number that lets me check my voicemail message. See comments in bug #1766 for more information. I haven't tried other "special" numbers that let me contact my phone company.

*123 works fine in 2007.2 and FSO.

Change History

comment:1 Changed 6 years ago by Nytowl

I've tried a number of different star and pound special numbers. Any number that starts with a pound or star causes the dialer to exit (sigsegv)

comment:2 Changed 6 years ago by regina_kim

  • Milestone set to Om2008.9

comment:3 Changed 6 years ago by regina_kim

  • Owner changed from openmoko-devel to zecke
  • Component changed from unknown to Qtopia

comment:4 Changed 6 years ago by haduong@…

Trevino on the community ML has a fix with explanation, patch and binaries for testing available:
http://lists.openmoko.org/pipermail/community/2008-September/030814.html
http://lists.openmoko.org/pipermail/community/2008-September/030834.html

comment:5 Changed 6 years ago by will

  • Milestone changed from Om2008.9 to Om2008.10

comment:6 Changed 6 years ago by zecke

Attach patches to the bug tracker and set HasPatch?. My GPRS connection does not allow to go fishing for patches.

comment:7 Changed 6 years ago by Treviño

It should be fixed in #2038

comment:8 Changed 6 years ago by tick

  • Owner changed from zecke to tick
  • Status changed from new to accepted

comment:9 Changed 6 years ago by tick

  • Status changed from accepted to in_testing

According to 2032, switch to in_testing

comment:10 Changed 6 years ago by tick

Sorry correct the typo it's #2038

comment:11 Changed 6 years ago by tick

  • Status changed from in_testing to assigned

It's not there yet. reopen it.

comment:12 Changed 5 years ago by tick

  • Status changed from assigned to in_testing
  • HasPatchForReview unset

I can dial *123. Though there are no such service in TW.
Please test it somewhere have these services. :-) thanks

comment:13 Changed 5 years ago by sushama

  • Owner changed from tick to john_lee
  • Status changed from in_testing to assigned

Testing image:02-dec-2008
Test date:03-dec-2008
Used a far eastone sim to dial out *131*1# (this service gives the balance credit left in the sim)
but the neo does not give back a message with the balance credit. Tried the same with a samsung phone and it does give a message with the credit left
the logread is as follows -

Dec 3 09:23:00 om-gta02 user.notice root: AtChat? : T : "AT+CUSD=1,002A003100330031002A00310023,15"
Dec 3 09:23:01 om-gta02 user.notice root: AtChat? : F : "+CME ERROR: 3"
Dec 3 09:23:08 om-gta02 user.notice root: AtChat? : percentCSQ event, rssi: 31

comment:14 Changed 5 years ago by john_lee

  • Cc tick@… added

tick,

could you help explaining the log then reassign it if necessary? thanks!

comment:15 follow-up: ↓ 16 Changed 5 years ago by Treviño

As I've said in #2038, imho the implementation isn't correct.

The first I did was tricky, but it works since it before sets the codec to GSM to place the request using the command AT+CUSD="<string-number-coded-using-current-codec>",<code-for-codec> with the <string-number-coded-using-current-codec> in plain ascii text and the <code-for-codec> set to the "known" 15 value; then decodes the answer using the current codec.

As I've previously said using the AT+CUSD="<string-number-coded-in-ucs2>",15 couldn't work since "15" is not the code for the ucs2 codec (that according to GSM 03.38 should be 01001000 and so 72 in decimal).

I've not tested the code living in git for this task in my phone since I'm still using my implementation, but I figure that I've already given all that I know about this.

comment:16 in reply to: ↑ 15 Changed 5 years ago by zecke

Replying to Treviño:

As I've said in #2038, imho the implementation isn't correct.

The first I did was tricky, but it works since it before sets the codec to GSM to place the request using the command AT+CUSD="<string-number-coded-using-current-codec>",<code-for-codec> with the <string-number-coded-using-current-codec> in plain ascii text and the <code-for-codec> set to the "known" 15 value; then decodes the answer using the current codec.

As I've previously said using the AT+CUSD="<string-number-coded-in-ucs2>",15 couldn't work since "15" is not the code for the ucs2 codec (that according to GSM 03.38 should be 01001000 and so 72 in decimal).

I've not tested the code living in git for this task in my phone since I'm still using my implementation, but I figure that I've already given all that I know about this.

My issue:

  • People can not test this in Taiwan as supplementary services are not used there.
  • I'm working on something not GTA0X/phone related and didn't have a spare cycle to work on this since leaving Taipei...

The issue with your implementation:

  • Temporarily switching the codec will open a can of worms. It opens a race with phonebook and SMS handling. and as an expert in racing I can assure you, you will not win this.

How to proceed:

  • I think the approach taken by the implementation in git (using CUSD, specifying a codec, touching the set of files) is the right thing to do.
  • The implementation, e.g. the codec value might be wrong...

How to continue:

  • Some one with time and a neo might just issue AT+CUSD and change the 15 to 72 (as from the bug)...
  • Extra points for testing this within a network with supplementary services..
Note: See TracTickets for help on using tickets.