Join Now

Hi, I'm Paul, but you can also call me Todd and I won't get upset.

At the moment I'm willing to try anything to get ODD to work

Paul.Todd | 10 July, 2008 17:41

Having spent the morning trying to persuade Carbide that I do not have an N73 at the end of the USB cable despite what windows and my self know I found this reddit post quite amusing

One major feature request for ODD - can we have meaningful error messages: what does error code 20 mean? 

The app installs and runs fine on the device and installs one via ODD but the moment it starts the process, ODD reports an error.

So my once monthly saga into trying to get ODD working on my Dell continues.....

 (More)

General | Permalink | Comments (3) | Trackbacks (0)

Symbian signed and openess

Paul.Todd | 08 July, 2008 12:11

I hav'nt posted for a while, primarily becuase I am behind in whitepaper writing and partially due to work commitments.

Anyway I am working on a new set of posts explaining the OMA-DM adapters and how to access them so this will be starting in a few days as I have realized the existing documentation for this is woefully inadequate for something so powerful.

However, onto the cheese as people say... David Woods has an excellent article on Symbian Signed over at his blog

 However there  are some fundemental issues I disgree with.

1. As he was part of the Symbian Signed  design team he knows it inside and out. Most developers do not know or understand the nuances of capabilities and how they apply to which dll is loaded with what capabilities and it even goes as far as signing a sis file. Even the documentation is far from clear. One only needs to look a the rule for dynmically loading a dll vs statically loading a dll. Whilst David may know it as he designed Symbian, a Windows or j2me developer will not.
In the past I have had  more than my fair share of problems explaining to other dev's why it is a bad idea to run services with the TCB priviledge on Windows and its orders of magnitude more complex to explain capabilities on Symbian.

2. how the hell can someone rewrite an OS and forget to document which API's have have what capabilities. Enough said really (whilst ROFL)

2.1 How can someone rewrite an OS and forget to put in decent logging and helpful error messages when an application fails to run or install (still ROFL'ing)

3. Getting decerts for other competing products is a piece of cake compared to Symbian. MS, RIM and Apple are simple and it just works.(Not ROFL anymore)

4. If anything I see device manufacturers using capabiltiies to prevent people distributing applications that can compete with other solutions or preferred suppliers. Lets say I write an antivirus application with the new Symbian Foundation API's, I am still beholden to the device manufacturer to grant the capabiltities to install and run my product.No where does it say Symbian Signed will be the point at which an application will be rejected for quality concerns. The way I see this playing out is that manufacturer will still stop developers getting devcerts with all capabilities (eg TCB) under the guise that they only work with "partners" and not with ISV's so developing for example an antivirus solution will never happen because of manufacturer lockout. Whereas Symbian Signed should be the gatekeeper for applications to be run on the phone it is much more likely that capabilties will still be the paper work nightmare they currently are.

For these issues, I unfortunatly I still see Symbian Signed as a revenue black hole for ISV's in the moribund ISV app market.

So enough with the "open source" washing PR and more with concrete support for developers because ultimately we are the people going to deliver the applications that sell the phones and lets make sure they are Symbian and not Android or Limo phones.

 If you really want to impress me, say that the Symbian Foundation and Nokia are adopting the principals and practises of Maemo and not Microsoft.

General, Symbian C++ | Permalink | Comments (4) | Trackbacks (0)

 
 
Powered by LifeType 
     
     RDF Facets:
     
     
     qfnZtypeQUqfnTypeZBlogContentQ
     qfnZtypeQUqfnTypeZCommunityContentQ
     qfnZtypeQUqfnTypeZWebpageQ
     qmarsZlanguageQUxhttpE3aE2fE2fswE2enokiaE2ecomE2flanguageE2d1E2fenX