I am also ok with axis2c prefix ,
Nandika
On 9/28/05, Damitha Kumarage <[EMAIL PROTECTED]> wrote:
Hi,
Samisa Abeysinghe wrote:
> Sanjiva Weerawarana wrote:
>
>> On Wed, 2005-09-28 at 21:44 -0600, Lilantha Darshana wrote:
>>
>>
>>> IMHO, I would suggest prefixing all names with 'axis2c' makes difficult
>>> to read names. It should be easy to read and find names by their
>>> own. Sometime
>>> searching names by enabling "find whole word only" makes difficult on
>>> having these prefixes. Since we know these code are from axis2c we
>>> might
>>> want to drop the prefix, if you do not have a strong reason to use it.
>>>
>>
>>
>> Going back to my old C programming days, it is absolutely critical to
>> prefix all names with a unique constant .. the pain of link errors is
>> simply not worth it.
>>
>> I'd suggest s/axis2c/axis2/g though.
>>
>>
> Yes, the obvious rationale for the prefix was the namespace conflicts,
> again going along with Axis C++ experiance, where at one point we
> really had to go back and introduce namespaces.
>
> I am OK to change axis2c to axis2 :)
Remember we have named our files also with axis2c prefix like in
axis2c_om_container.h.
So I think keeping all names with prefix axis2c is ok to be consistent
with file names as well.
Changing files names is not a good idea because I think what we gain is
not worth the effort(again
checkin to svn repo with different name)
cheers
damitha
>
>>
>>
>>> Further of using underscores ('_') with often used names needing us
>>> typing two keys in key board very often :-).
>>
>>
>> Yeah but again its a reasonably widely used model. The other option is
>> of course almost Java-like CamelCase: Axis2cFoo Axis2cBar etc..
>>
>>
>>
> We used the APR source as out master guide, when it came to selecting
> style.
>
>> Sanjiva.
>>
>>
>>
>>
>>
>
>
