On 12/18/14 16:28, Adriano dos Santos Fernandes wrote:
> On 18/12/2014 11:16, Alex Peshkoff wrote:
>> On 11/18/14 14:40, Adriano dos Santos Fernandes wrote:
>>
>> .
>>
>>> As an API writer of a project that doesn't follow standards (talking
>>> specially about std exceptions), you need to creat
On 18/12/2014 11:32, Dimitry Sibiryakov wrote:
> 18.12.2014 14:28, Adriano dos Santos Fernandes wrote:
>> So like selecting the types with "using" (or explicitly), user would
>> select the functions too.
>Funny. What will be returned from the function found by dlsym()?
>
Each different IMaster
18.12.2014 14:28, Adriano dos Santos Fernandes wrote:
> So like selecting the types with "using" (or explicitly), user would
> select the functions too.
Funny. What will be returned from the function found by dlsym()?
--
WBR, SD.
---
On 18/12/2014 11:16, Alex Peshkoff wrote:
> On 11/18/14 14:40, Adriano dos Santos Fernandes wrote:
>
> .
>
>> As an API writer of a project that doesn't follow standards (talking
>> specially about std exceptions), you need to create an API which is
>> usable by different ways, that can be mixe
On 11/18/14 14:40, Adriano dos Santos Fernandes wrote:
.
> As an API writer of a project that doesn't follow standards (talking
> specially about std exceptions), you need to create an API which is
> usable by different ways, that can be mixed (each project uses in its
> way). Current code mi
On 11/18/14 18:10, Adriano dos Santos Fernandes wrote:
> It seems I'm able to transform a policy class template into a namespace
> with a policy applied, without C++11 features, which can then be "used"
> and accessed without prefix.
>
> That's with cloop test.
>
> Firebird code is more complex. I'
It seems I'm able to transform a policy class template into a namespace
with a policy applied, without C++11 features, which can then be "used"
and accessed without prefix.
That's with cloop test.
Firebird code is more complex. I'll try the same approach on it.
Adriano
---
18.11.2014 13:18, Adriano dos Santos Fernandes wrote:
> You just don't understand. You understand nothing. It's difficult in
> this way!
But there is no alternative.
> Read the code! Read the archives!
I told you many times that your code is completely unreadable. Didn't you
listen?..
--
On 18/11/2014 09:56, Dimitry Sibiryakov wrote:
> 18.11.2014 12:40, Adriano dos Santos Fernandes wrote:
>> When you have an IAttachment in DtcStart, or IStatus in any other (UDR)
>> class, you're doing it wrong. IAttachment is bound to a policy.
>So, all you need to make it right is to unbind IA
18.11.2014 12:40, Adriano dos Santos Fernandes wrote:
> When you have an IAttachment in DtcStart, or IStatus in any other (UDR)
> class, you're doing it wrong. IAttachment is bound to a policy.
So, all you need to make it right is to unbind IAttachment from a policy,
no?..
> As an API writer
On 18/11/2014 08:54, Alex Peshkoff wrote:
> On 11/18/14 03:50, Adriano dos Santos Fernandes wrote:
>> Hi!
>>
>> Currently the API is problematic in any non usage scenario different
>> than one used internally in FB code.
>>
>> We have the "I" classes witch are typedefs for a "default" policy, a
>>
On 11/18/14 03:50, Adriano dos Santos Fernandes wrote:
> Hi!
>
> Currently the API is problematic in any non usage scenario different
> than one used internally in FB code.
>
> We have the "I" classes witch are typedefs for a "default" policy, a
> policy that is just not suitable for C++, as it ign
I'm really curious what level of absurd you can reach.
PS: KISS.
--
WBR, SD.
--
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashb
Hi!
Currently the API is problematic in any non usage scenario different
than one used internally in FB code.
We have the "I" classes witch are typedefs for a "default" policy, a
policy that is just not suitable for C++, as it ignore exceptions and
avoids every piece of the policy idea.
Even if
14 matches
Mail list logo
