[fdo-internals] resignation from FDO PSC

Haris Kurtagic haris at sl-king.com
Tue Jul 20 05:01:59 EDT 2010


Hi Traian,

I understand what you are saying and that is certainly most common
error I saw in code using FDO.
One of my first "not-like" in fdo was recounting everything. I would
prefer reference and pointers and const pointers approach.
As a helping thought, I am think of FDO commands as RISC against CISC
:) , risc has it's advantages

BTW, I would like to have C interface to FDO and even could be that I
will put some time in it .

Haris

On Tue, Jul 20, 2010 at 2:00 AM, Traian Stanev
<traian.stanev at autodesk.com> wrote:
>
> Hi Haris,
>
> Personally I don't have a problem with it either. However, imagine yourself in a situation where you work with, say, 10 programmers with Java or .Net background, doing natural, obvious things like:
>
> class->GetProperties()->GetCount();
>
> Or
>
> identifiers->GetItem(i)->GetName();
>
> Then you have to spend time fixing it and educating people, which hurts your own productivity. It's a situation that happens often.
>
> Using a subset of the API will not fix this, since it's still the same API. So in order to keep people productive, at runtime we end up either wrapping things in something that is less error prone, or copying data to runtime structures that are easy to use. If there is no wrapper, you end up with all the disadvantages I mentioned and you didn't refute directly.
>
>
> Traian
>
>
>
> -----Original Message-----
> From: fdo-internals-bounces at lists.osgeo.org [mailto:fdo-internals-bounces at lists.osgeo.org] On Behalf Of Haris Kurtagic
> Sent: Monday, July 19, 2010 4:20 AM
> To: FDO Internals Mail List
> Subject: Re: [fdo-internals] resignation from FDO PSC
>
> Hi Traian,
>
> It is unfortunate that you have lost confidence in the API.  I
> personally feel that FDO has great future. I never had major problems
> with FDO, it were problems with providers or fdo clients.
> The most often complain I heard is that it is big interface and lot to
> learn before use. Probably it is true, but there is solution for it.
> We can always create subset of interface and make it much easier to
> use for most popular use cases.
>
> Haris
>
> On Sat, Jul 17, 2010 at 11:58 PM, Traian Stanev
> <traian.stanev at autodesk.com> wrote:
>>
>> Hi Haris,
>>
>> The problem is not RFC minutia that can be discussed away in a heated mailing list thread. It's more strategic. Here are some details which perhaps will make it more clear.
>>
>> Currently, there are two conflicting uses of FDO. One is as data import/export API, and the other is as runtime API on which to build vertical functionality (like rendering, analysis, etc). But:
>>
>> * It is not a great import/export API, because it is not nearly as comprehensive as something like FME for example. Even GDAL/OGR supports more data sources than FDO.
>>
>> * It is a poor runtime API, because:
>>    - it is extremely verbose -- it is not uncommon to reduce code size by a factor of 3 or 4 by switching to a well designed wrapper.
>>    - it is not designed for ultimate performance (virtual refcounting, thread safe refcounting, things that aren't exactly cheap)
>>    - it is error prone (very easy to accidentally leak memory)
>>    - different providers all have different capabilities, making it impossible to write coherent logic unless one makes assumptions about the source provider or capabilities.
>>
>> I don't think it's fundamentally possible to have an API which is optimal for both uses at the same time, due to the wildly different requirements. My resignation simply reflected this lack of confidence in the FDO API.
>>
>> Traian
>>
>>
>>
>>
>> -----Original Message-----
>> From: fdo-internals-bounces at lists.osgeo.org [mailto:fdo-internals-bounces at lists.osgeo.org] On Behalf Of Haris Kurtagic
>> Sent: Saturday, July 17, 2010 3:40 AM
>> To: FDO Internals Mail List
>> Subject: Re: [fdo-internals] resignation from FDO PSC
>>
>> Hi Traian,
>>
>> I am sad that you have come to this point, you are great asset to our project.
>>
>> I have been looking in public discussion on FDO lists and couldn't see
>> any "hot" debate about any issue. The most discussed issue I saw was
>> Romica and myself about join RFC.
>>
>> It brings us to point which was said few times already that we would
>> need to discuss more on public lists no meter how much we agree or
>> not.
>>
>> It is never too late for open debate and I would always love to hear
>> what you have to say.
>>
>> Haris
>>
>> On Fri, Jul 16, 2010 at 5:12 PM, Traian Stanev
>> <traian.stanev at autodesk.com> wrote:
>>> Hello,
>>>
>>> I am resigning from my FDO PSC membership position. Lately, I have had trouble reconciling my personal beliefs about changes to FDO versus the direction supported by my employer. Therefore I think it's best to remove the conflict of interest by giving up my vote. I would like to emphasize that this is entirely my own decision and I was not pressured into it.
>>>
>>> Thanks,
>>> Traian
>>> _______________________________________________
>>> fdo-internals mailing list
>>> fdo-internals at lists.osgeo.org
>>> http://lists.osgeo.org/mailman/listinfo/fdo-internals
>>>
>> _______________________________________________
>> fdo-internals mailing list
>> fdo-internals at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/fdo-internals
>> _______________________________________________
>> fdo-internals mailing list
>> fdo-internals at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/fdo-internals
>>
> _______________________________________________
> fdo-internals mailing list
> fdo-internals at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/fdo-internals
> _______________________________________________
> fdo-internals mailing list
> fdo-internals at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/fdo-internals
>


More information about the fdo-internals mailing list