[SciPy-user] OpenOpt svn broken on Windows

dmitrey dmitrey.kroshko at scipy.org
Thu Jul 17 03:55:15 EDT 2008


Matthieu Brucher wrote:
>>> Let me explain this.
>>> You are complaining that you have to connect several optimizers
>>> yourself to OO.
>>>       
>
>   
>> Matthieu, IIRC I had been never complaining that I have to connect
>> several optimizers to OO.
>>     
>
> This is what you were saying some hours ago:
>
> "> And it is not your job to try to get a workaround (IMHO) ;)
>   
> Neither connecting IPOPT nor ALGENCAN 2.x version  (as well as some
> other work done during my GSoC 2008) are not mentioned in my
> GSoC-assigned schedule (still I had spent some more time because some OO
> users were very interested in these issues). So it's not my job to get
> ALGENCAN workaround."
>
> Sorry, but for me, this is complaining.
Maybe for you, but not for me (and I guess for others as well).
>  And I'm not talking about the
> mails you sent some months ago (in which you were saying that you
> wanted to get paid to connect some solvers).
>   
Why don't talk? Give a link to the messages (because I can't remember 
those ones), let's discuss it.
>   
>> Please give me a link to my message where I say "it is not my job to
>> connect solver X or solver Y".
>>     
>
> See my above quote.
>   
I see your quote above and don't find the words.

>   
>> Let me also remember you that OO had been started even before GSOC 2007,
>> and in GSoC 2008 rules (Python dept) they clarified that all code
>> produced by student is his own one (still it should be available with
>> BSD license). If you want to fork openopt to other package and develop
>> that one by yourself - please inform me ASAP.
>>     
>
> Let me also inform you that OO is a scikit to which I contributed a
> lot as well. And to which I will keep contributing (generic framework
> + some other lines).
>   
Let me remember you when I had been insisted to connect GenericOpt to 
OpenOpt I said "GenericOpt should stay available as separate package", 
and I have obtained agree (and AFAIK it remains available as separate 
package). So please enough speculating on GenericOpt. You have 
GenericOpt page in scikits webserver, any time you can remove GenericOpt 
from OpenOpt svn and continue to develop and use it as a standalone 
package (I would gladly hear from you of this decision).

The situation is like I commit something to Linux Kernel and then I 
commit whatever I want to any Linux files saying "I'm Linux developer 
too, I had committed something so I will continue more and more, no 
matter what you think".

D.



More information about the SciPy-User mailing list