Re: [eclipse-clp-users] 'Problem is infeasible!' in eplex goes to stdout

From: Kish Shen <kisshen_at_cisco.com>
Date: Wed, 22 Oct 2014 18:06:17 +0100
Hi Sergii,

I am afraid there will be further delays to the updated CBC. 
Unfortunately the new CBC is failing one of our tests during the build 
and test process, so I am reverting back to using the old CBC until this 
new bug is resolved.

It looks like a bug in CBC at the moment, once I confirm this is the 
case, I will need to report it to the CBC people and wait for their fix, 
so I am not sure how long this may take...

Cheers,

Kish

On 21/10/2014 21:27, Kish Shen wrote:
> Hi Sergii,
>
> I am actually in the middle of building the new Cbc binaries, which I
> have verified does resolve your issue.
>
> Sorry this is taking so long. It took us a while to get our daily build
> system working again, and then I was away earlier this month.
>
> Hopefully the Cbc binaries will be built and installed later today or
> tomorrow, and then our next ECLiPSe build will use these new Cbc binaries.
>
> Cheers,
>
> Kish
>
> On 21/10/2014 05:07, Sergii Dymchenko wrote:
>> Hi,
>>
>> I've download today's version 6.1 #193 for Linux 64 bits and it looks like
>> the issue is still present.
>> Any estimates when you will be able to update the binary distribution with
>> the updated CBC version?
>>
>> Sergii.
>>
>>
>>
>> On Thu, Sep 11, 2014 at 3:08 PM, Sergii Dymchenko <kit1980_at_gmail.com> wrote:
>>
>>> Hi Kish,
>>>
>>> Thanks for letting me know. I hope the migration will come out well.
>>>
>>> I understand that CBC/CLP solvers are third-party solvers and Eplex can
>>> also interface to other MP solvers.
>>> But in reality if a person new to ECLiPSe wants to try some examples he
>>> will just download ECLiPSe binaries and it's very unlikely that he will buy
>>> other solvers (I believe free GLPK is very slow for integer programming
>>> even compared with CBC), install, configure them, etc.
>>>
>>> Sergii.
>>>
>>>
>>>
>>> On Thu, Sep 11, 2014 at 2:55 PM, Kish Shen <kisshen_at_cisco.com> wrote:
>>>
>>>> On 11/09/2014 06:07, Sergii Dymchenko wrote:
>>>>
>>>>> Hi,
>>>>>
>>>>> Do you have any estimate when you will be able to update the
>>>>> Cbc we use by compiling it with the macro to suppress the printing of
>>>>> these messages?
>>>>>
>>>>> I'd want to publish an article that have some examples using eplex, and
>>>>> I'd
>>>>> like the examples to work with the latest stable ECLiPSe version if
>>>>> someone
>>>>> wants to test.
>>>>>
>>>>> Sergii.
>>>>>
>>>>
>>>> Hi Sergii,
>>>>
>>>> Unfortunately we had a sudden and unexpected change to the machines we
>>>> build ECLiPSe on, and we are in the process of migrating to new machines,
>>>> but there are some difficulties, and it is unlikely we will be able to
>>>> produce new distributions in the next few days.
>>>>
>>>> John Forrest (the main implementer of the CBC and CLP solvers) have
>>>> kindly updated CBC with the changes so that those messages are no longer
>>>> sent directly to stdout. I had wanted to update the CBC in our binary
>>>> ECLiPSe distributions, but this will need to wait till we restart our daily
>>>> builds.
>>>>
>>>> The CBC/CLP solvers are third-party solvers and not part of ECLiPSe;
>>>> these solvers are distributed with our binary distributions for convenience
>>>> and because their license allow us to do so. Eplex can also interface to
>>>> other MP solvers, and your examples should run regardless of the actual MP
>>>> solver used.
>>>>
>>>> Cheers,
>>>>
>>>> Kish
>>>>
>>>>
>>>>
>>>>
>>>
>>
>
>
> ------------------------------------------------------------------------------
> Comprehensive Server Monitoring with Site24x7.
> Monitor 10 servers for $9/Month.
> Get alerted through email, SMS, voice calls or mobile push notifications.
> Take corrective actions from your mobile device.
> http://p.sf.net/sfu/Zoho
> _______________________________________________
> ECLiPSe-CLP-Users mailing list
> ECLiPSe-CLP-Users_at_lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/eclipse-clp-users
>
Received on Wed Oct 22 2014 - 17:06:47 CEST

This archive was generated by hypermail 2.2.0 : Thu Dec 04 2014 - 01:16:36 CET