You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mahout.apache.org by "Grant Ingersoll (Created) (JIRA)" <ji...@apache.org> on 2011/11/04 01:57:32 UTC
[jira] [Created] (MAHOUT-870) Driver or Job? Let's pick one and be
consistent.
Driver or Job? Let's pick one and be consistent.
-------------------------------------------------
Key: MAHOUT-870
URL: https://issues.apache.org/jira/browse/MAHOUT-870
Project: Mahout
Issue Type: Improvement
Reporter: Grant Ingersoll
Priority: Minor
Some things are Drivers, some are Jobs, but they all do the same thing. Since we often use these as the main interface into our stuff, let's make it easier for people to find things by being consistent w/ our naming.
My vote is Driver.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
RE: [jira] [Created] (MAHOUT-870) Driver or Job? Let's pick one
and be consistent.
Posted by Jeff Eastman <je...@Narus.com>.
+1 For some time my pattern has been to use a Driver for each independent M/R algorithm and a Job to execute more than one Driver to accomplish some higher level of functionality (e.g. seq2sparse). But now, a lot of the drivers are themselves orchestrating atomic drivers. Adding the sequential execution mode has also made the drivers less atomic than I had originally intended.
-----Original Message-----
From: Frank Scholten [mailto:scholten.f.b@gmail.com]
Sent: Friday, November 04, 2011 1:50 AM
To: dev@mahout.apache.org
Subject: Re: [jira] [Created] (MAHOUT-870) Driver or Job? Let's pick one and be consistent.
On Nov 4, 2011, at 2:26, Shannon Quinn <sq...@gatech.edu> wrote:
> My habit has been naming anything immediately accessible via bin/mahout a "driver", anything else a "job".
I favour this naming convention.
>
> I'm also inclined towards "job".
>
> On 11/3/11 9:06 PM, Sebastian Schelter wrote:
>> I'd favor Jobs :)
>>
>> On 04.11.2011 01:57, Grant Ingersoll (Created) (JIRA) wrote:
>>> Driver or Job? Let's pick one and be consistent.
>>> -------------------------------------------------
>>>
>>> Key: MAHOUT-870
>>> URL: https://issues.apache.org/jira/browse/MAHOUT-870
>>> Project: Mahout
>>> Issue Type: Improvement
>>> Reporter: Grant Ingersoll
>>> Priority: Minor
>>>
>>>
>>> Some things are Drivers, some are Jobs, but they all do the same thing. Since we often use these as the main interface into our stuff, let's make it easier for people to find things by being consistent w/ our naming.
>>>
>>> My vote is Driver.
>>>
>>> --
>>> This message is automatically generated by JIRA.
>>> If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
>>> For more information on JIRA, see: http://www.atlassian.com/software/jira
>>>
>>>
>
Re: [jira] [Created] (MAHOUT-870) Driver or Job? Let's pick one and be consistent.
Posted by Frank Scholten <sc...@gmail.com>.
On Nov 4, 2011, at 2:26, Shannon Quinn <sq...@gatech.edu> wrote:
> My habit has been naming anything immediately accessible via bin/mahout a "driver", anything else a "job".
I favour this naming convention.
>
> I'm also inclined towards "job".
>
> On 11/3/11 9:06 PM, Sebastian Schelter wrote:
>> I'd favor Jobs :)
>>
>> On 04.11.2011 01:57, Grant Ingersoll (Created) (JIRA) wrote:
>>> Driver or Job? Let's pick one and be consistent.
>>> -------------------------------------------------
>>>
>>> Key: MAHOUT-870
>>> URL: https://issues.apache.org/jira/browse/MAHOUT-870
>>> Project: Mahout
>>> Issue Type: Improvement
>>> Reporter: Grant Ingersoll
>>> Priority: Minor
>>>
>>>
>>> Some things are Drivers, some are Jobs, but they all do the same thing. Since we often use these as the main interface into our stuff, let's make it easier for people to find things by being consistent w/ our naming.
>>>
>>> My vote is Driver.
>>>
>>> --
>>> This message is automatically generated by JIRA.
>>> If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
>>> For more information on JIRA, see: http://www.atlassian.com/software/jira
>>>
>>>
>
Re: [jira] [Created] (MAHOUT-870) Driver or Job? Let's pick one
and be consistent.
Posted by Shannon Quinn <sq...@gatech.edu>.
> the redundancy of MahoutDriver!
Ok, that is an excellent point and very well may have convinced me.
>
> On Nov 3, 2011, at 9:26 PM, Shannon Quinn wrote:
>
>> My habit has been naming anything immediately accessible via bin/mahout a "driver", anything else a "job".
>>
>> I'm also inclined towards "job".
>>
>> On 11/3/11 9:06 PM, Sebastian Schelter wrote:
>>> I'd favor Jobs :)
>>>
>>> On 04.11.2011 01:57, Grant Ingersoll (Created) (JIRA) wrote:
>>>> Driver or Job? Let's pick one and be consistent.
>>>> -------------------------------------------------
>>>>
>>>> Key: MAHOUT-870
>>>> URL: https://issues.apache.org/jira/browse/MAHOUT-870
>>>> Project: Mahout
>>>> Issue Type: Improvement
>>>> Reporter: Grant Ingersoll
>>>> Priority: Minor
>>>>
>>>>
>>>> Some things are Drivers, some are Jobs, but they all do the same thing. Since we often use these as the main interface into our stuff, let's make it easier for people to find things by being consistent w/ our naming.
>>>>
>>>> My vote is Driver.
>>>>
>>>> --
>>>> This message is automatically generated by JIRA.
>>>> If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
>>>> For more information on JIRA, see: http://www.atlassian.com/software/jira
>>>>
>>>>
> --------------------------------------------
> Grant Ingersoll
> http://www.lucidimagination.com
>
>
>
>
Re: [jira] [Created] (MAHOUT-870) Driver or Job? Let's pick one and be consistent.
Posted by Grant Ingersoll <gs...@apache.org>.
Bikeshedding, I know, but we have MahoutDriver and a whole bunch of other Drivers. Most of clustering is Driver-based, Recommenders seem to be Job-based, classification seems to have a mix (TrainNaiveBayesJob, TestNaiveBayesDriver!)
We just need to have less confusion and make it easier for people to get in and start navigating the code. Consistency in naming makes that easier.
I personally like the association of Driver with Mahout (Driver of Elephants) and the redundancy of MahoutDriver! Jobs are boring things we hate going to! Driving is cool!
On Nov 3, 2011, at 9:26 PM, Shannon Quinn wrote:
> My habit has been naming anything immediately accessible via bin/mahout a "driver", anything else a "job".
>
> I'm also inclined towards "job".
>
> On 11/3/11 9:06 PM, Sebastian Schelter wrote:
>> I'd favor Jobs :)
>>
>> On 04.11.2011 01:57, Grant Ingersoll (Created) (JIRA) wrote:
>>> Driver or Job? Let's pick one and be consistent.
>>> -------------------------------------------------
>>>
>>> Key: MAHOUT-870
>>> URL: https://issues.apache.org/jira/browse/MAHOUT-870
>>> Project: Mahout
>>> Issue Type: Improvement
>>> Reporter: Grant Ingersoll
>>> Priority: Minor
>>>
>>>
>>> Some things are Drivers, some are Jobs, but they all do the same thing. Since we often use these as the main interface into our stuff, let's make it easier for people to find things by being consistent w/ our naming.
>>>
>>> My vote is Driver.
>>>
>>> --
>>> This message is automatically generated by JIRA.
>>> If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
>>> For more information on JIRA, see: http://www.atlassian.com/software/jira
>>>
>>>
>
--------------------------------------------
Grant Ingersoll
http://www.lucidimagination.com
Re: [jira] [Created] (MAHOUT-870) Driver or Job? Let's pick one
and be consistent.
Posted by Shannon Quinn <sq...@gatech.edu>.
My habit has been naming anything immediately accessible via bin/mahout
a "driver", anything else a "job".
I'm also inclined towards "job".
On 11/3/11 9:06 PM, Sebastian Schelter wrote:
> I'd favor Jobs :)
>
> On 04.11.2011 01:57, Grant Ingersoll (Created) (JIRA) wrote:
>> Driver or Job? Let's pick one and be consistent.
>> -------------------------------------------------
>>
>> Key: MAHOUT-870
>> URL: https://issues.apache.org/jira/browse/MAHOUT-870
>> Project: Mahout
>> Issue Type: Improvement
>> Reporter: Grant Ingersoll
>> Priority: Minor
>>
>>
>> Some things are Drivers, some are Jobs, but they all do the same thing. Since we often use these as the main interface into our stuff, let's make it easier for people to find things by being consistent w/ our naming.
>>
>> My vote is Driver.
>>
>> --
>> This message is automatically generated by JIRA.
>> If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
>> For more information on JIRA, see: http://www.atlassian.com/software/jira
>>
>>
Re: [jira] [Created] (MAHOUT-870) Driver or Job? Let's pick one
and be consistent.
Posted by Sebastian Schelter <ss...@apache.org>.
I'd favor Jobs :)
On 04.11.2011 01:57, Grant Ingersoll (Created) (JIRA) wrote:
> Driver or Job? Let's pick one and be consistent.
> -------------------------------------------------
>
> Key: MAHOUT-870
> URL: https://issues.apache.org/jira/browse/MAHOUT-870
> Project: Mahout
> Issue Type: Improvement
> Reporter: Grant Ingersoll
> Priority: Minor
>
>
> Some things are Drivers, some are Jobs, but they all do the same thing. Since we often use these as the main interface into our stuff, let's make it easier for people to find things by being consistent w/ our naming.
>
> My vote is Driver.
>
> --
> This message is automatically generated by JIRA.
> If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
> For more information on JIRA, see: http://www.atlassian.com/software/jira
>
>
[jira] [Resolved] (MAHOUT-870) Driver or Job? Let's pick one and
be consistent.
Posted by "Sean Owen (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAHOUT-870?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Sean Owen resolved MAHOUT-870.
------------------------------
Resolution: Won't Fix
"Job" is the Hadoop terminology. Though Job = one MapReduce. I imagine it's still the right-er convention. But as this is minor, nobody's acted on it, the names are in common use now by third parties, and there are far bigger differences among the implementations, I suggest there's not much point changing this.
> Driver or Job? Let's pick one and be consistent.
> -------------------------------------------------
>
> Key: MAHOUT-870
> URL: https://issues.apache.org/jira/browse/MAHOUT-870
> Project: Mahout
> Issue Type: Improvement
> Reporter: Grant Ingersoll
> Priority: Minor
>
> Some things are Drivers, some are Jobs, but they all do the same thing. Since we often use these as the main interface into our stuff, let's make it easier for people to find things by being consistent w/ our naming.
> My vote is Driver.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira