[gmx-users] making maxwarn a hidden option

[Justin A. Lemkul]

ms wrote:
> On 29/12/10 23:47, Justin A. Lemkul wrote:
> 
>> I think the root problem boils down to a lack of documentation of this
>> feature. For most routine use, -maxwarn should not be used, similar to
>> -missing with pdb2gmx.
> 
> Yes, but it depends. In my systems I routinely have to use both to get 
> the system right, because it's a custom coarse grain and both programs 
> spit warnings due to quirks of my (surely non-usual) system -that I am 
> aware of and (as far as I know and I tested) are safe to ignore.
> 

It sounds very much like your systems are in the minority - those for which 
-maxwarn is essential :)

>> It is your last safeguard when fatal errors
>> occur, but if it is implied that using it is somehow routine or
>> convenient, then we begin to undermine the use of all those informative
>> notes and warnings that grompp prints.
> 
> Well, no: grompp should instead print *more* stuff, and explain more in 
> detail what the warning refers to, showing (for example) the line of the 
> .mdp or .top file it refers to and stuff like that, and pointing to an 
> exhaustive manual section to understand it.
> 
> This doesn't undermine: quite the opposite, it empowers the user with 
> *knowledge* that then will use to *decide* what to do.
> 

What I meant was that -maxwarn allows a user to casually bypass that which 
grompp is already printing, something that has caused problems at least 2 or 3 
times now in the last week, I believe.

I know of at least one commonly-used tutorial supplies example grompp commands 
with -maxwarn and no explanation as to what it's doing or why it's being 
invoked.  What we need to do is avoid this blind practice.  I think grompp has a 
sufficient level of verbosity for most cases.  In most (if not all) instances, 
the lines in the .mdp and/or .top are clearly indicated.

<snip>

>> The description of this flag is currently "Number of allowed warnings
>> during input processing." Perhaps it should say something like "Number
>> of allowed warnings during input processing. Not for normal use and may
>> generate unstable systems." I would also think that a description should
>> be added to the grompp -h text, like:
>>
>> "The -maxwarn option can be used to override warnings printed by grompp
>> that otherwise halt output. In some cases, warnings are harmless, but
>> usually are not. The user is advised to carefully interpret the output
>> messages before attempting to bypass them with this option."
>>
>> If that sounds agreeable, I'll put in an enhancement request.
> 
> This makes a lot of sense and I agree in full.
> 

Great, thanks.

-Justin

> Thanks,
> Massimo
> 

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================