[Rivet] AIDA branch closed (gzip next...)

[James Monk]

hmm, well actually there’s an argument that HepMC files should be gzipped as a matter of course - it increases performance, decreases disk usage, and there’s no reason they shouldn’t be.  In that more optimal scenario, it’s a bit weird that Rivet would need files piping through gunzip in order to read them, in just the same way that asking a user to set LD_PRELOAD would be a slightly more convoluted workflow than they’d anticipate.  

cheers,

James

On 22 May 2014, at 12:16, Andy Buckley <andy.buckley at cern.ch> wrote:

> Oh, absolutely: I was just mentioning that it exists! I'm not sure we
> want to make gzipped HepMC files a "normal" workflow either...
> 
> Andy
> 
> 
> On 22/05/14 11:57, David Grellscheid wrote:
>> Please let's not make LD_PRELOAD a normal workflow. It's meant for
>> developers who know what they're doing.
>> 
>>  David
>> 
>> 
>> 
>> On 22/05/2014 18:49, Andy Buckley wrote:
>>> Hi James,
>>> 
>>> You can also do this "non-invasively" via zlibc and LD_PRELOAD, which
>>> avoids the Boost wrangling... but if you like it and it's disabled by
>>> default then please merge your gzip branch on to the default branch and
>>> we'll put this feature out with v2.2. I will direct user complaints to
>>> you :-P
>>> 
>>> Andy
> 
> 
> -- 
> Dr Andy Buckley, Royal Society University Research Fellow
> Particle Physics Expt Group, University of Glasgow / PH Dept, CERN