Project

General

Profile

Bug #13654

artmod for analyzer prints "art/Utilities/InputTag.h" instead of "canvas/Utilities/InputTag.h"

Added by Wesley Ketchum about 3 years ago. Updated almost 3 years ago.

Status:
Closed
Priority:
Normal
Category:
Infrastructure
Target version:
Start date:
08/24/2016
Due date:
% Done:

100%

Estimated time:
1.00 h
Spent time:
Occurs In:
Scope:
Internal
Experiment:
MicroBooNE
SSI Package:
cetpkgsupport
Duration:

Description

filter is ok. Haven't tried producer

this is on art v2_03_00


Related issues

Has duplicate art - Bug #13833: artmod still produces #include "art/Utilities/InputTag.h"Rejected09/10/2016

History

#1 Updated by Kyle Knoepfel about 3 years ago

  • Status changed from New to Feedback

artmod has been superceded as of art 1.14.00 in favor of cetskelgen. Please give cetskelgen a try.

#3 Updated by Wesley Ketchum about 3 years ago

OK, this looks ok. I'm guessing it has all the same options that artmod had?

Also, would be good to make sure this is updated across the board. art workbook v0.91, dated July 18 2016, still mentions artmod. Also, can a message display when running artmod to say that this has been depracated? Or ... if we shouldn't use artmod, can it be disabled (or have a message saying "use cetskelgen"?

#4 Updated by Adam Lyon about 3 years ago

A g-2 person got hit by this just this morning. Can artmod be removed or can cetskelgen turn into a new artmod? Right now we have an artmod that people know but does the wrong thing,

#5 Updated by Adam Lyon about 3 years ago

Adam Lyon wrote:
I like Wes' idea - have artmod point people to cetskelgen. Also, note that $CETLIB_FQ_DIR/bin is not an obvious place to look for something like artmod.

#6 Updated by Wesley Ketchum about 3 years ago

(While offering suggestions, artmod is kinda a nice name for what we've been using it for [generating art modules]. I think it's gonna take me a while to remember cetskelgen. And as Adam said, we've been trying to train a generation of students and PDs to think about and use artmod...)

#7 Updated by Rob Kutschke about 3 years ago

As I understand it, artmod creates skeletons for art modules while cetskelgen knows how to make skeletons for many kinds of art plugins.

Does it make sense to do the following:

  1. make artmod a symlink to cetskelgen
  2. modify cetskelgen to check argv\[0] to check if the executable was invoked as cetskelgen or as artmod
  3. If it was invoked as artmod, change default behaviour to be as close to artmod as possible.

If this gets us close enough to the current behaviour of artmod, then regular artmod users will see this as if it were a minor evolution of artmod.

#8 Updated by Kyle Knoepfel about 3 years ago

We will discuss this at this week's stakeholders meeting.

#9 Updated by Marc Paterno about 3 years ago

  • Status changed from Feedback to Assigned
  • Assignee set to Christopher Green
  • Estimated time set to 1.00 h
  • Experiment MicroBooNE added
  • Experiment deleted (-)
  • SSI Package art added
  • SSI Package deleted ()

Per the discussion at the stakeholder meeting, we will update artmod to indicate that cetskelgen should be used instead.

#10 Updated by Marc Paterno about 3 years ago

  • Has duplicate Bug #13833: artmod still produces #include "art/Utilities/InputTag.h" added

#11 Updated by Christopher Green about 3 years ago

  • Category set to Infrastructure
  • Status changed from Assigned to Resolved
  • Target version set to 2.05.00
  • % Done changed from 0 to 100
  • SSI Package cetpkgsupport added
  • SSI Package deleted (art)

Implemented with commit cetpkgsupport:fb390e4: artmod invocation will warn of deprecation (and mention cetskelgen as its replacement) unless art is detected in the environment of a sufficiently old version that artmod is still useful.

To reassure some of the commenters above: while cetskelgen has some extra features relative to artmod, it may be used in exactly the same way as before. Its principal differences lie in its modularity, allowing specifications for particular plugins to reside in the same package as the actual definition of same; and in its flexibility, now supporting new plugin types such as the timestamp plugins for art::EmptyEvent, or art::ResultsProducer plugins for output modules.

#12 Updated by Kyle Knoepfel almost 3 years ago

  • Status changed from Resolved to Closed


Also available in: Atom PDF