I have followed this thread with a somewhat bemused attitude. There have been comments, coming from both sides of the discussion, that are somewhat jaundiced in the viewpoint expressed. Being a former programmer and application architecture developer, I thought it may be helpful to consider some things that have been misconstrued.
First, I should be able to choose any operations in any sequence without affecting the pipeline provided the architecture is truly independent of those selections. By this statement, I mean that if the pipeline order is predetermined, then my order of choosing operations should have no effect on the pipeline. This would also imply that the order of displaying the operations should have no such effect. I stress this is true if the architecture supports it. I have designed systems where this is true, so I know it can be done. Second, if the architecture is in some way dependent on the order that operations are displayed, this would be a significant overhaul of the code. When an architecture is designed and selected for a project, it is imperative that all code adhere to the constraints of the architecture. To disregard those constraints is to create a system that is not only not stable, but is also not consistent, correct. and clean. Since only the developers know the architecture intimately, all users can do is request changes, enhancements, etc. and rely on the developers and architects to approve or disapprove those requests. Third, it is impossible for a user to say whether a change is easy or not. What may seem easy could be incredibly difficult, once again depending on the architecture. So it is an exercise in futility to argue for an enhancement when one has no idea of the issues involved in incorporating that enhancement into an existing architecture. As in the paragraph above, one can only ask those who understand the architecture and issues. Finally, it serves no purpose, in my opinion, to denigrate anyone who sincerely advocates for some change when that person does not understand the essence of what they are requesting. Rather we should be willing to give direction and/or answers that promote understanding. If someone persists in advocating for something after reasonable efforts have been made to answer that individual, then do not keep relighting the fire by being rude or obnoxious. Just let the matter drop. It will die of it's own accord. I apologize for this lengthy missive. I just think this went way beyond where we need to be. On 2/17/21 12:05 PM, Remco Viëtor wrote: > On mercredi 17 février 2021 17:39:23 CET Maurizio Paglia wrote: >> First of all I desire to point out it was not my intention to be rude or >> blame nobody. >> Nevertheless the reply " that's quite easy - because none of those who want >> this feature become >> active, fork the code and simply implement the stuff ... " is really >> impolite. >> I cannot write C and cannot start learning C at 52 only to send a pull >> request that will be ignored... > Keep in mind that that particular reply was given *after* OP's question was > answered with an explanation why changing the current interface was not a > good > idea (according to devs and several users). > And *after* a few others jumped in claiming that it (or something similar) > should be implemented because "intuitive interface". > > I got the impression they thought that it couldn't be all that difficult to > do... So it was proposed they do the work (after about 70 messages in the > thread). > > > Oh, and while you claimed you were not asking to change the dt GUI, in > practice what you are asking *is a change in the interface*. Currently, > changing the module order has a well-defined effect: it changes the pipeline > order *to reflect the module order in the interface*. And changing that link > between user action and effect is a change in the interface (and a rather > important one). > > "The interface" is not just how things look on the screen, it's also how > interacting with it influences the program's behaviour and results. > > > > > ____________________________________________________________________________ > darktable user mailing list > to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org > ____________________________________________________________________________ darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
