Hi Oleg, First of all I am not against giving it a try. Let's try it.
P.S. Did not really get what was the main problem related to IGNITE-10399 [1]. I suppose it should go: simple problem -- fast fix. [1] https://issues.apache.org/jira/browse/IGNITE-10399 чт, 29 нояб. 2018 г. в 16:04, oignatenko <oignate...@gridgain.com>: > > Hi Ivan, > > You are right that these things will distract people and from this > perspective it is very well justified that vast majority of style deviations > (currently, all of them) are marked yellow. These are non-critical and if > developer ignores them nothing immediately bad happens. > > (For the sake of completeness one can argue that style deviations contribute > to technical debt but it's tangential to current discussion and here we can > probably consider them just harmless for simplicity.) > > The problem with five of these inspections that are proposed to change is, > these became different after IGNITE-9983 and above reasoning doesn't work > for these anymore. These five inspections are now checked at Teamcity and > when there are new deviations it reports problems. > > Essentially this means that developer introducing new violations in these > inspections is going to be distracted anyway - if they ignore at the coding > phase they still will be chased by the warnings from Teamcity. You can check > details of IGNITE-10399 because it shows a good example of how it goes. > > So what we're discussing here is essentially not about whether to distract > developer or not (because they will be distracted anyway) but when it is > more convenient to distract - at coding time or after Teamcity check. > Granted, delaying this to TC checks felt okay to me before we tried it but > observing how it really goes (in mentioned above IGNITE-10399) made me > curious if maybe we could try another option by raising this issue at coding > time instead. > > This is the whole point of this change, to let us try how it would go if we > warn developers about inspections impacting Teamcity in the time of coding. > As I wrote in beginning of this thread I briefly tried it myself and it > looked quite promising. > > To avoid misunderstanding, I would like to make it clear that at this point > it is not supposed to be a one way change because my testing was too brief > to say for sure if this is the way to go. Current plan is that we give it a > try for a while and later - depending on how folks feel - decide whether to > keep these inspections red or revert them back to yellow. > > Does that make sense Ivan? > > regards, Oleg > > > > > -- > Sent from: http://apache-ignite-developers.2346864.n4.nabble.com/ -- Best regards, Ivan Pavlukhin
