1

Topic: Technical debt

2

Re: Technical debt

Hello, Ziaw, you wrote: Z> Each release reminds natural cataclysm and you need to guess only what it will be this time. It not a debt. This bankruptcy already ... Collectors already knock on a door. Z>  now stand so many that you understand: your program is so bad what more cheaply to rewrite it.  ... to Rewrite a pancake. And to produce a new technical debt. Therefore as in anew rewritten past miscalculations will be considered, but in a new variant will be imported  the new. For instead of an old minefield, we, solving "and well it , I will go on an adjacent minefield" we again we begin with zero, catching new mines, for a field new. To be played with a technical debt it is possible and it is necessary. But not each time from zero. It is necessary to correct, get rid of a technical debt not simply. And to analyze the reasons on which it appears. And not  in vacuum, "a pier here  my tin", the pier "all of us " is and so it is clear. And why it is is specific, in this most specific place, what exactly us  on "" the code (save-paste, and other guns to shoot to itself at a foot). What exactly provoked!?! And how to write so, and how to project, to avoid these provocative moments. Here then there is an experience, and the result will be better. And so, "to rewrite all ", especially in architectural sense, it is the infinite run on a circle. If we  on the first circle, threw and went on "other stadium", and then again threw and already on third "stadium" - whence, here whence thought that on a circle of Iks all will be excellent!?!

3

Re: Technical debt

Hello, Ziaw, you wrote: Z> Thanks, that who read up. For me one of the major concepts of development and the higher art of the architect I consider ability to balance, play a technical debt with a technical debt. Emmas... All  as though, but generally the technical debt and the other postponed risks is (, should be) in the majority of courses for QA/PM Work with a technical debt in the same way, as with all other resources of the project: , by in advance painted criteria we add in , sooner or later we take in hand and we repair. Not, it is possible , but it from discharge "anybody in the project does not track for  / certificates / ". One suffices Someone , someone attacks a rake to the last.

4

Re: Technical debt

Hello, Sinix, you wrote: S> Emmas... All  as though, but generally the technical debt and the other postponed risks is (, should be) in the majority of courses for QA/PM Probably. I work in small commands, I combine  and the architect. S> work with a technical debt in the same way, as with all other resources of the project: , by in advance painted criteria we add in , sooner or later we take in hand and we repair. S> not, it is possible , but it from discharge "anybody in the project does not track for  / certificates / ". One suffices Someone , someone attacks a rake to the last. The initial post is more work of art, it is not necessary to be in earnest about it too.

5

Re: Technical debt

Hello, Carc, you wrote: a C> It not a debt. This bankruptcy already ... Collectors already knock on a door. A C> Ugu ... to Rewrite a pancake. A C> And to produce a new technical debt. Therefore as in anew rewritten past miscalculations will be considered, but in a new variant will be imported  the new. For instead of an old minefield, we, solving "and well it , I will go on an adjacent minefield" we again we begin with zero, catching new mines, for a field new. You truly caught sense. That you so bombs?> What exactly provoked a C!?! And how to write so, and how to project, to avoid these provocative moments. The C> Here then appears experience, and the result will be better. Yes, and the last paragraph too. The C> And so, "to rewrite all ", especially in architectural sense, is the infinite run on a circle. Such council somewhere seemed to you? Or recalled the rake? Anything, on them very many came, it is normal. C> If we  on the first circle, threw and went on "other stadium", and then again threw and already on third "stadium" - whence, here whence thought that on a circle of Iks all will be excellent!?! Again you correctly understood all)))

6

Re: Technical debt

Hello, Sinix, you wrote: Z>> Thanks, that who read up. For me one of the major concepts of development and the higher art of the architect I consider ability to balance, play a technical debt with a technical debt. S> Emmas... All  as though, but generally the technical debt and the other postponed risks is (, should be) in the majority of courses for QA/PM S> Work with a technical debt in the same way, as with all other resources of the project: , by in advance painted criteria we add in , sooner or later we take in hand and we repair. Indeed. We, for example, add it at once in TFS (but not at once in iteration) and periodically we view the list, whether it is time to repair already.

7

Re: Technical debt

8

Re: Technical debt

Hello, Carc, you wrote: a C>>> ... Zachetnenko at you , what infantile odd fellow it is necessary to be Ugu, that on a minus to take offense and go  the last messages of other person. The poor fellow

9

Re: Technical debt

Hello, MTD, you wrote: MTD> Hello, Carc, you wrote: a C>>>> ... MTD> Zachetnenko at you , what infantile odd fellow it is necessary to be Ugu, that on a minus to take offense and go  the last messages of other person. That poor fellow not, surprised your response, somehow earlier I you not so here saw ... Went to esteem your messages ... Mum Well and something I suddenly . Yet did not pay attention that your messages in the core it in forums "sacred wars", "policy" ... Well all is clear: the boxer-theorist. Excuse, if a backlog, henceforth I will consider, more not to repeat. For in a policy, and others I rarely look, it in   on a zavalinka and so suffices.

10

Re: Technical debt

Hello, Carc, you wrote: a C> That not That yes a C> surprised your response That the surprising? You wrote judgement, I disagreed. Who knew, what you such sensitive? The C> Went to esteem your messages ... It it and is - . A C> Well all is clear: the boxer-theorist. , the poor thing as at you burns

11

Re: Technical debt

Hello, vmpire, you wrote: V> Indeed. V> we, for example, add it at once in TFS (but not at once in iteration) and periodically we view the list, whether it is time to repair already. Tools for control of a debt business the tenth, principal to learn it to see. Personally I type till now this experience and while edge I do not see. The intuition often works, I do not know, on what courses it give.

12

Re: Technical debt

Hello, Ziaw, you wrote: Z> Tools for control of a debt business the tenth, principal to learn it to see. Personally I type till now this experience and while edge I do not see. The intuition often works, I do not know, on what courses it give. The intuition here helps slightly. It is useful experience  on a rake (to estimate danger of the code) + approximate representation about dependences and plans under the project More likely. If the piece of the code with a debt is not problem in respect of errors or productivity, and in it is not planned any editings to isolate and let lives as is to . Otherwise (and situation plus when the nearest iterations demand the considerable editings in this piece) - the code the good candidate on repairing.

13

Re: Technical debt

Reminds the conservation law of complexity the Author (): Igor Tkachyov Data: 12/6/2002. Often developers think that can simplify (to pay less). But upon they simply take a technical debt. But full with the finance says analogs what to pay less without a debt taking all probably. The talent of the architect consists in defining where simplification, and where a debt taking.

14

Re: Technical debt

Hello, igor-booch, you wrote: IB> Reminds the conservation law of complexity the Author (: Igor Tkachyov Data: 12/6/2002. Yes, this article very much was pleasant to me in due time. IB> often developers think that can simplify (to pay less). But upon they simply take a technical debt. IB> the talent of the architect consists in defining where simplification, and where a debt taking. It agree. To see . and not to confuse it to simplification all the same art. I do not agree that it can learn on courses or to control it any tools. Earlier I used code metrics, but now somehow hammered, practically new they do not inform anything to me.

15

Re: Technical debt

Hello, Sinix, you wrote: S> the Intuition here helps slightly. It is useful experience  on a rake (to estimate danger of the code) + approximate representation about dependences and plans under the project More likely. The intuition is the concentrated experience. You do not like its usage in engineering craft? S> if the piece of the code with a debt is not problem in respect of errors or productivity, and in it is not planned any editings to isolate and let lives as is to . To understand, whether editings are planned in it experience is necessary. S> otherwise (and situation plus when the nearest iterations demand the considerable editings in this piece) - the code the good candidate on repairing. Or on a rewriting at the moment of scheduled editings. It too can be cheaper. We compare risk * rewritings and repairing cost now. While left even the right part we have no numerical methods of an exact estimation, the decision is accepted on the basis of intuition.

16

Re: Technical debt

Hello, Ziaw, you wrote: Z> the Intuition is the concentrated experience. You do not like its usage in engineering craft? Here not in "it is pleasant-not business is pleasant". A problem in spread of knowledge in a command. It is possible to transfer a specific dial-up of examples, with intuition of problem S>> If the piece of the code with a debt is not problem in respect of errors or productivity, and in it is not planned any editings to isolate and let lives as is to . Z> to understand, whether editings are planned in it experience is necessary. Review of plans under the project and readings  normally suffices me. We tell, if I know that in the project problems with sending of letters (besides from ) from time to time get out and in  hangs  about adding of promo-mailings - time to prick up the ears. At worst rescues analysis  if it appeared heavy. Something went not so? It is a high time not to be at war to last, and to stop and valuably to disassemble a problem. Well, i.e. To play back, find the reason, to outline in mind the plan of repairing and only then something to do. S>> otherwise (and situation plus when the nearest iterations demand the considerable editings in this piece) - the code the good candidate on repairing. Z> or on a rewriting at the moment of scheduled editings. It too can be cheaper. We compare risk * rewritings and repairing cost now. While left even the right part we have no numerical methods of an exact estimation, the decision is accepted on the basis of intuition. Yes, certainly. The main thing to make the actual decision about "we take in hand", and further who repairs - that and selects a method.

17

Re: Technical debt

Hello, Sinix, you wrote: S> it is possible to transfer the Specific dial-up of examples, with intuition of a problem Yes,  in a command are transferred, but very slowly. S> review of plans under the project and readings  normally suffices me. We tell, if I know that in the project problems with sending of letters (besides from ) from time to time get out and in  hangs  about adding of promo-mailings - time to prick up the ears. I would try to combine refactoring and  in one task or at least one executor that it could consider both tasks at performance of everyone. S> at worst rescues analysis  if it appeared heavy. Something went not so? It is a high time not to be at war to last, and to stop and valuably to disassemble a problem. Well, i.e. to play back, find the reason, to outline in mind the plan of repairing and only then something to do. All it is true in short-term perspective. When the development cycle goes on five-years periods, the analysis passes. It is simply impossible to remember or sort out all in . At me one 14 summer project worried 4 .

18

Re: Technical debt

Hello, Ziaw, you wrote: Z>... At all points