Commit 0e1df5b36e0a67fb66c9e7c78e4fc2aba00986eb

Authored by Melissa Wen
1 parent 777dcd13

[i3eSW] Reviewing the last 3 benefits

ieeeSW/releaseEng3/IEEE_ThemeIssue_ReleaseEng_CD.md
@@ -222,46 +222,50 @@ to production. @@ -222,46 +222,50 @@ to production.
222 222
223 ### Shared Responsibility 223 ### Shared Responsibility
224 224
225 -When the government technicians were responsible for deploying the project, the  
226 -developers lost track of what happened after code was delivered. After adopting  
227 -CD, they felt more responsible for what was getting into production. CD  
228 -influenced developers on taking ownership of the project. In the end of the  
229 -project, we noticed that the entire team was working to improve the CD pipeline  
230 -since they want to their new features in production.  
231 -  
232 -Interestingly, the CD pipeline also made the government requirements analysts  
233 -feel more responsible for the project. They were an active part of the pipeline  
234 -and that engaged them on the whole process. In the end, they were even actively  
235 -creating issues and discussing them during the development process.  
236 -  
237 -  
238 -[//]: # (TODO - depois deles entrarem de fato no pipeline, ou seja, validar em ambiente de homologação, criando issues e comentando nas issues do repositório é que de nosso processo empírico de desenvolvimento predominou até o fim do processo)  
239 -  
240 -  
241 -### CD pipeline protocol between Government and Development  
242 -  
243 -In the beginning of the CD pipeline use, a bottleneck arose at the acceptance  
244 -tests step due to delays in reviewing new features and starting the next step.  
245 -These delays occured because the government analysts, responsible for  
246 -reviewing, were sometimes busy or didn't have schedule this work. Furthermore,  
247 -after the acceptance of the new code, there was a bureaucracy in the government  
248 -IT infrastructure that often made us to wait until 3 days to get the production  
249 -environment access and then to start the deployment step. This problem was  
250 -softened when we clarify our pipeline for these analysts and they organized  
251 -their schedule to speed up reviews and production access requests.  
252 -  
253 -### Work in small batches developed trust in our relation with government analysts  
254 -  
255 -In the first three releases, government requirements analysts were validating  
256 -all features to be released only at the end of a delivery cycle which often  
257 -took almost four months. However their superintendents requested monthly  
258 -reports about the project progress and this brought pressure on them which in  
259 -turn put it on us. When we started to make continuous deliveries, we really delivered  
260 -intermediate and candidate releases. Thus, the analysts could have validate a  
261 -small set of features, making possible more accurate feedbacks and better  
262 -reports for regular meetings with their superintendents. As a result, we  
263 -gained their trust and they also gained trust of their chiefs about the SPB  
264 -project management. 225 +Before the adoption of the CD, the developers team could not track what happened to the code
  226 +after its delivery, since government technicians were the only responsibles
  227 +for deploying the project. The implementation of the referred
  228 +approach influenced developers on taking ownership of the project because it
  229 +made them feel equally responsible for what was getting into production.
  230 +
  231 +Interestingly, the CD pipeline had the same effect on the team of requirements analysts.
  232 +They were an active part of the pipeline and became more engaged on the whole process.
  233 +
  234 +After the incorporation of the pipeline into the work process, analysts
  235 +became more active in opening and discussing issues during the platform evolution.
  236 +Additionally, developers worked to improve the CD pipeline in
  237 +order to speed up the process of making available, in the production environment,
  238 +new features for the platform.
  239 +
  240 +
  241 +### Synchronicity between Government and Development
  242 +
  243 +Despite the positive impacts that the CD pipeline brought to the project, its
  244 +implementation was not easy at first. The good performance of the CD pipeline
  245 +depended on the synchronicity between the teams of developers and government
  246 +analysts, so that the work of one could be initiated immediately after
  247 +the delivery of the work by the other. Initially this concern was not
  248 +contemplated in the agenda of the governmental team, which generated delays in
  249 +the validation of the new features of the release. This situation combined with
  250 +governmental bureaucracy (up to 3 days) to release access to the production
  251 +environment resulted in additional delays for the deployment step to begin.
  252 +This problem was softened when the analysts realized the impact of
  253 +these delays on the final product and decided to allocate the revisions in its
  254 +scale of work and to request the access to production in time.
  255 +
  256 +### Strengthening trust in our work relation with the government
  257 +
  258 +Continuous delivery was also a tool that helped to strengthen trust in the
  259 +relationship between developers and government analysts, as well as between the
  260 +latter group and its superiors.Before using CD, analysts had access to the
  261 +features developed only at the end of the release, usually every 4 months.
  262 +However, this periodicity did not meet the requirements of their directors, who
  263 +demanded monthly reports on the progress of the project. With the
  264 +implementation of the CD, intermediate versions became available, allowing
  265 +analysts to perform small validations over time. The constant monitoring of
  266 +the development work brought greater security to the governmental nucleus and
  267 +improved the interactions of these with the team of developers.
  268 +
265 269
266 ## Challenges 270 ## Challenges
267 271