Brainstorm 2010/Garage and developer services

(New page: {{Brainstorm 2010}} == Garage and developer services == Subversion, autobuilders, issue trackers, mailing lists... what can be done to improve the Maemo application developer experience...)
(summarize http://talk.maemo.org/showthread.php?t=53003)
 
(3 intermediate revisions not shown)
Line 4: Line 4:
== Garage and developer services ==
== Garage and developer services ==
-
Subversion, autobuilders, issue trackers, mailing lists... what can be done to improve the Maemo application developer experience by the community?
+
Garage, subversion and git repositories, autobuilders, issue trackers, mailing lists... what can be done to improve the Maemo application developer experience by the community?
 +
 
 +
 
 +
* requests for improvements
 +
** do not auto-upload to public repo on succeeding build, give developer a method to check the actual result prior to "publishing" the package (gri)
 +
** have a purge button, so a successfully built package that doesn't look like intended, can be removed/overwritten by a modified version, without need for incrementing version number (synt: bundyo, nicolai)
 +
 
 +
* debate about alternatives
 +
** qgil says there are better alternatives nowadays, so why put effort into improvements of the current service
 +
** Ferenc Szekely points to functions of garage that are hard to implement in any of the alternative services, like upload privileges to extras
 +
** Jaffa about the current setup:
 +
*** #advantages:
 +
*** familiarity between projects
 +
*** cross-project collaboration
 +
*** streamlined tools for Maemo
 +
*** single authentication/authorisation point
 +
*** svn
 +
*** #disadvantages / requests for improvement:
 +
*** poor communication in the past (move to vcs.maemo.org and drop.maemo.org)
 +
*** concerns about qgil's suggestion to go into "maintenance mode" might cause bit rot
 +
** qgil calls for staying on topic, which is "what are the top priority tasks at maemo for the next 6 months". Then he claims maemo developer services has pretty much all it needs, and most of the developers will migrate to meego in a short while, so we shouldn't waste effort on improving maemo developer services
 +
** Jaffa (and Tekojo) disagrees on the meego part of qgil. In the scope of this brainstorm impact of meego on maemo still unclear.
 +
** Tekojo nicely summarizes: "''What it needs to do it does it already.'' Here I agree. Garage has been a fairly low maintenance, reliable service for a long time. Don't poke at it without a good reason. And if such reasons come up, Ferenc knows where to poke"

Latest revision as of 17:35, 24 May 2010

Image:Ambox_notice.png
This article is continued discussion from the Maemo brainstorm 2010
Please see the Brainstorm 2010 agenda for more.


[edit] Garage and developer services

Garage, subversion and git repositories, autobuilders, issue trackers, mailing lists... what can be done to improve the Maemo application developer experience by the community?


  • requests for improvements
    • do not auto-upload to public repo on succeeding build, give developer a method to check the actual result prior to "publishing" the package (gri)
    • have a purge button, so a successfully built package that doesn't look like intended, can be removed/overwritten by a modified version, without need for incrementing version number (synt: bundyo, nicolai)
  • debate about alternatives
    • qgil says there are better alternatives nowadays, so why put effort into improvements of the current service
    • Ferenc Szekely points to functions of garage that are hard to implement in any of the alternative services, like upload privileges to extras
    • Jaffa about the current setup:
      • #advantages:
      • familiarity between projects
      • cross-project collaboration
      • streamlined tools for Maemo
      • single authentication/authorisation point
      • svn
      • #disadvantages / requests for improvement:
      • poor communication in the past (move to vcs.maemo.org and drop.maemo.org)
      • concerns about qgil's suggestion to go into "maintenance mode" might cause bit rot
    • qgil calls for staying on topic, which is "what are the top priority tasks at maemo for the next 6 months". Then he claims maemo developer services has pretty much all it needs, and most of the developers will migrate to meego in a short while, so we shouldn't waste effort on improving maemo developer services
    • Jaffa (and Tekojo) disagrees on the meego part of qgil. In the scope of this brainstorm impact of meego on maemo still unclear.
    • Tekojo nicely summarizes: "What it needs to do it does it already. Here I agree. Garage has been a fairly low maintenance, reliable service for a long time. Don't poke at it without a good reason. And if such reasons come up, Ferenc knows where to poke"