<html><head><meta http-equiv="Content-Type" content="text/html; charset=us-ascii"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;" class=""><br class=""><div><br class=""><blockquote type="cite" class=""><div class="">On Apr 16, 2020, at 7:46 PM, Mateusz Loskot <<a href="mailto:mateusz@loskot.net" class="">mateusz@loskot.net</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><div class="">On Fri, 17 Apr 2020 at 01:45, Mike Taves <<a href="mailto:mwtoews@gmail.com" class="">mwtoews@gmail.com</a>> wrote:<br class=""><blockquote type="cite" class=""><br class="">It seems that many projects have shifted from AppVeyor to AZP for<br class="">performance benefits.<br class=""></blockquote><br class="">Or GitHub Actions, which is an incarnation of AzP.<br class=""></div></div></blockquote><div><br class=""></div>Having done both GitHub Actions and AzP, I think Actions has more convenient expressiveness of job flow, a big win of integration with GitHub's API, and large directory of user-contributed actions.</div><div><br class=""></div><div>We migrated PDAL to AzP last summer after our experience at the OSGeo Code Sprint in Minneapolis with AzP on the <a href="http://gdal.org" class="">gdal.org</a> refactor. A big benefit was to simply have *one* CI platform instead of two or three. It was also wall-time faster, more concurrent, and most importantly had no cost to get better performance for open source projects (you need to pay for that from Travis or AppVeyor, which the PDAL project had done).</div><div><br class=""></div><div>GitHub Actions takes it even a step further, and it would be my choice for GDAL if someone has the capacity to invest the time to bring it forward. </div><div><br class=""></div><div>Howard</div><div><br class=""></div></body></html>