I'd like to thank everyone that helped make this release a success. Often times people focus on coder contributions, but software releases are decidedly multi-disciplinary, and I'd like to specially credit four teams in particular for being well organized, thorough, and dependable.
Before the release really even got started, a strong foundation was built for us by the Bug Triaging Team. Especially after such a lengthy development period, with such extensive changes, there can be an overwhelming number of bugs. But due to how thorough the triaging was, we actually had very good transparency into what needed fixed the most.
In the past, we've learned that packaging of our windows and osx builds can make or break the release, and even a small glitch can throw everything out of wack. We certainly had our share of packaging challenges to overcome, and more that still need solved, but the Packaging Team developed a good strategy for charting us through to the release, to ensure we'd have functional packages for all platforms within two days.
Even now the Translation Team is hard at work on the release announcements and our website. This cycle we allocated ample time to their translation of the application, and I think that paid off in spades.
I think this may be the first release where our Web Team played such a crucial role. They established a solid infrastructure both for editing the website and for mirror-caching our downloadables; we have exercised both these functions extremely heavily the last few days, and while we've had a few brief (but quickly fixed) glitches, the infrastructure appears to be shouldering all the loads without complaint.
~ ~ ~
Going forward, my plan is to keep us on course to deliver at least a couple new major releases each year. So, like perhaps shoot for a 0.92 release this summer or early autumn. Maybe even more frequently than that, but to achieve this I think we will need to tighten up in a few areas.
First is marketing. Unsurprising after 4 years since the last release, our marketing abilities seem to have atrophied quite severely. We recognized this would be a problem and attempted to fix it; it's my fault we didn't get things organized here in time, but fortunately all the other teams filled in. We can try again for 0.92, and I will work harder to ensure we have a good solid marketing team by then.
Second is the release mechanics themselves, which are quite byzantine. Fortunately no one else needed to deal with this but me! A number of steps are manual which should be automated, other steps are redundant but can't be skipped for one reason or other, and there's some steps that have to be carried out by specific people or at specific times. The result is that release management requires handling by someone with release experience to know what knobs need turned and levers need pulled, when and by how much. Instead, we need to systematize all this logic in scripts, so that releases become more like a push-button exercise that any one of us could do.
A third area is user documentation. The only reason this hasn't been a release issue is because we haven't got any! :-) We get away with not shipping documentation thanks to having a quite excellent tutorial writing community and well-crafted release notes; users also thankfully have a selection of third-party commercial documentation to fill the gap. But to be considered as professional software, a user manual is a must. Since it'll need updated each release, this implies the need to create a Documentation Team.
Bryce
P.S. here's some of the press we've gotten so far:
Our news announcement: https://inkscape.org/en/news/2015/01/30/inkscape-version-091-is-released/
Was rather a last minute rush. Hopefully we can do better next time.
Slashdot announcement: http://news.slashdot.org/story/15/01/31/1914255/inkscape-version-091-release...
Amongst the usual comment chaff is some strong praise. "Best of Breed," "One of the handful of apps that together make up a good base set," "One of the coolest things it does [is] converting bitmapped images to vectors," "Have always preferred Inkscape", "Plays nicely with Inkcut - vinyl cutting," "My current business depends on Inkscape. I use Inkscape's 'gcodetools' plugin to generate gcode for a CNC mill." "I spoke with the very helpful people on Inkscape's IRC Channel for this tip." "Is it just me, or is this whole thread making you feel nostalgic for the good ol' Slashdot days when we gabbed about stuff that was a lot more fun to gab about?"
Much of the critical feedback is stuff we're already working on: UI design looks out of date, needs porting to Gtk3, needs more functionality before it can supplant Illustrator, Gtk+ is poor on Windows, need a more native L&F on OSX, needs proper CMYK / printing support.
http://fedoramagazine.org/inkscape-0-91-released/
An article by our own Ryan Lerch. A great overview of the release, and would have made a much better news announcement than the one we put out.
http://www.phoronix.com/scan.php?page=news_item&px=Inkscape-0.91-Release...
Unfortunately, this one was a leak of the release before we were ready for it to get out. I suppose they wanted to get the scoop, but with no packages available nor the news release posted, there was little to go on and it caused users some confusion and misunderstanding. I fault myself here for not having the marketing end of things better organized, so we could have kept the release hush-hush until everything was ready to go.
http://www.reddit.com/r/graphic_design/comments/2u6kxi/new_release_of_inksca...
Reddit also scooped the news release. But the comments here were a bit more interesting. A lot of pro-Inkscape cheer. A few areas where work is needed: Working with text, Saving as .eps, CMYK spot color.
Hopefully Monday we'll see wider uptake of our release.