Blog by Sumana Harihareswara, Changeset founder
Getting Autoconf Unstuck
Hi, reader. I wrote this in 2020 and it's now more than five years old. So it may be very out of date; the world, and I, have changed a lot since I wrote it! I'm keeping this up for historical archive purposes, but the me of today may 100% disagree with what I said then. I rarely edit posts after publishing them, but if I do, I usually leave a note in italics to mark the edit and the reason. If this post is particularly offensive or breaches someone's privacy, please contact me.
For most of this year, Zack Weinberg and I have been working on a pretty ambitious project:
- to make a fresh release of GNU Autoconf, a crucial free and open source build tool that hadn't had a new release since 2012
- to get paid for that
- to help put Autoconf on a more sustainable footing so it doesn't have to get rescued again a little while down the road
Autoconf 2.70 is due out this month; if you use Autoconf, check out the 2.69e beta and test it soon since Zack aims to make the release on December 8th.
If you hear "Autoconf" and think "I don't even know what that is or why it is important", you can read my LWN story about the rejuvenation & what's next.
(I am proud that a person said "That's one of the best pieces of technical writing that I've read in a long time." about my article.)
Several companies use/depend on Autoconf internally and would like for Autoconf and the entire Autotools toolchain to get back on track. There's lots of code out there already depending on autoconf. Converting it would be risky and expensive. Plus, competing build systems don't cover all the edge cases Autoconf does. If this makes you nod, check out the 2.69e beta and test it.
But also, the funding we got has run out, so we're trying to get some corporate sponsorship to make 2.71 even better (including building out a robust continuous integration system), and get the project on a sustainable footing. We'd like to:
- test Autoconf with complicated autoconf scripts and find and fix more regressions
- set up proper CI so we can find regressions on lots of OSes
- get the hundreds of disorganized patches and bug reports filed, so we can prioritize and assess our backlog
Even a donation as small as USD $5,000 could help make substantial progress. If you want to directly pay Changeset to work on this, email me and let's talk. Or: the Free Software Foundation, a 501(c)3 nonprofit, collects donations on behalf of the GNU Toolchain (see their list of Working Together for Free Software Fund project areas), and your organization can make a tax-deductible donation to the FSF targeted at GNU Toolchain maintenance.
