configure: update to autoconf 2.71#2281
Conversation
|
Current stable releases of e.g. Debian and Fedora come with autoconf 2.69. I suggest to wait with this PR until these two important Linux distros provide autoconf 2.71 in their latest stable releases. |
This is not an urgent issue. I was foremost curious how much changes was needed to update and apparently there was not much. All the same, I tagged this to 8.4 and had no intention to suggest to backport it. |
I just checked https://repology.org/project/autoconf/versions :
|
|
bookworm as been released |
|
@nilason Please can you rebase this PR? |
#2189 is stuck as requires @wenzeslaus approval to get merged. I'll be mostly off starting from the next week thus (unless someone else merges it), no action till the new year. |
If it's okey with you, I'll go ahead with the rebase (and possibly merge if so agreed). |
Yes, feel free to do so. I'll manage with merging changes into #2189 (and, hopefully, one day be able to merge it too). |
|
Rebased. Looks good overall, but Travis' Clang build fails. Seems unrelated to me... |
There is a new Travis docker build, created in Dec 7. |
|
Ahh, Travis is a good boy again. |
|
According to https://repology.org/project/autoconf/versions
I believe the day has come to upgrade and merge this PR. Note: Although the changes are small, it is preferable to merge this in good time before next release, to be able to test it properly. |
wenzeslaus
left a comment
There was a problem hiding this comment.
I can't really judge the details, but overall this looks hopeful and checks are passing. I like the use of #error instead of raw text choke me in the code.
|
@nilason Please free to merge this PR (there are already three approvals)... |
|
I had maybe one last question, but it got merged when writing the question. Where the changes in i.cluster/global.h intended in this PR (to fix something), or was a merge/rebase artifact? I didn't read a mention of it in this thread |
It was intended, obsolete configure tests/macros with 2.71. |
|
Perfect then! |
Update GRASS configure to autoconf 2.71 (from present 2.69).
From autoconf 2.70 announcement:
(see more)
Announcement of 2.71:
This is a smaller jump than the one from 2.13 to 2.69 (#1845), and (hopefully) avoids issues with patched distributions of autoconf, e.g. #2225 and #2280.
autoconf 2.71 News