Linux 6.9 will be the first to top ten million Git objects Linus Torvalds has released version 6.8 of the Linux Kernel. “So it took a bit longer for the commit counts to come down this release than I tend to prefer,” Torvalds wrote on the Linx kernel mailing list on Sunday, “but a lot of that seemed to be about various selftest updates (networking in particular) rather than any actual …
As it should be, honestly. Numeric punctuation is the one thing that imperial and metric should have agreed upon a single, worldwide standard for. Never mind the size of a git repo, we've lost space probes due to people from decimal-comma countries misreading "3,000km" as "three kilometers" instead of the correct interpretation of "three thousand".
I feel your pain. Not just the comma as the decimal separater but to make things especially interesting, the period as the thousand separater.
Thankfully, at least for web interfaces since HTML5 an input type of number will accept something like 3.210,5 (if the page is declared as being in German) and convert it to 3210.5 before sending it to the server.
> Numeric punctuation is the one thing that imperial and metric should have agreed upon
Let’s take the first step: SI should have agreed a single form of punctuation rather than permit the continuation of different national conventions.
> decimal-comma countries misreading "3,000km"
Interestingly, The recommended solution is to use space, so that becomes 3 000km, thus leaving the only dot/comma in the decimal point position. Only catch is the space gives rise to parsing problems…
Is it the whole git archive having 10^7 objects, or are those the actual objects for the single version kernel? The article headline seems to indicate the latter, Linus' quote states the former.
So before any of the "bloatware" crying fraternity show up: nothing to see, move along. This is not the issue you are looking for.
"Is it the whole git archive having 10^7 objects, or are those the actual objects for the single version kernel? The article headline seems to indicate the latter, Linus' quote states the former."
From TFA "6.9 will be the first to top ten million Git objects"
Linus: "This is the last mainline kernel to have less than ten million git objects" (speaking of 6.8).
IOW, single version kernel.
I regularly do fetch/clone operations, and it is (slightly) amusing to see that the percentage progress indicator goes by number of objects, not their actual size. So it will quite often seem to stall at a certain percentage value while pulling down a few large objects, only to suddenly leap ahead when it gets to a lot of small ones.
If you are missing one for your favourite architecture, why not create it yourself?
As an example, think of how Raspbian came into being: it was because Debian did not have an optimized build for the particular combination of architectural features (older ARM instruction set + hardware floating point) that the first Raspberry Pi was using. In other words, there was already Debian-for-ARM, but it was for the wrong kind of ARM. So these two guys took it upon themselves to recompile the whole of Debian for that particular architecture. They got the bulk of the job done in six weeks.
So you see, it’s not that hard, is it?
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
