| 1 |
On Sun, Jul 02, 2017 at 03:55:54AM +0000, Duncan wrote: |
| 2 |
> William Hubbs posted on Sat, 01 Jul 2017 11:53:59 -0500 as excerpted: |
| 3 |
> |
| 4 |
> > See this article for why using liblua as a shared library is not |
| 5 |
> > recommended. |
| 6 |
> > |
| 7 |
> > http://article.gmane.org/gmane.comp.lang.lua.general/18519 |
| 8 |
> > |
| 9 |
> > Yes, it talks about the interpretor, but it goes further and really |
| 10 |
> > discourages even making a shared library available. |
| 11 |
> |
| 12 |
> PMFJI, but... |
| 13 |
> |
| 14 |
> That reply is from 2005 and is apparently specific to (32-bit) x86's |
| 15 |
> extreme shortage of general purpose registers. Back then it made sense |
| 16 |
> as 32-bit x86 was the dominant arch, both for gentoo, and (apparently) |
| 17 |
> for that lua discussion (which was in the debian context). |
| 18 |
> |
| 19 |
> That x86 general lack of registers was one of the big pressures behind |
| 20 |
> the switch to amd64, before system memory sizes increased to 4GB+, and |
| 21 |
> applies far less to today's dominant amd64, with x86 now legacy/embedded. |
| 22 |
> |
| 23 |
> Now it may well be that lua performance remains register sensitive even |
| 24 |
> with the increased number of registers available in amd64 and other |
| 25 |
> modern archs, but quoting an 11+-year-old email written in the extremely |
| 26 |
> register-restricted 32-bit x86 context does little to argue that point. |
| 27 |
> |
| 28 |
> So... got any equivalent links to posts with more modern figures? |
| 29 |
|
| 30 |
That link actually came from our current lua ebuilds. The shared library |
| 31 |
is offered, but the comments in the ebuild basically say the same thing |
| 32 |
and cite that link. Also, x86 is still a stable supported arch in |
| 33 |
Gentoo. |
| 34 |
|
| 35 |
> All that said, in FLOSS, he who volunteers, makes the rules, and |
| 36 |
> particularly given the upstream opposition meaning more gentoo-level work |
| 37 |
> required, if there's nobody willing to support lua in gentoo with dynamic |
| 38 |
> linking... |
| 39 |
|
| 40 |
You are correct. Basically we have to write our own build system and |
| 41 |
keep it up to date for every version of lua in order to support this. If |
| 42 |
someone can convince upstream to support it, this would be far better for |
| 43 |
everyone. |
| 44 |
|
| 45 |
> ... people unable or unwilling to volunteer to support their preferred |
| 46 |
> alternative get to live with one they don't like, whether that be |
| 47 |
> accepting what their existing distro provides even if it's not optimal, |
| 48 |
> switching to another, or supporting their own patches or builds apart |
| 49 |
> from gentoo. |
| 50 |
> |
| 51 |
> At least gentoo makes the latter case relatively easy compared to some |
| 52 |
> distros. I did it with kde twice, when gentoo/kde wasn't supporting |
| 53 |
> builds without semantic-desktop. =:^) |
| 54 |
> |
| 55 |
> And in this case it appears there's even someone already doing it and |
| 56 |
> making their work public via the lua overlay. =:^) |
| 57 |
|
| 58 |
The plan is to migrate what can be migrated from there to the |
| 59 |
tree once everything is up to date. |
| 60 |
|
| 61 |
William |
Archives