| 1 |
On Tue, 16 May 2006 15:56:38 -0700 Brian Harring <ferringb@×××××.com> |
| 2 |
wrote: |
| 3 |
| This whole thing seems a bit dumb; it's not that far off from someone |
| 4 |
| coming along with a non-compliant c compiler, and arguing that |
| 5 |
| they're still compliant, they just dropped the stupid stuff they |
| 6 |
| didn't like. |
| 7 |
|
| 8 |
And this is why your whole argument is nonsensical. C is a documented |
| 9 |
and fixed standard (or rather, several of them). |
| 10 |
|
| 11 |
| > | Add binpkgs, and try it- you'll get some fun behaviour there. |
| 12 |
| > |
| 13 |
| > As we're not emulating Portage's binary package format, it's not an |
| 14 |
| > issue at all. |
| 15 |
| |
| 16 |
| Doesn't matter *how* you bundle the ebuild data (cpio/tar/whatever), |
| 17 |
| the issue of reloading the env still will exist, the container for |
| 18 |
| the data doesn't matter. |
| 19 |
|
| 20 |
What's your point? |
| 21 |
|
| 22 |
| > A lot of the ebuild handling (especially environment) is done in |
| 23 |
| > C++. You're missing all kinds of things by only looking at the bash |
| 24 |
| > code. |
| 25 |
| |
| 26 |
| And portage does a lot of crazy shit in python for env handling |
| 27 |
| also. Sooner or later however, it hands control over to bash with a |
| 28 |
| pregenerated environment for the ebuild to execute in- what I keep |
| 29 |
| pointing out (and you keep dodging) is that the ebuild env *must* be |
| 30 |
| consistant, regardless of the actual pkg manager implementation. |
| 31 |
|
| 32 |
And hey, we do provide a consistent environment. |
| 33 |
|
| 34 |
| You deviate from the standard of the tree, you break your support |
| 35 |
| for the tree. Pretty simple, users being the ones who see the mess. |
| 36 |
|
| 37 |
Ok, so the *tree* is the standard now, not Portage? That's good, |
| 38 |
because the tree is the standard we're using. |
| 39 |
|
| 40 |
| > | Feel free to suggest them- since it's initial discussion, your |
| 41 |
| > | comments on it haven't really progressed beyond "y'all did it |
| 42 |
| > | badly", without naming a solution that works. |
| 43 |
| > |
| 44 |
| > I gave you two that worked. One, the .ebuild.n thing, which avoids |
| 45 |
| > the filename incompatibility issue and the source issue. Two, the |
| 46 |
| > "eapi as a function" thing, which avoids the source issue. |
| 47 |
| |
| 48 |
| 1) sucks- folks aren't going to be happy when they have |
| 49 |
| mplayer-1.0.1.ebuild.25 |
| 50 |
|
| 51 |
Yes, but unfortunately it's the only way around Portage exploding |
| 52 |
horribly when it encounters something it doesn't understand. |
| 53 |
|
| 54 |
| 2) EAPI as a function is no different then EAPI as a var, just |
| 55 |
| massively slower since you have to shell out more. |
| 56 |
|
| 57 |
Untrue. If the eapi function checks that the eapi is supported and |
| 58 |
bails with an understandable error if not, the rest of the file doesn't |
| 59 |
have to be source compatible. |
| 60 |
|
| 61 |
| > Which is a good thing. Rather than emulating one of Portage's |
| 62 |
| > sillier misfeatures, which only came about because of the whole |
| 63 |
| > "single repository" concept being hard-coded, we did it properly. |
| 64 |
| |
| 65 |
| So... let me see if I grok this- last response, state it supports |
| 66 |
| eclass unified overlays (blurb above). Now you're stating that it |
| 67 |
| doesn't, because it's one of portage's misfeatures. Please don't |
| 68 |
| waste the bandwidth doing that again. |
| 69 |
|
| 70 |
Try defining your terms properly. It supports overlays with a unified |
| 71 |
eclass directory. It doesn't support overlays with unified eclasses |
| 72 |
from different directories. |
| 73 |
|
| 74 |
| Stating it "won't be a silent failure" does not make it so- line 1023 |
| 75 |
| of portage.py directly contradicts that. |
| 76 |
|
| 77 |
There are other ways of making Portage fail non-silently, as you know |
| 78 |
fine well. |
| 79 |
|
| 80 |
| If your parent parsing implementation handled N parents on a single |
| 81 |
| line (rather then parent per line as you do now), portage would |
| 82 |
| explode rather then silently use the left most. Your implementation |
| 83 |
| isn't doing that however... |
| 84 |
|
| 85 |
Disallows spaces in paths. Which, admittedly, is probably a good thing. |
| 86 |
|
| 87 |
| > | Bit retarded here, but seriously, work it out in the overlay |
| 88 |
| > | (like most herds do), then try to bring it to the tree. |
| 89 |
| > |
| 90 |
| > Oh, we already went through that stage. |
| 91 |
| |
| 92 |
| So... where's the sample profile? :) |
| 93 |
|
| 94 |
Elsewhere in the thread. |
| 95 |
|
| 96 |
| > | That and the thread is getting fairly wide in discusion, rather |
| 97 |
| > | then the profile specific "does alt pkg manager X cruft belong in |
| 98 |
| > | the tree?" |
| 99 |
| > |
| 100 |
| > Well yes, because rather than discussing the issues, you're trying |
| 101 |
| > to turn this into your personal crusade against Paludis. |
| 102 |
| |
| 103 |
| Sorry you see it that way, meanwhile kindly address the points I've |
| 104 |
| been raising |
| 105 |
|
| 106 |
What points? When you start coming up with points relevant to a Paludis |
| 107 |
profile and stop using this as a personal vendetta I might be |
| 108 |
interested. |
| 109 |
|
| 110 |
-- |
| 111 |
Ciaran McCreesh |
| 112 |
Mail : ciaran dot mccreesh at blueyonder.co.uk |
| 113 |
|
| 114 |
|
| 115 |
-- |
| 116 |
gentoo-dev@g.o mailing list |
Archives