| 1 |
Ciaran McCreesh wrote: |
| 2 |
> On Tue, 24 Feb 2009 17:04:28 +0100 |
| 3 |
> Luca Barbato <lu_zero@g.o> wrote: |
| 4 |
>> Ciaran McCreesh wrote: |
| 5 |
>>> On Tue, 24 Feb 2009 08:08:23 +0100 |
| 6 |
>>> Uh, your benchmarks are nonsense. |
| 7 |
>> Provide your nonsensical ones. |
| 8 |
> |
| 9 |
> You're doubling the number of files that have to be read for an |
| 10 |
> operation that's almost purely i/o bound. On top of that, you're |
| 11 |
> introducing a whole bunch of disk seeks in what's otherwise a nice |
| 12 |
> linear operation. |
| 13 |
|
| 14 |
I see words, not numbers. |
| 15 |
|
| 16 |
>>> That is not how metadata checks work. |
| 17 |
>> Explain how they work, regen works that way... |
| 18 |
> |
| 19 |
> If metadata is valid, ebuilds aren't opened at all. An optimal |
| 20 |
> implementation can slurp up the entire directory in one go and then |
| 21 |
> start pulling out cache entries as it needs them, not having to go back |
| 22 |
> to the ebuild directory or read its contents at all. Then it can open |
| 23 |
> and read cache files in a carefully selected order to avoid having to |
| 24 |
> do any more opens than necessary. |
| 25 |
|
| 26 |
So? if the cache is valid then you don't have to source them at all. If |
| 27 |
you have to regen it, well you have to read everything. |
| 28 |
|
| 29 |
>>> By parsing the ebuilds you're talking doubling the number of file |
| 30 |
>>> reads required to get the job done, and massively increasing the |
| 31 |
>>> number of seeks required. |
| 32 |
>> Apparently it doesn't impact anything. |
| 33 |
> |
| 34 |
> Please show the patch you created (for Paludis, since Portage doesn't |
| 35 |
> yet do a lot of the optimisations it could here) that demonstrates this. |
| 36 |
|
| 37 |
Paludis isn't portage. |
| 38 |
|
| 39 |
>>> But that isn't even the main issue. The main issue is that even if |
| 40 |
>>> you retroactively pretend that all ebuilds are in a format they're |
| 41 |
>>> not, and ignore the breakage, and then wait for a year for package |
| 42 |
>>> managers to try to parse your new format, you *still* can't change |
| 43 |
>>> name or versioning rules. |
| 44 |
>> why? when portage would breanch if I put an ebuild with a wacky |
| 45 |
>> version AND there is a valid cache for that telling its eapi 99 ? |
| 46 |
> |
| 47 |
> Because it has to parse that version. Also, the package manager can't |
| 48 |
> tell whether or not a cache entry is valid if it doesn't recognise the |
| 49 |
> EAPI in the cache entry. |
| 50 |
|
| 51 |
unknown isn't unsupported? |
| 52 |
|
| 53 |
>>> Again, these are all things that have been discussed at length |
| 54 |
>>> previously. Please either come up with a legitimate technical |
| 55 |
>>> objection, or admit that you've seen the light. |
| 56 |
>> the glep doesn't show any of those nor reference to it, as I said |
| 57 |
>> before, do your homework and probably more people will be happier |
| 58 |
>> with your proposals. |
| 59 |
> |
| 60 |
> Why should it? The C++ standard doesn't explain why you should use it |
| 61 |
> instead of Java... |
| 62 |
|
| 63 |
In fact many people do wonderful things with java and many more just do |
| 64 |
over engineered mess with C++? |
| 65 |
|
| 66 |
lu |
| 67 |
|
| 68 |
-- |
| 69 |
|
| 70 |
Luca Barbato |
| 71 |
Gentoo Council Member |
| 72 |
Gentoo/linux Gentoo/PPC |
| 73 |
http://dev.gentoo.org/~lu_zero |
Archives