| 1 |
On Thu, 28 May 2009 21:46:48 +0200 |
| 2 |
Patrick Lauer <patrick@g.o> wrote: |
| 3 |
> > And just how do you plan to enforce that? What measures will you |
| 4 |
> > take to ensure that there's no way for developers or users to |
| 5 |
> > modify the repository? |
| 6 |
> I can think of many simple methods. Like a tarball with a checksum. |
| 7 |
|
| 8 |
...which a user can modify once it's been extracted. |
| 9 |
|
| 10 |
> Or a zipfile. |
| 11 |
|
| 12 |
...which a user can modify once it's been extracted. |
| 13 |
|
| 14 |
> Or a git repository. |
| 15 |
|
| 16 |
...which a user can commit to without telling the package manager that |
| 17 |
the cache is now invalid. |
| 18 |
|
| 19 |
> That's all implementation details. But I think we can agree without |
| 20 |
> too much pain that it is possible to have a reasonably tamper-proof |
| 21 |
> format that we can consider readonly for these purposes. |
| 22 |
|
| 23 |
No, I do not in the slightest bit agree that there is an easy way of |
| 24 |
ensuring that what the package manager sees hasn't been tinkered with |
| 25 |
by a user or developer who "just wants to try a quick change out". |
| 26 |
|
| 27 |
> > We can't make changes because the package manager needs to know the |
| 28 |
> > EAPI in order to parse versions, since once we make changes versions |
| 29 |
> > will be defined in terms of EAPI. |
| 30 |
> |
| 31 |
> ... why? You're stating one possible scenario as the only potential |
| 32 |
> solution again. That ignores the other methods that were mentioned on |
| 33 |
> this mailinglist and in other places where you lurk. Please stop |
| 34 |
> being so dishonest. |
| 35 |
|
| 36 |
No-one has provided a viable way of extending the version format that |
| 37 |
doesn't require EAPI changes. So unless you're talking about your |
| 38 |
"start a whole new tree" idea, which has already been debunked to |
| 39 |
death... |
| 40 |
|
| 41 |
> > We want to make changes because, as has been stated several times |
| 42 |
> > previously, allowing 1.1_rc1 but forbidding 1.1-rc1 is entirely |
| 43 |
> > silly and arbitrary. |
| 44 |
> See Intercal versioning. Also silly and arbitrary to avoid it, but |
| 45 |
> noone has provided a proposal to accomodate nonincreasing and |
| 46 |
> nonmonotonic versioning systems. I doubt you would want those allowed. |
| 47 |
|
| 48 |
No-one is suggesting making changes to match silly upstream versions. |
| 49 |
What we are suggesting is making changes to match sensible and |
| 50 |
reasonable upstream versions. |
| 51 |
|
| 52 |
-- |
| 53 |
Ciaran McCreesh |
Archives