1 |
On Wed, 19 Dec 2018 21:33:29 -0500 |
2 |
Rich Freeman <rich0@g.o> wrote: |
3 |
|
4 |
> I'm not sure how necessary this is, but another way to do this is to |
5 |
> just use tags, perhaps date-based (eg year-month). Perhaps this could |
6 |
> be combined with some level of QA as well to ensure the tree is clean |
7 |
> at the time it was tagged. From the command line this would be |
8 |
> simpler than copy/pasting hashes from some webpage, but it obviously |
9 |
> clutters the repo. Granted, it isn't much clutter if you only do it |
10 |
> monthly. |
11 |
|
12 |
Ew. Please no. Even when used appropriately, tags create a lot of mess |
13 |
when dealing with repos on a regular basis. Using them to simply |
14 |
communicate metadata is just wrong. |
15 |
|
16 |
My suggestion would probably be easier with some instrumentation in |
17 |
portage if we worked out how to do it, eg: |
18 |
|
19 |
emerge --sync-to=2018-12-21 |
20 |
|
21 |
*maybe* it could be done with a ref spec that doesn't collide with the |
22 |
tag/head space, enough that they show up in git ls-remote, but |
23 |
otherwise don't involve reference copying when people do naive git |
24 |
clones on stock configuration ( because syncing a bunch of tags that |
25 |
will never be useful after you've synced them is um... ) |
26 |
|
27 |
The downside though of that is using non-standard ref names will mean |
28 |
mirrors won't clone them by default. |
29 |
|
30 |
> Git fetch does not seem to support any kind of relative refspec. You |
31 |
> need a hash/branch/tag/ref. Git ls-remote just lists refs and not |
32 |
> history. |
33 |
|
34 |
> If super-unreliable connections are the concern it probably would be |
35 |
> cleaner to just use the previous suggestion of providing bundles with |
36 |
> resume support. They can be downloaded and then pulled/fetched from. |
37 |
> Do we really have that much of a need for this? |
38 |
|
39 |
Indeed, there's also the opportunity to replicate bundles via |
40 |
bittorrent, but not sure how much demand there is for that either. |