| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
| |
TODO: Use another approach instead (parse a manifest and use that
to checkoutout the source code)
Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@cyberdimension.org>
|
| |
|
|
| |
Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@cyberdimension.org>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This should improve speed a lot:
- Increasing the number of fetch jobs speeds up a lot the
downloads when resuming from a previously interrupted
download session as checking for new commits can have
a big latency, so checking many as repositories as
possible at the same time speeds up things a lot.
With Replicant 4.2 0004, using "-j 100" worked well for
me, however it didn't seem to have reached 100 parallel
download jobs, so using more would not improve
performance in my case.
- As for the number of compression threads, one has to consider
it carefully: xz -9e uses quite a big amount of RAM when compressing,
so using too much threads can result in xz being killed because
it used too much RAM. In another hand using 1 thread per core
seems to divide the amount of time spent compressing by the number
of threads.
Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@cyberdimension.org>
|
| |
|
|
| |
Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@cyberdimension.org>
|
|
|
The following text is part of the GPLv2:
If distribution of executable or object code is made by offering
access to copy from a designated place, then offering equivalent
access to copy the source code from the same place counts as
distribution of the source code, even though third parties are not
compelled to copy the source along with the object code.
Since we currently host images (which contains many executable or
object code) on ftp.osuosl.org, it would be best to also ship
the corresponding source code on the exact same server.
This way we could be pretty sure that the corresponding source
code is available at the same place.
This script can generate the corresponding source code tarballs
for all the Replicant 4.2 and 6.0 versions.
For that it simply downloads the Replicant source code of a given
tag (which corresponds to a release).
To save space it doesn't download the history (it uses --depth=1)
and it compresses the results with xz -9e, but we still kept the
git and repo information in case the build system uses it somehow.
Keeping the repo and git information also makes it compatible with
the current build instructions on the wiki.
Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@cyberdimension.org>
|
