[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#635504: ITP: flashcache -- write back block device cache for Linux

Hash: SHA1


On 22.09.2011 04:38, onlyjob wrote:
> Congratulations upon mentors.debian.net hardware upgrade - well done.

Thanks, but unrelated :)

> As you may already know upstream made number of changes, notably:
>   *  updated documentation
>   *  initramfs hook for early flashcache_load invocation
>   *  updated Makefiles (upstream kindly accepted my minor patch
> (similar to yours) for $(DESTDIR) compatibility)

Yes. I merged upstream's most recent commit and prepared a new upstream
release, including your patch and thus removed mine.

> I updated the man pages by incorporating recent upstream documentation update.

I updated the man pages. However you can also do yourself, if you intend
to become my co-maintainer. Same holds for everything else you want to
improve - but see below.

> I'm quite inexperienced with git so I couldn't figure out how to
> access your repository.
> Is my Alioth account onlyjob-guest allowed to push?

Yes, it technically is. However I'd ask you to send me git format
patches instead of pushing changes, I can review first. Once I'm
convinced I can fully trust your work, I'll ask you to push directly
instead. I made you official co-maintainer, by the way [1]. Don't worry
on the Maintainer/Uploaders difference. That has merely historic
reasons, I don't have any more powers within the package than you.

> It appears to me that you might want to make a separate branch from my
> package but ultimately it is up to you.

I see no point in two packages at all. I think I merged all improvements
from your package into mine. If you disagree, let me know if you miss
As far as I know, you are only worried on the date string and I think I
solved that. Besides, your version number is wrong anyway, as its
1.0+... not 1.0~. Please compare outputs of:

$ dpkg --compare-versions '1.0+1' '<=' '1.0' && echo "True"
$ dpkg --compare-versions '1.0~1' '<=' '1.0' && echo "True"

> I'm not sure if you want to keep your workaround for mounting
> flashcache volumes on boot after introducing initramfs hook by
> upstream.
> (Frankly I didn't have a chance to test this functionality yet.)

Yes, but the initramfs hook does not setup anything. It only provides
the binaries in the initrd. It still does not provide any mechanism to
setup such volumes upon boot. You would need to hack a initramfs script
to execute a mount script upon boot.

> What do you think about version number differences between our packages?
> (I really like my get-latest-source target)

I updated mine to be more precise, i.e. I added hh:mm to the version
number. I think that's enough. Besides we do pretty much the same thing
in our targets, except that your package has more build dependencies
because you use bashisms and Perl. Both are essential packages for now
(well, perl-base at least), but there are some thoughts to remove bash
from that list. I'd prefer to keep dependencies low.

> Maybe we have at least one thing in common after all: I noticed you've
> chosen Perl to write flashcache_mount - I think it is nice because I
> love Perl.

There are surely more common things. Since we're interested in the same
package, that's a very visible common thing, isn't it?

I think we can stop CC:ing the bug now. Let's do any further technical
discussion in private.


- -- 
with kind regards,
Arno Töll
IRC: daemonkeeper on Freenode/OFTC
GnuPG Key-ID: 0x9D80F36D
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/


Reply to: