Re: Comment posted on sqldeveloper-package
Hi, Phil.
It'll be a lot of work, but I have to change my email address, I've
been having problems with BigFoot service, and only yesterday noticed
it was failed.
Probably it doesn't matter who's name appears on the page as long as
the package gets sponsored which, from my experience, isn't that easy
and the reason I gave up in 2012 of keeping uploading my newer
versions.
I develop all my packages in "testing", but I've tried to have them
backward compatible, so I believe that all of them will work on
previous releases counting from the time they were published.
I've noticed the key issue and I believe you are right, it's the 1024
bit that might be issuing the warning, I'll look into it also.
I didn't notice you had issued a package, just noticed the bug report,
I haven't been using 0.2.4 since Nov 2012. The code base became quite
different from 0.2.4 to 0.3.0, not sure how to add an out of order
entry to changelog, if I put it in sequence with the version number
the date is completely wrong, if I do it by date the version is out of
order. In fact your package is more of a fork than an inline version.
Since it's native, the changelog isn't just for the uploaded versions,
it's for all the ones I've made, even those I haven't made public,
like those prior to 0.2.3.
Never used Alioth, just GitHub, and for a long time just my own
machine. No issue was raised before about this, so I'm guessing it's
not a requirement.
I've revised your "stretch-pu" package and though the changelog isn't
quite accurate (--no-tgz-check is used, not because sqldeveloper isn't
free but, because this is a native package and doesn't need a ".orig"
folder or archive), the patch to make-sqldeveloper-package is
pristine, so I see no issue in it.
Regards,
--
Lazarus
On Mon, Oct 30, 2017 at 9:55 AM, Phil Morrell <debian@emorrp1.name> wrote:
> Hello Lazarus, pleased to meet you.
>
> I attempted to get in touch with you over the summer, but
> unfortunately only got bounces from bigfoot.com.
>
> On 29 October 2017 at 22:57, mentors.debian.net
> <support@mentors.debian.net> wrote:
>> Lazarus Long made the following comment about the sqldeveloper-package package:
>>
>> Not sure what happened, but although I uploaded this package, it went under Phil's name.
>>
>> I'm guessing it's an interface bug, and not a problem. If I'm wrong please advise.
>
> I currently have an upload prepared for stretch-pu:
> https://bugs.debian.org/879850 I'm also guessing from the URL
> structure that there can be only one (first come, first served)
> uploader on mentors.debian.net per package but since your version was
> correctly signed, it was added to the existing page. If you like, I'm
> happy to reverse that by deleting the package page and re-uploading
> after you, just let me know.
>
> As for v0.4.2, thanks for fixing the outstanding bugs, I was only
> interested in getting it working on stretch and unfortunately I'm not
> at work today, so can't test the .deb. I did have to use dget
> --allow-unauthenticated on your .dsc otherwise dscverify complained
> about the signature, but gpg --verify did not. Hmm, dget-ing my own
> upload had the same issue - do I need to explicitly specify the
> keyring somehow, or is the 1024 bit key causing issues with gpg
> modern?
>
> In the changelog, please include my nmu entry verbatim (as it has
> reached unstable and instead of ack in 0.4.0), so bugs.debian.org
> correctly shows the graph for #868673. I don't know if the deb9u1
> changelog entry also needs including, I've never done a stable update
> before. I also thought the changelog should only contain versions
> actually uploaded to unstable, rather than UNRELEASED ones, but that
> might just be the behaviour of git-buildpackage.
>
> As the package is debian native, does the VCS need to point to alioth?
> I know it's fine for there to be a mirror, accept contributions, or
> for the upstream to be on a non-free service, but I thought all the
> debian packaging needed to be DFSG hosted.
>
> Thanks,
> --
> Phil Morrell (emorrp1)
Reply to: