Launchpad learning curve and suboptimality

[Sampo]

This is my first fime using launchpad.net. I am experienced Unix sysadmin so I expected
things to be easy. Well they were not. Launchpad could be improved in several ways

1. First page of launchpad does not have any link to report bug. Especially
    figuring out how to report a bug about launchpad itself took time.
2. Requiring user to already know the project name is a major hurdle. It should
   be possible to report bug primarily by name of the binary involved.

In context of trying to add "Also affects distribution/package" to
Bug #431975: downloading a torrent to an encrypted home partition hangs and uses 100% CPU

The URL is: https://bugs.launchpad.net/ubuntu/+source/ecryptfs-utils/+bug/431975/+distrotask

3. Distribution menu lacks my distribution: LinuxMint12. Either the menu should
    at all times contain all distribution names (impossible) or option "Other" should
    be offered.

   Given that LinuxMint is quite mainstream distribution, I am surprised to see it
   missing from the menu. If it is missing because it is considered competition
   and there is no desire to support it, then it should clearly say so and perhaps
   point the user to the appropriate LinuxMint bug tracker.

4. The requirement to name a source package is quite onerous and difficult
    for a user to satisfy (the page claims it is optional, but in practise it is impossible
    to submit the page without it).

5. Clicking "Choose" brings up dialogbox. Since I think bug #431975 is a kernel
    related bug, I typed in search area "kernel". This results

    "Too many matches. Please try to narrow your search."

    which is not helpful. While I understand that word "kernel" may match
    gazillion things, I also think that "kernel" is the most succinct and accurate
    way to describe the module I am trying to refer to. In fact it is impossible
    to imagine any other way of describing it. How can I possible narrow the
    search? A kernel is a kernel and there is nothing more to it.

    Perhaps the search box could show at least first 100 results and have a way to
    make negative searches so I could eliminate all the other clutter that
    obscures the simple module called kernel.

6. Repeat (5) for keyword "init". There really is a Unix process called init and
    there is no other accurate way to call it. It should be possible to find
     "init" without getting confused with the gazillion "initialize" matches.

Reading the thread on https://bugs.launchpad.net/ubuntu/+source/ecryptfs-utils/+bug/431975

"This bug was filed against a series that is no longer supported and so is being marked as Won't Fix. If this issue still exists in a supported series, please file a new bug.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team."

7. The bug in question clearly is a serious and real bug. It also has not gone away (hence my
    attempt to report it as still open on newer distributions). It is blatant disregard of
    of the bug reporter's time and effort that a robot can nullify all work done
    in reporting and tracking the bug.

    Such automated mechanisms to close bugs that in reality have not been closed
    do not contribute to quality in open source software. The robot may clean statistics,
    but it also causes the bug to persist in real life as pressure is taken off the module
    maintainer. This means focus shifts on new features that create even more bugs
    rather than quality increasing. I'd rather have better quality software than more
    buggy features.

I'll keep posting more suboptimality as I try to make my amendment to #431975
or report new bug.

Feature request:

8. Please support Single Sign-On in launchpad.net. In particular, I would appreciate
    SAML2 support (but OpenID-Connect is also OK). I am even willing to pitch in
    system integration effort - I am lead developer of zxid.org.

Cheers,
--Sampo

[Sampo (sampo)]

Suboptimality #9

When trying to submit https://bugs.launchpad.net/ubuntu/+bug/913787 (eventually successful),
I first saw the list of potentially related bugs, clicked "No, I need to submit new bug",
then composed the bug and Submitted it.

I was then taken again to related bugs screen to see if mine was duplicate and apparently
my bug was not yet in the system. This is quite bad as I could have lost my valuable
bug report just due to some navigation error or bug in process (majority of online
shopping carts are lost due to too many steps in process and some step having
bug - by adding more steps you compound the chances of some bug manifesting).

I thenk clicked "No, I need to submit new bug report" again and I returned to the
composition screen (luckily the system did not loose my composition - I definitely
feared this would happen).

Evidently it is impossible to submit a bug whose summary line causes a bug to be
considered potentially related. I was forced to change the summary so that
it would not match any of the other bugs. In effect I had to make the summary
less meaningful just to work around the brain damage of the duplicate bug
detector. Thus your bug tracking system requires me to degrade the quality
of my bug reports. Now that's broken.

Also, given that I am a compentent Unix sysadmin and open source author myself,
I wonder if I really need to jump through all these hoops to contribute. Could you
flag me as competent user so all this could be avoided in future?

Cheers,
--Sampo

[Martin Pool (mbp)]

Hi, Sampo,

Thanks for using Launchpad and for sharing your feedback. I agree most or all of the things you mention are real bugs or suboptimalities, but most of them already have bug reports: for instance bug 210943 asks for accepting openid from authorities other than login.ubuntu.com.

We want to have one bug for each distinct bug, rather one for every user who has feedback. So, if you find other issues, please do tell us, but it's probably better to post to the launchpad-users list <https://launchpad.net/~launchpad-users> than putting them all in here.

#9 may be a new bug but I don't really understand what's happening there.

[Sampo (sampo)]

Marking this as "invalid" is a slap in the face of the bug reporter.

I was just trying to improve your platform. Do not expect such collaboration
in future.

Indeed this is the central problem of the open source "culture": the autocratic
statistic chasing admins eliminate all the bugs as invalid before the developers
have chance to fix them.

Wiith such practises, let the open source burn in hell. I will change my stance
to closed source where at least the support can be arranged.

--Sampo

P.S. It is not like my report was poorly researched, but rather
that you did not want to address it.