launchpad-dev team mailing list archive

[cheater cheater <cheater00@xxxxxxxxx>]

Hi guys,
sorry about the late (about a month) reply; I got quite a bit
real-lifed. Still am, but the OS community is important enough for me
to take the time out to revise this issue. Additionally I wanted to
give you guys some time to react and see what happens.

Danilo,
thanks for your comment. I have posted it on the page from my account.
I now understand that Launchpad's "basic unit" is the Project; I think
this is not communicated well enough by Launchpad; in fact I think
it's not communicated at all. Until I have read, and today re-read,
your comment I had thought that a project is in some non-obvious way
related to a repository, but that this link is not required and not
necessary. Basically, what I would be looking for in Launchpad is a
page - and a well-styled one, with a graphical overview of the process
- that says: "Want to put your code on Launchpad? It's easy! Just: 1.
start a new project 2. create a repository 3. upload code, create
bugs, wiki entries, etc 4. add build recipes and translations". The
whole process should not have more than 3 stages until I can start
working; it should not have more than 5 stages total. Of course, "add
build recipes" and "add translations" are equivalent and therefore
both in step 4, in that they both can be the next step from step 3.
Additionally, for someone who's not registered yet, I think Launchpad
should not take more than 4, 5, or 6 separate page loads from the
front page, through me registering, to me getting a bzr link to check
out an empty repo on my computer. In addition, the only data I should
need to enter should be my email address, requested account name and
password, requested project name, and project description; and what
features I want the project to have. Nothing more, no other complexity
to ward me off from working on Launchpad. I think this is entirely
achievable.

Thanks a lot for your reply, Matt.

 * Getting started with code hosting is hard/you need a project/+junk
is an unfortunate term

I think your proposed solution to that are right.

 * The profile page is overwhelming/there's no dashboard

You got part of that right.. however, you say "your 'area' is should
be easy to find and should guide you towards all those things you want
to do in Launchpad". I say "towards *just* those things". This is a
big difference and I think it nails down the cause of what's wrong
with the profile page right now, for my use. I don't want a page that
does everything; I want a page that does just the bare minimum that I
need. Too much stuff means the useful things become useless for me not
being able to find them.

In fact, I would suggest the ability to completely turn off features
of launchpad I don't need in my project. I don't need translations,
recipes, and other stuff; don't show it to me until I turn it on.
Ambushing the user with all that is very bad and it's the biggest
source of problems I have been having. In addition, no projects *start
out* (and by this I mean the situation where there's 0 lines of code
written) with package management builders, with translations, with bug
trackers, etc. You first want to create an experimental version,
probably working on your own for a few days, and only then start
adding those features *as they are needed*. This needs to be addressed
because it is very, very confusing.

 * Private branches are only available if you pay

I agree to disagree; in my opinion private branches can be very good
and facilitate easier release of things to the public; however this is
not the right place to discuss our difference on this specific topic
as it is of a deeper, philosophical and social, nature.

 * Launchpad lacks polish

I wouldn't say polish! I would say it lacks the paint, padding, and
the plaster is showing everywhere! I don't want to sound negative, but
this time I have to express strong disagreement!

 * Launchpad isn't social

It sounds like you guys need a mindstorm and a general plan in that
direction; I'm hoping this will happen; can someone responsible for
arranging this sort of thing please comment on whether this could
enter the ideas and planning stage, and if so, when?

 * Launchpad is too geared towards things other than just hosting code

I'm not sure I meant that; it's great Launchpad does other things, but
what I mean is that for my use the features I'd describe as core
functionality are not properly implemented yet.


In addition to this - have you guys created any issues in a tracker of
some sort that would allow tracking progress on this? I suggest having
a top-level "meta" issue for the general topic here and separate
issues that address the specific problems I have run into. (I would
find the time to be the stakeholder/QA on those). Please, do tell me
this is at this moment traceable.

Additionally, I have noticed this mention in the team leader meeting minutes:
> Matt:
>    * Happy to have the Usability and Communications Specialist job adverising out.

Is this in part related to the issues I have raised? Would this person
be responsible for improving the status of that?

Thanks



On Tue, Aug 30, 2011 at 23:58, Matthew Revell
<matthew.revell@xxxxxxxxxxxxx> wrote:
> Hey,
>
> On 30 August 2011 11:04, cheater cheater <cheater00@xxxxxxxxx> wrote:
>
>> I have recently tried switching from Bitbucket and hg to Launchpad and
>> bzr as Launchpad seemed "free-er" and bzr is the main DVCS I use. I
>> have not done so due to some issues I have run into.
>
>> I have brought up some problems I have experienced with Launchpad's
>> usability in an article, additionally making a few comparisons with
>> solutions in Bitbucket. After bringing this up in #launchpad I got a
>> fairly good response and they asked me to bring it up in this mailing
>> list.
>
> Thanks for taking the time to write up your experience in such detail
> and for raising the issues here.
>
>> Link below. Happy reading. Bear in mind some opinions might be strong
>> but I feel strongly about the future of services like Launchpad and
>> wish them the best.
>>
>> http://cheater.posterous.com/launchpad-not-an-alternative-for-bitbucket
>
> I'll address here some of the points from your post. I agree with what
> Danilo has said but I want to add some more. Let me know if Danil or I
> have missed something you'd like me to address.
>
>  * Getting started with code hosting is hard/you need a project/+junk
> is an unfortunate term
>
> It should be easier to get started with hosting your code on
> Launchpad. As Danilo notes in his reply, the fact that Launchpad does
> more than code hosting and that it has a project model are both
> important, and part of the reason that it appears more complex than
> other hosting sites, but they shouldn't be a barrier to getting
> started.
>
> We need that big "Push your code to Launchpad" button. If you don't
> have a project, it'll guide you to create a "personal", perhaps rather
> than "junk", branch that you later have the option of associating with
> the project.
>
>  * The profile page is overwhelming/there's no dashboard
>
> The profile/person page was designed so that other people could learn
> about you and your contributions to free software. I think it's not
> doing that job as well as it good be and it's not a substitute for
> having dashboard.
>
> We should rework the person profile pages and add dashboards.
> Launchpad should bring you the information you need to do your job ...
> your "area" is should be easy to find and should guide you towards all
> those things you want to do in Launchpad.
>
> Oh, and a lot of people really appreciate karma, even if it just a bit of fun :)
>
>  * Private branches are only available if you pay
>
> This is something I'm perfectly comfortable with. Canonical funds
> Launchpad for a number of reasons, not least of which is that it's a
> contribution to the free software community. It's a two-way deal:
> we'll provide Launchpad, its hosting, its full-time development team,
> etc for free use by free software projects because in exchange for all
> the investment we, as Canonical, make we get a richer free software
> ecosystem thanks to the free software development work that happens on
> Launchpad.
>
> We're also happy to provide private branch hosting but if the code
> isn't free/open, there's not the same incentive for us to provide that
> service, so we ask for a contribution to our costs.
>
>  * Launchpad lacks polish
>
> It's true, we lack polish in places. This is something we must and will fix.
>
>  * Launchpad isn't social
>
> When Launchpad was first around, it was more social than other
> solutions. It has a real focus on people, rather than just code and/or
> projects. But we need to catch up and go beyond that and address what
> me, you and many others want. That's something we're going to fix.
> Launchpad has masses of data about the work people do and the
> relations between people and projects. We can track more and we can
> make better use of what we have.
>
>  * Launchpad is too geared towards things other than just hosting code
>
> This is a plus point :) As a developer, are you sure you're not
> interested in how that code gets to people who use it? Either way,
> Launchpad's breadth of functionality should not stand in the way of
> you getting your work done easily.
>
> I know there's more to what you wrote but I hope that Danilo and I
> have addressed most of it.
>
> The summary of my reply is this: thank you so much for reminding us of
> where Launchpad can be better. It's true that we've heard much of this
> before but this sort of feedback is invaluable to us.
>
> Launchpad should never stand in the way of hosting, sharing and
> otherwise working on open source code. We need to make it easier to
> get started with Launchpad and to move on from those first steps.
>
> You'll find a lot of support from us at Canonical, and the Launchpad
> community generally, for addressing much of what you wrote. We have a
> lot of work already lined up but over the coming months the usability
> of Launchpad is going to be one of my top priorities. We have a full
> time designer/web developer, Huw Wilkins, on the team who'll be
> working closely with me and the Launchpad engineers to fix the
> usability issues you've addressed, along with others, and the general
> polish of Launchpad.
>
> It's an exciting time to be part of the Launchpad community. Thanks
> for taking the time to tell us where you feel Launchpad can be better.
>
> --
> Matthew Revell
> Launchpad Product Manager
> Canonical
>
> https://launchpad.net/~matthew.revell
>