KDE 4.0.1 Released

As was stated previously, GNOME 2.0's release was similar, and from what 
I've ready they pushed out updates weekly (or nearly that often). A lot of 
people were confusing plans for KDE 4 as promises for what'd be in KDE 
4.0.0.

It was noted in KDE's Commit Digest that after the 4.0.0 release was made, 
there was a sizeable increase in the amount of activity in the SVN 
servers, leading one to believe that the release of 4.0.0 attracted more 
developers (or the developers were getting better/more information).

It seems odd to me how principles that have worked so well for Free 
Software for so long (such as Release Early, Release Often) are suddenly 
seeming to be forgotten. Also KDE 4.0's development time was originally 
said to be too long, but then it suddenly became too short ;)

> The bugs in the KDE4 rc series did not need wide spread use
> to find them, just a small amount of attempted usage.

you're referring to a handful of issues with plasma, i assume. since 4.0 many non-obvious
issues have been reported against various components in that release. remember, it's a bit
more than a window manager and a desktop shell. so, no ... many of "the" bugs (whatever that
means, exactly) in the 4.0 release were not found until people started using the various apps
and libraries underneath. visiting bugs.kde.org shows this pretty well.

even in plasma there were several interesting issues as well as feature requests that only
came out since the release and more people using it.

suggesting that because some issues are obvious that there are no non-obvious issues is odd.

> was get the release machinery sorted

that is indeed one of the positive things that happened as a result, yes.

other things that happened: third party devs could more easily start earnest work on porting,
kde devs shifted in thinking from "one day it'll be out there, until then let's continue blue
sky development" to "release cycle driven development" ... there's more to this than a single
simple answer.

> and have not seen any "lack of testers" until the .0 version is out

as you note it has to be readily available. a lot of people were waiting for that fated
"official release. the speed of communication has increased and development has moved forward
since the release, which sort of shows that despite your expectations, this move has yielded
dividends.

> 3rd party ISV's porting to KDE4 from KDE3, may be they
> do follow the user base,

it's not about them following the user base, it's about them being able to easily source an
API that they know won't shift beneath them. until 4.0, there were API and ABI changes being
made, and that's really, really hard to deal with as an application developer (trust me, this
drove me *mad* trying to get plasma together during this time> The bugs in the KDE4 rc series
did not need wide spread use
> to find them, just a small amount of attempted usage.

you're referring to a handful of issues with plasma, i assume. since 4.0 many non-obvious
issues have been reported against various components in that release. remember, it's a bit
more than a window manager and a desktop shell. so, no ... many of "the" bugs (whatever that
means, exactly) in the 4.0 release were not found until people started using the various apps
and libraries underneath. visiting bugs.kde.org shows this pretty well.

even in plasma there were several interesting issues as well as feature requests that only
came out since the release and more people using it.

suggesting that because some issues are obvious that there are no non-obvious issues is odd.

> was get the release machinery sorted

that is indeed one of the positive things that happened as a result, yes.

other things that happened: third party devs could more easily start earnest work on porting,
kde devs shifted in thinking from "one day it'll be out there, until then let's continue blue
sky development" to "release cycle driven development" ... there's more to this than a single
simple answer.

> and have not seen any "lack of testers" until the .0 version is out

as you note it has to be readily available. a lot of people were waiting for that fated
"official release. the speed of communication has increased and development has moved forward
since the release, which sort of shows that despite your expectations, this move has yielded
dividends.

> 3rd party ISV's porting to KDE4 from KDE3, may be they
> do follow the user base,

it's not about them following the user base, it's about them being able to easily source an
API that they know won't shift beneath them. until 4.0, there were API and ABI changes being
made, and that's really, really hard to deal with as an application developer (trust me, this
drove me *mad* trying to get plasma together during this time.)

> could be release as production quality 4.0.0 with the
> desktop environment, with new things like plasma given an "alpha"
> or "beta" label.

so your objection is to the semantics, not that there was a release? i just want to understand
what exactly your objections are. personally, i'm more interested in the fact that we're
releasing and releasing regularly now rather than the semantics. call it whatever you want,
really.

i even suggested at one point, only half jokingly, that we release this as KDE4 v0.1 so people
could "get" the idea more clearly that this was the first release of the KDE4 series.

> KDE4 was put out without useful documentation for testers (KDE4

documentation is a long standing problem in free software. there aren't many good and
motivated documentation writers around, but they do exist. unfortunately we make life very
hard for them. (i spoke with some people about this last week at l.c.a actually.)

we ask them to use docbook (or send in unformatted text, oi!), don't provide great
collaboration tools and don't provide a clear way to collaborate with developers or get access
to development snapshots.

this is why, at least in part, documentation often lags. =/

> then how will the motivation be generated to complete this task
> when it's not necessary for release to end users

documentation will always be necessary, and now that there is a release people can use the
odds of docs appearing actually goes up. plasma documentation, for instance, was not written
because there was no release (i'm not kidding, this is exactly what the documenters wanting to
work on it communicated)

yes, this is ass backwards, not unique to KDE and something that those with a real interest in
documentation should try and solve. i already have too much on my plate to try and squeeze
this issue in as well. maybe you or someone like you could put some time and effort into it.
the world will thank you, i'm sure =)

personally, i think wiki style editing with the ability to create offline versions and turn it
into non-html documentation is the holy grail.

> What I saw was premature release and call for bugs,
> rc1 being done despite Plasma being in bits and useless,

once again, plasma is not KDE4. you're completely overemphasizing its role in 4.0.0 and
missing the point and purpose of making a release.

> imminent release, they can't be fixed, as they're not crashers or data
> corruption; so a large proportion of the user base are going to feel
> ignored.

and if i ask you to let us make a release, so we can move past the crashers and data
corruption issues and get to regular bugfixing (such as the above 4.0.1 release) what do you
say? i'd like to pay attention to your needs, but every time i try to it seems it isn't good
enough in some new way. interesting ...

> The X.0.0 smells like a commercial software con,
> "we need the revenue and will do it right after".

neat. so it's a con, and we're like those dirty commercial software people. and it was about
revenue or something. i call foul and raise a red card in your direction. that was way over
the line, and completely out of line with reality.

> FOSS X.0 releases were generally very high
> quality compared to commercial, because of extensive alpha & beta test,
> with many sysadmins trusting beta's enough on production systems.  In

well, in *my* experience of using f/oss since the mid-90s, free software projects used to be
given all the time in the world to release 0.1, 0.1.n, 0.2 ... etc working up to a 1.0 after
*years* of slow efforts. the reason beta's were so high quality was because projects were
small and development times in the years were acceptable. these things have changed almost
purely due to user expectation and the involvement of corporate interests, so welcome to your
brave new world.

personally, i see 4.0.0 as "KDE4 0.1". it's the same old "early release, will eat your
children, 0.2 coming soon" party.

> run I do not see a load of Ubuntu GNOME users, feeling that
> they should give KDE4 a spin, running into bad bugs and confusion,
> dismiss it as badly thought out "eyecandy" and thus miss out on
> KDE for ever more is going to help anyone.

firstly, converting GNOME users is not our main priority.

secondly, it would help if people such as yourself might actually try some of the apps like
gwenview and okular that really shine and shift the conversation to those positive things
rather than making my every morning the "rake plasma over the coals again and blame the world
on aseigo's project" show. it's discouraging for me, but worse it's totally framing the wrong
discussion.

thirdly, i have lots of faith in the coming releases. 4.1 is already shaping up nicely and
we're just one month into devel. those who don't get the concept now will later. on the topic
of GNOME, you could say the same about both their 1.0 and 2.0 releases.

*hugs*, aseigo.

The release makes sense if you think from the standpoint of developers as 
opposed to users.

Think process as opposed to consumption.

Derek