Re: Half-year review of Stskeeps's work as distmaster
 

For my own part, I believe you've been doing an excellent job. Nokia's thrown us a lot of curveballs over the two years since Maemo Reconstructed was started and you have to roll with them as best you can. The TI drivers ended up taking much longer than expected, Mer ended up being a much larger undertaking than expected, and things were all, generally, more difficult than any of us would've believed going into it. We all learn as we go along, and things rarely meet expectations in software (why does everything in Maemo always come back to managing expectations? . . .), but I've see some amazing progress so far.

Hell, there's an argument to be made for just how fantastic a job you've done given that MeeGo seems to fulfill many of the goals of Maemo Reconstructed, Mer and the distmaster position (whether you or any of us had a strong influence on that, I have no idea, though). ;)

So, to summarize, thanks for digging in doing all of the dirty work and doing your damndest to help encourage and coach other people to help to it, too. It's a tough job shifting the direction of a ship as big as Nokia's, but I think you've done admirably so far.

As for places to improve:

It's become especially tough to follow all of the information flowing around the Maemo Community these days, and, as always, communication is still key. I don't think a few self-congratulatory blog posts about progress would go amiss. As we can already see from some responses here, perception is a large factor and it will always suffer if things are visible enough (trust me, I know, I've been playing a facilitation role for nearly 2 years ;)). Some summarizing of both major and minor progress would make it easier for everybody to track and to appreciate that activity is, in fact, taking place, whether or not they set out to find it.

Proper management of expectations is always important. It's something that we've all tripped on with Mer, and something that I don't think will be an issue in the future, as, well, now we know. ;)

Re: Half-year review of Stskeeps's work as distmaster
 

Gladly - keep in mind that I work 30 hours a week (and for 3 months of the time, 20), so I can't be everywhere or do everything, but here it goes (in no particular order, just as I find them):

* I've helped to port Maemo5 to Beagleboard and Zoom2 (http://omappedia.org/wiki/Maemo_on_OMAP_Project)
* Helped push theme templates to the open and written guides how to deal with their packaging. This is shown in the themes people are developing today and how since Theme maker dev was kinda stalled.
* Ported Imagemagick to extras-devel related to that
* Helping to bootstrap the community SSU for Diablo (mostly spearheaded by lma though)
* Bootmenu on N900 + getting support for it into PR1.1
* Mer on N900 proof of concept (some might say this was a detour but up front: it was an avenue for more people to contribute to the core system as a lot of new coders came in with N900 launch who didn't have a N8x0)
* Mer development in general.
* Maemo Summit (two presentations)
* Bring out the toolchain for the N8x0 initfs
* Create a way for applications to be build against GLESv2+EGL apis without having to be built against the closed source libraries
* Mapping openness in Fremantle PR1.1 and Diablo
* Enabling OS porters to N900 how to use battery charging, RTC, etc
* Proposal for project management on maemo.org
* Interview: TheLinuxLinkTechShow 11-11-09
* Presenting Maemo-on-OMAP and Mer on TI eTechDays online conference to attract more people to the platform
* Loads of time getting the TI MBX drivers to a semi-usable state.
* How to approach a open sourcing queue in a way that makes it productive
* Dual booting another OS on N900 (Mer) and Making flashable rootfs's for N900
* Mer^2 prototype work, mostly OBS work - an experiment in how close to a Fremantle system you could get and recompile everything in Fremantle.
* The licensing change requests queue - a gateway to have better open sourcing process and evaluating these
* Work on making a OBS implementation of autobuilder and building towards Fremantle in this
* Fighting for public R&D in MeeGo
* The vendor social contract
* Help push out fiasco-gen in PR1.1 SDK so we can build own FIASCO images.
* MeeGo ARM bringup on N810 (thanks to the great kernel work by others)
* And a lot of random talk, work and poking internally - not very transparent, but I think it has helped.
* And a lot of talk and pointing in the right direction for people, which isn't that transparent but I've hoped helped people..

Re: Half-year review of Stskeeps's work as distmaster
 

I shouldn't have been so flippant earlier, sorry.

I'm mostly impressed with the work of Stskeeps but then, I make a point of following what he does. As has been mentioned, most of that effort won't be immediately obvious to the community at large.

So I'd have to agree with GeneralAntilles that the one area of improvement would be in communication. Make strong use of the available channels, including twitter. Yes, there have been some good updates, especially here, but I think the visibility could have been just a bit higher.

Anyway thanks Carsten and I look forward to your continued efforts!

Re: Half-year review of Stskeeps's work as distmaster
 

Not sure how more clear I could had made it that it was subjective opinion by an end-user. If those opinions are not requested then just ignore my feedback. I'm also pretty sure that Stkeeps can speak for himself so not quite sure why you felt the need to add your 2 cents.

Your comment "perhaps it'd be best to start out questing for information and painting is a picture of your own perceptions, rather than immediately jumping into an analysis based on incomplete information." is also rather contradictory as I said I was an end-user reading this forum and this is the channel he asked for feedback in. Not sure how much I would have to dig and what channels I need to go to in order to meet your expected level of analysis. I read the forum and quite a few about Mer.

I find your subsequent comment rather offensive with the above in mind: "If you never get out into the wide world of Maemo, it seems awfully presumptuous of you to set out to grade someone performance based on such a thin slice of the community. Especially for the distmaster, whose work rarely directly affects self-described "end-users" (primarily because it is both platform-focused and long-term)."

This was the channel he asked for feedback in. I don't utilize any other channels. If there was a pre-requisite for 'reviewing' him - as he requested - I wish it was called out in the opening post.

So please ignore my 'review' and replace it with:
"Great work, keep it up, no complaints! :D" as I don't want to be presumptuous and review something I have no idea about.

Don't worry, no more comments from me in this thread.

Re: Half-year review of Stskeeps's work as distmaster
 

Just quickly answering this one - Nokia only pushed TI to do this, but the absense of a driver is not a fault of Nokias, it was not their driver to give.

I was quite surprised to hear that TI announced they had a working driver for N8x0 at Maemo Summit and started planning for this arrival.

Now, imagine my surprise when it turns out they (TI) did not actually try it on anything but their development platforms and not with the N8x0 devices, in fact, had not been developed wit N8x0 nor did they have any to test on?

That screws up things quite a bit when a small task of integration turns into a driver porting task. As it was a high priority of the community, I spent a -lot- of nights working on this, with some kind of workable result. Now, there is still some work to be done in that area, but it is out of my expertise. Maybe it returns now with the MeeGo for N8x0 port.

Re: Half-year review of Stskeeps's work as distmaster
 

Yes, this never really came to fruition (yet) except in very few cases where I've pointed people in the right direction on where to do some fixes. We had a bit of a inactive council last session so there really never came any wishes for things for me to work on.

Agreed, but this also (as stated) requires some management to prioritize.

Re: Half-year review of Stskeeps's work as distmaster
 

I would still like to see more discussion and explanation for how Mer turned from a crock of gold into just a crock.

Re: Half-year review of Stskeeps's work as distmaster
 

Yes, I'm getting to your post - don't worry. I will answer posts but not immediately, just woke up an hour ago ;)

Re: Half-year review of Stskeeps's work as distmaster
 

In the meantime, I have a bit more to say.

There were days, weeks, and months when I checked the Mer project constantly for new developments. I even installed it a number of times. It looked great.

Then you seemed to me to say that the only real thing holding Mer back was a couple drivers that needed to be released in non proprietary format, or something like that. There ensued one heck of a long period of waiting, in which it was said time and again that we were waiting for the two drivers or whatever.

Then finally, finally, the two drivers arived. And what followed was -- nothing.

That's the last progress I remember.

What ever happened to the Mer that was waiting for the two drivers?

My impression is that a usable version of Mer never came.

That means that the long period of waiting for the two drivers was what? Were the two drivers just a story?

Was I wrong to be under the impression that a usable version of Mer was almost ready to go?

Re: Half-year review of Stskeeps's work as distmaster
 

As Stskeeps as already mentioned in the thread, TI severely misrepresented the state of the drivers to be released. Many months worth of volunteer community effort went into getting them into a semi-usable state. Then a series of game-changers came along (culminating in the MeeGo chaos) that ate up a lot of time and required shifts in the Mer development direction.