andi- has quit [Remote host closed the connection]
andi- has joined #spectrum
tilpner has quit [Quit: tilpner]
cole-h has quit [Quit: Goodbye]
tilpner has joined #spectrum
<qyliss>
When will I learn that procrastinating on writing a status update just means that the status update I eventually end up writing just gets longer and longer and takes longer to write
<MichaelRaskin>
Maybe you need to do market segmentation? Write «not really full status updates» like fifty words with no editing beyond spell-checking while you procrastinate, then a) the Real Update can just skip a ton of details, b) it doesn't matter whether Real Update happens
<qyliss>
Yeah, something like that is probably a good idea.
<qyliss>
I do like the full status updates, but they are extremely time consuming to produce
<qyliss>
I just hit 1000 words in the one I'm working on right now, and I'm maybe half done?
<Profpatsch>
qyliss: I can recommend logging a few sentences each day (and also designs and things) and then just condensing for the status updates.
<MichaelRaskin>
Maybe indeed you need to have a form of updates shorter than an introduction to a CS paper
<qyliss>
yeah
<qyliss>
last status update was similarly long, apparently
<MichaelRaskin>
Stop writing paper introductions and switch to abstracts. Or IRC messages even.
<qyliss>
I liked colemickens' idea of a "This Week In" sort of thing
<qyliss>
Even if some weeks there wouldn't be all that much to say, I'm sure there'd always be something.
<qyliss>
The question I always have about this sort of stuff, though, is is anybody going to read it?
<qyliss>
The occasional high-level status update, I feel, has a much larger audience than a weekly report from the trenches.
<qyliss>
But of course that doesn't matter if I never write the status updates :P
<colemickens>
The curator of https://weekly.nixos.org/ would probably link to it and could give you an idea of the traffic they get, maybe?
tg has quit [Quit: Leaving]
<colemickens>
I think weekly reports end up potentially getting on HN/lobsters, end up driving traffic to high level updates.
<colemickens>
Or at least that's been my experience with Matrix, TWIM.
<qyliss>
Oh really? I'd have expected status updates to be more their sort of thing. Interesting.
<colemickens>
And weekly.nixos.org exposes me to a lot I miss even though I spend a lot of time in IRC :)
<qyliss>
(Although that hasn't happened to any status updates of mine yet, that I know of)
<MichaelRaskin>
qyliss: please take one step back and look at waht exactly you are doing for SpectrumOS right now. However you package the updates, it is foundational work that can only be described for people interested in the inner side of things.
<colemickens>
hm maybe its more of a subreddit thing
<qyliss>
That's true.
<qyliss>
But I still think there's a difference between people interested broadly in whether I'm making progress vs what exactly I did this week.
<qyliss>
maybe I'm wrong though
<MichaelRaskin>
I remember you being worried about people not being able to judge progress properly
<MichaelRaskin>
It is easier to feel «hmmm, it's indeed a nice week of work» than cutting into enough details to understand why something took two months
<colemickens>
"these weeks in" along with a link to "project status" at top
<qyliss>
Yeah, that's true.
<colemickens>
and "project status" links to a github issue. seems ideal
<MichaelRaskin>
Single-sentence dailies could be nice, but then there is too much variance risk
<qyliss>
Yeah
<MichaelRaskin>
Like, get stuck on hunting a silly upstream bug, everyone is let down
<qyliss>
(I don't work every day either :P)
puck has quit [Ping timeout: 256 seconds]
<MichaelRaskin>
I assumed so, but it is your thing to confirm or deny, not mine to guess in public
<qyliss>
that's very considerate of you
<MichaelRaskin>
You are definitely doing impressive work overall, even if the order of milestones negotiated, the order of milestones you would now prefer to have negotiated and the order of milestones I hoped you would set are three different orders.
<qyliss>
lol
<qyliss>
thank you :)
<MichaelRaskin>
Even if — in the sense that disliking some things in Wayland makes me put off looking at the details of Wayland, thus making me less able to fully appreciate the Wayland-specific parts of effort
<Profpatsch>
The way I see it this is a R&D project, high risk high reward. It’s expected to not consistently produce visible output.
<Profpatsch>
(though that means it’s even more important to have a consistent log, even if that’s about internal work instead of visible output)
<MichaelRaskin>
Re: risk — I have not seen the list of milestones, but what milestones I have seen mentioned look pretty amenable to 80/20 implementation. It's another story that qyliss goes 120/200 on them.
<pie_[bnc]>
XD
puck has joined #spectrum
cole-h has joined #spectrum
spacekookie has quit [Quit: **agressive swooshing**]