@eyedeekay
&eche|on
&kytv
&zzz
+R4SAS
+RN
+RN_
+T3s|4
+acetone
+dr|z3d
+hk
+not_bob
+orignal
+postman
+weko
+wodencafe
An0nm0n
Arch
Coucou
Danny
DeltaOreo
FreefallHeavens_
Irc2PGuest21083
Irc2PGuest53061
Irc2PGuest55294
Irc2PGuest56732
Irc2PGuest86132
Nausicaa
Onn4l7h
Onn4|7h
Over
Sisyphus
Sleepy
SoniEx2
T3s|4_
aargh2
anon4
b3t4f4c3__
bak83_
boonst
carried6590
cumlord
dr4wd3
eyedeekay_bnc
hagen_
onon_1
plap
poriori_
profetikla
r3med1tz-
rapidash
shiver_
solidx66
u5657
uop23ip
w8rabbit
x74a6
eyedeekay
There PPA copy-overs are all queued up should just be a few minutes
eyedeekay
re: checkins, Easy-Install embedded routers are always built from the latest i2p-* tag so you shouldn't hurt anything by checking in
eyedeekay
But take your time
zzz
eyedeekay, I'm copying the last two for you (kinetic and noble)
zzz
eyedeekay, where did all the changes go I put on the deb home page?
zzz
oh you made it a symlink
zzz
fixed
zzz
don't know why you skipped kinetic and noble but fyi launchpad will glitch if you try to copy too much at once
zzz
you have to check that they all made it
zzz
eyedeekay, noble done, kinetic is eol
zzz
eyedeekay, noble copied to deb; also copied to trixie
zzz
eyedeekay, http is still redirecting to https, you need to fix it before we can roll this out
zzz
eyedeekay, I updated the version info at the bottom manually, I guess that's fine for now if you can't get to doing it automatically
eyedeekay
I'll fix the https and the automation as soon as I finish making coffee
zzz
fyi I can login to gitlab clearnet; git.idk.i2p still returns 422 and there's a raft of google results for 'gitlab 422'
zzz
clearnet gitlab is totally borked though, all the internal links are to git.idk.i2p
zzz
and https to i2p doesnt work anymore
eyedeekay
What do you mean https to I2P doesn't work anymore?
zzz
I mean git.idk.i2p doesn't work any more
eyedeekay
I'm logged in to it looking at your user account as we speak
zzz
hmph
zzz
ok will investigate further
zzz
i think the http-over-i2p 422 problem is related to cookies or something but I tried several times clearing them, no luck
eyedeekay
I'm going to try and see if the login is something on my end but from here it looks like HTTPS is working
eyedeekay
re: the internal links yeah one of my rewriters got broken when I was doing all those gitlab tweaks earlier this month, I'll have to add that back to the reverse proxy config
zzz
for me the https always times out but it could be my problem because I'm cooking with all my keepalive changes
eyedeekay
I had the https side of stats time out for about an hour last night, don't know if that could be related
zzz
maybe
zzz
I've seen some strange hangs also
eyedeekay
clearnet http on deb repo should work now
zzz
lets see
zzz
yup
eyedeekay
do it with curl and not a browser, browser will try and make you use HTTPS still
zzz
eepget
zzz
are we ready to tell people? maybe here first?
eyedeekay
Yeah the word is getting out anyway
zzz
ok let's get a couple people here to try it and report before we put it on blast
eyedeekay
Somebody posted this this morning: old.reddit.com/r/i2p/comments/18o96ew/updating_i2p_on_debian_vps_mini_guide so apparently got it to work somewhat independently
zzz
some process discussion for the next release:
zzz
eyedeekay, I assume you'll continue being the release manager?
eyedeekay
Yes I will
zzz
great, and I assume we want to continue as we have been, where big changes get MR'ed and review requested?
eyedeekay
Yes please, IMO that gives the best opportunity at contextualizing a big change
zzz
agreed
zzz
so you'll set and publish the schedule for all the milestones for this cycle
eyedeekay
Yeah planning on making the forum post today. How do you feel about early-April for the next release?
zzz
3 months would be late March ofc. Doesn't matter to me, haven't looked at the calendar to see when easter is
zzz
next q
zzz
I'll have 3 big changes - i2ptunnel, snark, susimail - each 1-2k lines of diff
zzz
after review, I'll probably check in each in smaller chunks
zzz
but for the reviews, would you like to see each in one big MR, or in a bunch of smaller MRs?
eyedeekay
March 31 is Easter this year
zzz
so weekend before or after probably best, your choice
eyedeekay
A guideline would be that if it touches at least 3 files and extends beyond say, 100 lines of changes I'd like to see an MR for it
eyedeekay
Let's do weekend after that's where I was leaning before
zzz
ok re: schedule, ok re: MR
zzz
that would be a 16 week cycle
zzz
re: my big changes, the question is if it's a 1500 line change, would you want to review that in one MR or split into several MRs?
eyedeekay
Hm. Several MR's would be better if they can be tested separately, for example, if merging part of a feature is viable without merging the whole thing
eyedeekay
But only if we track all the MR's on the same issue
eyedeekay
But if it all works together to accomplish the desired change and makes more sense that way, put them all in the same MR 1500 lines or not
eyedeekay
So if you think we'll be more successful if they're broken into smaller MR's, file an issue for each feature and when you make MR's tag that issue but don't close it with the MR
zzz
it's not really different features
zzz
more like a stack of changes, like they do on linux
zzz
which is more work but might make it easier to review, or maybe not
zzz
but it's definitely an "all works together" scenario
eyedeekay
If it's an all-works-together type of thing then my suggestion is that the whole changeset might be better so we can see all the moving parts at once, so to speak
zzz
ok
zzz
I think some of the trivial parts of the susimail changes can go in separately and w/o review, that will make the review easier
zzz
eyedeekay, please update website debian page instructions and remove the note I put in there
eyedeekay
Will do
zzz
thanks!
zzz
eyedeekay, any reason you changed ppa release name from x.x.x-xubuntu1 to x.x.x-x~ubuntu1 ?
zzz
*naming scheme
eyedeekay
I did it when I was trying to figure out why launchpad wouldn't take my upload and I had to bump the build number, one thing I tried was to change the naming scheme
zzz
I don't understand the scheme rules
zzz
I just try hard not to bork it
zzz
I don't recommend random change and hope
eyedeekay
It can be changed back, it wasn't the real underlying issue, but when that upload built and the others didn't I left it in place
zzz
was never a fan of squiggles, that was a kytv thing
zzz
if it works I guess it doesn't matter
zzz
but yes you have to bump something after you screw up
zzz
as you discovered
eyedeekay
Oh I knew that for years
zzz
yup
zzz
eyedeekay, I'm checking in a bunch of little stuff this morning, I'd like to bump to -1 when I'm done so people don't think they're on the release, ok?
eyedeekay
Yeah go ahead
zzz
thx :)
eyedeekay
thanks for letting me know :)
zzz
sure, I like to bump to -1 right away so we don't panic if there's any problems
eyedeekay
Good tip, makes sense
zzz
I'm down under 100 local files changed, lets see if I can find some more to push from my disk onto yours
zzz
down to 92, still looking
zzz
should I check in my dog slow vanity b32 generator or do we want to discourage that kind of thing?
zzz
ok thats it for -1
eyedeekay
I don't care about bruteforcing words in b32's if somebody wants it they can do it outside our code
eyedeekay
If it's slow then there's not much point anyway
eyedeekay
Are you using i2pgit.org or git.i2p.net to access gitlab?
zzz
I usually use i2p but as a workaround when I need to log in I'm using clearnet
eyedeekay
I am trying to get the rewrite filter right so it stops trying to send you to git.idk.i2p when you click on an issue
eyedeekay
I do that by rewriting everything on the fly with nginx since gitlab does not like that a bit
eyedeekay
being on multiple hostnames