IRC channel logs

2018-08-31.log

back to list of logs

***ChanServ sets mode: +o rekado
***Server sets mode: +cnrt
***Server sets mode: +cnrt
***Server sets mode: +cnrt
<rekado>I’ve been thinking that maybe we should do a little more publicity work to let people know about the successes of the past 2 years.
<janneke>yes, for sure
<rekado>what do you think about adding a blog to bootstrappable.org?
<OriansJ>rekado: I prefer slow growth actually
<janneke>one thing that i was hoping/planning for, is to make some noise when the x86 bootstrap has been integrated
<rekado>okay
<rekado>I think this is too cool to keep quiet about it.
<janneke>i agree -- but it's much easier to "shout" when we have something released that people can inspect, and we're *very close*
<rekado>yes, that makes sense.
<janneke>i'm abivalent about blogging; i really like the idea but i'm feeling a bit hesitant about making a commitment
<rekado>oh, that’s okay.
<janneke>i feel much more relaxed after the mes 0.17 release "it is done, the rest is polishing"
<rekado>I wouldn’t want to force a regular announcement schedule upon anyone.
<janneke>so, i'm much more open to the idea than half a year ago, and i think there's much to gain from creating more exposure
<rekado>I just think that it would be good to have some progress reports once in a while for interested outsiders.
<janneke>in a way, my mes release announcements were kind of a blog, maybe :-)
<rekado>indeed :
<rekado>:)
<OriansJ>rekado: We do that when we have releases
<rekado>alright
<janneke>yes...so if i keep those announcements real short and put the stories in a blog, that would work
<janneke>*could
<rekado>would you mind if I collected the announcements from the mailing list and served them up on the website?
<OriansJ>I really don't have time for a blog, plus the amount of information one needs to understand stage0 could fit on the back of a postcard with lots of room to spare.
<OriansJ>rekado: that is fine
<rekado>okay. I might do that at some point.
<janneke>that would be great! you can also collect them (the mes ones) from: http://git.savannah.gnu.org/cgit/mes.git/tree/doc/announce
<OriansJ>akkartik: your issues have been addressed in the recent M2-Planet commit, thank you for letting me know :D
<OriansJ>janneke: I guess the next big question is when do we make mescc-tools a GNU project?
<janneke>OriansJ: i'm wondering...maybe mescc-tools should be part of stage0?
<janneke>and what about m2-planet?
<OriansJ>well stage0 was designed to be the platform specific ports of mescc-tools.
<janneke>i think we should wait a bit, mes is a fresh gnu project
<OriansJ>and M2-Planet could easily be merged into mescc-tools
<janneke>OK
<janneke>if and when this bootstrap idea has caught on @GNU, it will be obvious and easy to make the whole bootstrap GNU
<janneke>but if you want to go for it now, i'm all for it
<OriansJ>janneke: no rush
<janneke>possibly when we have mes.M2 working/finished, that would be a perfect time
<janneke>..but what do i know?
<OriansJ>janneke: clearly a great deal about bootstrapping and writing a C compiler and a Lisp interpreter ;-D
<rain1>you could take a break for now
<rain1>you've been working very hard
<rain1>and this current result is awesome
<OriansJ>rain1: well a vacation is planned, during which time I'll be completely unplugged
<OriansJ>Although I must admit, I have some ambitious goals for M2-Planet v1.1.0
<OriansJ>Multi-arch being the big one
<janneke>i found #glibc, and got some useful information
<OriansJ>good
<janneke>possibly the most obvious thing: what i'm trying to do is "unusual" and possibly has never been done
<janneke>building a new glibc, using a gcc that was built with a very old glibc
<janneke>i didn't imagine the issue could be there
<janneke>the most certain recipe is to just work our way up the versions, building contemporary versions of binutils, glibc, and gcc
<janneke>in any case, the jump that i'm attempting is probably too far out
<janneke>but i didn't imagine that using older gcc's could help...
<OriansJ>I'm currently attempting to figure out how >| followed by >> results in losing the >|
<janneke>rain1: http://paste.debian.net/1040200/
<OriansJ>what the heck am I missing https://paste.debian.net/1040202/
<janneke>OriansJ: what is >|, how is: "> seed.M1" not good enough?
<janneke>anyway, why not add a: "cat seed.M1" in between to inspect what's going on?
<OriansJ>>| is the same as > but will still overwrite files if you have set -o noclobber
<janneke>ah
<janneke>hmm
<OriansJ>in between, it is exactly what one would expect
<OriansJ>yet somehow writing to standard out, the fopen is reseting the file pointer in seed.M1
<janneke>and doing really silly: vm --rom > seed-2.M1; cat seed.M1 seed-2.M1 > seed-both.M1
<OriansJ>works perfectly
<janneke>or >| of course
<janneke>then use that workaround, add a comment about being flabbergasted and wait until the lymbic system (or the internet) presents the solution...
*janneke does see how unsatisfactory this probably feels
<OriansJ>unless you know a simple way to prefix a block of text ontop of a text file
<OriansJ>I could use sponge to do it though
<OriansJ> https://paste.debian.net/1040204/
<OriansJ> https://paste.debian.net/1040206/
<OriansJ>then works
<janneke>i have no idea what's going on -- looks like magic
<janneke>maybe take a few hours off and have a fresh look tomorrow?