sb0 changed the topic of #m-labs to: ARTIQ, Migen, MiSoC, Mixxeo & other M-Labs projects :: fka #milkymist :: Logs http://irclog.whitequark.org/m-labs
klickverbot has joined #m-labs
mumptai has quit [Remote host closed the connection]
<cr1901_modern>
rjo: The new simulator is awesome. However, being able to simulate mixed code bases is no longer possible. I recall talking to sb0 about adding verilog support using IPC, but as I recall it's not trivial to do
sandeepkr has joined #m-labs
sb0 has quit [Quit: Lost terminal]
sb0___ has joined #m-labs
<sb0___>
whitequark: can the int redefinition be made a bit lighter?
<rjo>
whitequark: about the doc upload failure: no idea what the problem is exactly. but it may (or may not) be triggered by the new directory (_sources).
<whitequark>
i think i'm doing tree removal wrong.
<rjo>
i also had to manually update the logo a while ago. either the doc build or the file syncing did not notice the logo change.
<whitequark>
it should be done in reverse depth first order, but i just do it however walktree gives me the files
<whitequark>
and it seems like my assumptions about walktree are incorrect.
<whitequark>
bb-m-labs: force build --props=package=libusb conda-win64
<bb-m-labs>
build forced [ETA 2m57s]
<bb-m-labs>
I'll give a shout when the build finishes
<rjo>
yes. i think it was actually on the build slaves. maybe sphinx is dumb.
<whitequark>
if you have a network failure, then what happens is all your processes are stuck in iowait
<rjo>
sure. pick your poison.
<whitequark>
plain scp would work here, but i tried avoiding shelling out.
<whitequark>
since, well, it's also fragile, but less so.
<whitequark>
i actually think it's enough to sort directories in reverse alphabetical order for this.
<whitequark>
ok. today was not meant for me to make mingw work.
<whitequark>
it consistently fails to download archives from the buildbot, which is very baffling, because doing that from my home works
<whitequark>
so it's not sourceforce's broken mirror selection or something, assuming it treats all HK IPs the same
<whitequark>
and I cannot upload the archived installation of mingw that I have because this ADSL has such bad QoS that uploading anything over 1MB timeouts; no acks come back
<GitHub138>
conda-recipes/master c9cae50 whitequark: libssh2: install DLL in the proper place (Library/bin).
<GitHub138>
conda-recipes/master 9fe47ac whitequark: libftdi: add Windows support.
<whitequark>
it seems like what happens is that it has a shared medium (the copper wire) and the outgoing packets hog that medium
<whitequark>
well, now that I recall how ADSL works, that's technically incorrect, as it uses two bands. hm. no idea why it breaks, then.
<rjo>
sidestory: yesterday on the train i was "borrowing" somebody else's MAC address. and obviously TCP sessions got reset all the time. but mosh worked flawlessly.
<whitequark>
they really need to implement non-text tunneling in mosh (it's on the roadmap)
<whitequark>
I've seen it many times that on airport wifi they didn't even bother to gate udp with the nag screen
<whitequark>
bb-m-labs: force build --props=package=libgit2 conda-all
<bb-m-labs>
build forced [ETA 4m32s]
<bb-m-labs>
I'll give a shout when the build finishes
<whitequark>
bb-m-labs: force build --props=package=libssh2 conda-all
<bb-m-labs>
build forced [ETA 4m32s]
<bb-m-labs>
I'll give a shout when the build finishes