so, the binutild-gdb folder I found is empty.

but now the internet works better, i'm trying out doing a full clone.

it is presently stalled.

843

it looks like the git:// protocol doesn't support partial clone, but I could probably do a partial clone from a github mirror of sourceware.org which are likely common

845

i'm trying a shallow clone again.  I guess i'll also look for github

it could make sense to work with the bitcoin code too, but gdb source is an important thing to have.

both clones are stalled, maybe due to long tcp delay kernel config somewhere.

oooh shallow clone unstalled.  847a.

found a binutils-gdb mirror on github https://github.com/bminor/binutils-gdb

848

i've started a partial clone and it's trying to find io time.  there it goes.  I think this one might work, partial clone is kinda smart.  we'll see.

file recovery still says it's in progress.

partial clone is asking for 132k objects.  shallow clone is asking for 36k.  partial clone is stalled at 44%.  shallow clone stalled at 10%.

maybe i'll look for that kernel config.

850a

hrm first resource I found says the upper retry time is harcoded.  but that doesn't mean the falloff curve is ... but it might be.

852

the partial clone finished its important first retrieval and is buzzing away downloading sparse object sets in retryable manners.

the shallow clone is at 87%

I also have a full clone going which thankfully has not given up.  maybe i'll pause partial clone.

maybe not.  they stall so much it seems unlikely they'd compete.

i'll look for tcp retry settings more.

some inhibition :/

856.  oh hey shallow clone finished!  hmm some kind of io contention trying to update the files.

hey everything unstalled all at once, file io and network both.

now they all stalled again.

it's 857.