For the month of February, 2020 I didn't manage to deliver any of my agenda from the last report. With that being said, not all was lost and a point of reporting is to learn your lessons and improve how you work with people moving forward.
At the beginning of the month, Diana Coman provided me valuable feedback and conversation on the 2nd attempt at the Why it's +ev to contribute to TMSR OS which resulted in deciding to move back up a node and write about why TMSR OS came to be. I didn't make meaningful progress on that or anything else on the list.
The first couple weeks of the month, the majority of my productive output was directed at a conference clients asked jfw and myself to help host which took place Feb 13-14th. We generated some JWRD leads, made presentations we can reuse and turn into articles and gained experience hosting a ~20 person, 2 day event.
The later half of the month I took a couple steps back to do some personal reflection on the past few months and further given the occasion I turned 30 last week. This included picking back up an old, resourceful habit of hand written journaling in an effort to examine some of my unexamined bad habits and replace them with more productive habits. A key shift I'm is looking in the accountability mirror at the close of each day and publishing on my weekly Young Hands plan a reflection on some questions to help me grow consistently in the right direction.
Moving to March, the first month of the Roman year and the resumption of military campaigns, my objective is to strike off the items from last months' list.
- bvt's February
- published his awk + posix shell + Ada Vtron in a 3 part series (1, 2, 3)
- Investigated TRB wedge
- bvt's March
- moving onto the regrind of keccak RNG patches for 2.6.32 kernel
- spyked's February
- published Gales build and install report on January 31st
- spyked's March
- TMSR OS Hardware research
- jfw's February
- reported completion of the offline component of Gales Bitcoin Wallet and a successful test transaction was confirmed.
- published on article on Bitcoin transactions and their signing that's building toward a series of publications presenting the code.
- pointed out to musl their move towards an infinite alphabet is misguided and does not have the consensus they may have imagined.
- jfw's March
- finish the Gales Bitcoin Wallet article series and genesis the code, which will include TRB patches to build on Gales Linux and a sendrawtransaction regrind.
- Ask about proposed clearsigning scheme.
- trinque's February/March
- reported to be closing in on another article in his Republican OS series, but it's not exactly clear what that will cover.
- No word from ave1
- lobbes reported he'll need to put down TMSR OS work for the short term as mp-wp delivery and changes on the home and saltmining front are higher priorities.
Last but not least, the proposed clearsigning scheme that was seeded could use the sowing of intelligent discussion.
@Dorion:
Current status of 2.6.32 (not 2.16, btw): I have a few trees that I'll have to materialize into vpatches:
- 2.6.32.71 genesis. The question to this is should it be proper genesis, or downgrade vpatch from the previous 4.9-genesised kernel. The problem here is that 4.9 genesis is 600 Mb, 2.6.32 genesis is 300 Mb; but the downgrade patch 4.9 to 2.6.32 is 700 Mb. While v.sh seems to work through this mess rather fast (more precisely, flow resolution is fast when compared to pressing to the rest of the process), there is quite some storage space overhead from the downgrade, also vpatching the tree together becomes a bottleneck for the work process. So, so far I am inclined to abandon 4.9 tree and start anew from 2.6.32 genesis.
- 2.6.32-add-kfifo. This tree ports 4.9 kfifo and generator for timeconst.h: in 2.6.32 this was computed using a perl script, which does not work with the perl version in Cuntoo, and anyway we don't want to bring perl to TMSR-OS with the kernel. TODO for this tree: testing that new kfifo work with the old code: the functions present in 2.6.32 are the same, so all should be OK.
- 2.6.32-add-keccak. Regrind with minor changes of the Keccak vpatch. No problem at all expected here.
A tree and vpatch that is still in the TODO is the one that brings the Keccak RNG. I will have to review the API differences between 4.9 and 2.6.32. It seems that 4.9 API is a strict superset of 2.6.32 API, so no problem is expected, but still, more care is needed here.
Also, I expect that this weekend I will still be continuing v.sh series as the result of the latest discussions.
Comment by bvt — March 5, 2020 @ 20:39
Was there anything further heard from the rEFInd guy or what did his "positive reply" consist in?
Comment by Diana Coman — March 7, 2020 @ 16:05
@bvt :
Thanks, fixed.
Good to hear on the speed at least, but I agree with the inclination.
Agreed. Thanks for the update on the TODOs, sounds good and keep it up!
Comment by Robinson Dorion — March 9, 2020 @ 18:08