Thursday, 2022-05-12

excited-mango[m]<lkcl> "and if it's instructions..." <- Can you send the link to the wiki?02:29
excited-mango[m]programmerjake: Oh I didn't realize that was the wiki, but in hindsight it makes sense. Has tons of useful information02:33
lkclexcited-mango[m], which is then a git repository, which is then backed up like everything else.06:38
lkclhttps://git.libre-soc.org/?p=libreriscv.git;a=summary06:38
ghostmansdprogrammerjake, it seems I managed to send the mail properly this time thanks to your tips and help!08:33
ghostmansdhttps://sourceware.org/pipermail/binutils/2022-May/thread.html08:33
ghostmansdhere you can see [PATCH v2 ...] lines08:33
programmerjakeyay!08:34
ghostmansdI've updated the "about us" page to reflect the recent changes, and then I recalled we had other tasks...09:08
ghostmansdhttps://libre-soc.org/ghostmansd/ lists there we several in progress... lkcl, programmerjake, what's their status? Did anybody work on these?09:09
ghostmansdIt'd be great if these can be visited by you and re-scheduled, including the budget. binutils is a huge task, it'd be great to have something besides it.09:10
programmerjakeafaict: #671 has had some work done on it, the other two have had no work done09:18
programmerjakethe work done on #671 is mostly just conversion to XLEN whenever we were writing new pseudo-code or whenever we were editing pseudo-code.09:19
programmerjakee.g. https://git.libre-soc.org/?p=openpower-isa.git;a=blob;f=openpower/isa/bitmanip.mdwn;h=d1521a762d1302f8a3c9502760d76fda8f711044;hb=ff0fa168b62d4c182592437ae6f54fa764dbe8a709:21
programmerjakei'll let lkcl to the scheduling and budgeting09:24
programmerjakedo*09:24
lkclghostmansd, things can be broken down into sub-tasks09:44
lkcli've been meaning to discuss that because there's (a) quite a lot of budget available and (b) a lot to get done09:45
lkclwe can almost certainly declare the auto-generation phase as "100% completed" for example10:19
lkclwhich (surprise) gets its own bugreport, and budget, then gets marked as "RESOLVED"10:19
lkclblah blah10:19
ghostmansd[m]Well, I wouldn't say that's 100% completed, since I still have plans to generate the insn fields as well...10:26
ghostmansd[m]But I get what you mean, yeah10:27
ghostmansd[m]Ok, I'll think of how to raise this task, then10:27
ghostmansd[m]Then, I guess, I should dedicate part of budget for task 550 to this new task, right? How is it done, should I simply substitute from the field?10:28
ghostmansd[m]Or do we have another task which should be reflected this way?10:30
lkclyes. just create a new task, take some off the "budget (EUR) for this task, excluding subtasks' budget:" field and add it to the new task instead10:49
ghostmansd[m]Ok, got it, thank you!11:15
ghostmansd[m]Could you, please, re-visit tasks marked as "in progress" at https://libre-soc.org/ghostmansd?11:16
ghostmansd[m]Should I take a budget for tests from the xlen task, for example?11:16
ghostmansd[m]Ah, I see, it already has it, I need to update the page11:18
ghostmansd[m]Ok, then, I guess, only the parent task needs an update regarding what is done/what needs to be done11:18
lkclghostmansd[m], you can edit the page yourself, it's part of the wiki :)11:20
ghostmansd[m]yehyeh, that's what I wrote, I need to update the page :-)11:20
lkclexcellent you worked it out https://git.libre-soc.org/?p=libreriscv.git;a=blob;f=ghostmansd.mdwn;hb=HEAD11:21
lkcl:)11:21
lkcluse that page to help yourself keep track, ok?11:21
lkclthere's an auto-generated version by running the budget-sync tool11:21
lkclhttps://git.libre-soc.org/?p=utils.git;a=blob;f=README.txt;hb=HEAD11:22
ghostmansdlkcl, I wasn't aware of this tool :-)11:47
ghostmansdI'm updating budget-sync-config.toml with my email and info11:47
ghostmansdlkcl, I'll need your help with budget-sync, because, once I've updated my email and profile, I have some errors with budget-sync tool. Part of them went away once I updated tasks so that RFPs are assigned to correct person (ghostmansd), but others don't.12:39
ghostmansdAlso, I cannot push changes to budget-sync-config.toml, with `fatal: remote error: access denied or repository not exported: /utils.git`.12:40
ghostmansdFWIW, the changes are, I added "[people."ghostmansd"]" section and swapped it with the legacy, ping me for patch12:41
ghostmansdPerhaps these questions should be addressed to programmerjake, considering how `git log` looks like :-)12:47
ghostmansdOK, seems there are only three issues left.12:53
ghostmansdhttps://pastebin.com/J8cmfFbz12:53
ghostmansdI'm not sure whether it is related to the recent changes directly, or these were present above.12:53
ghostmansdAnyway, ping me for patch to keep only these errors, or update the access for a repo.12:54
lkclyeah it's a global tool so those are not your problem :)14:28
lkclgimme 1sec to add you...14:28
lkcldone you should now be able to push14:29
ghostmansdlkcl, thanks, pushed :-)14:55
ghostmansdOK, so far only #48 and #191 show some errors14:58
*** ghostmansd <ghostmansd!~kvirc@broadband-188-32-220-29.ip.moscow.rt.ru> has left #libre-soc15:00
ghostmansdlkcl, programmerjake, this budget-sync is a lifesaver15:03
ghostmansdit's somewhat strange that fractional values must be written as "0.1" instead of 0.115:06
lkclyes, #48 and #191 are going to stay that way.15:23
lkclentire project of this scale, its finances would be completely unmanageable without budget-sync15:23
lkclghostmansd, if you're closing bugreports don't close ones that you could reasonably be paid for! https://bugs.libre-soc.org/show_bug.cgi?id=67917:03
lkclbut if you (and everyone) have received EUR for then yes17:04
ghostmansdafter so much time passed, it's difficult to even remember what this was about :-)17:04
ghostmansdfor 684, I think that one should get its budget, because that needed more work than just "fix something in scope of another task, aka drive-by fix"17:05
ghostmansdI'll re-check 67917:05
ghostmansdah, I think 679 can be considered as drive-by fix17:06
ghostmansdI'm trying to clean the task list since I found sync-budget script17:07
ghostmansdsome tasks were really drive-by fixes, so I didn't bother about budget there17:08
ghostmansdnot the case with 684 I think :-)17:08
lkclyehyeh17:08
lkclbtw you _can_ use the updater-program but for god's sake be careful17:09
lkclit's not documented but i think at least has a --help option?17:09
lkclyou can use it to automatically add the "paid date" and "submitted date" to the TOML17:09
lkcland can give it a *list* of bugs that you want to add those dates to17:09
* lkcl currently using google maps streetview to "drive" totally down the wrong side of a street...17:11
tplatenI did a whitespace cleanup in nmigen-boards/nmigen_boards/orangecrab_r0_2.py, now creating an account on gitlab17:28
tplatenI corrected the pins for subsignal ba18:28
tplatenAt a later time I want to add my BeagleWire FPGA Cape to nmigen-boards, I have once used nmigen to export a single sr-latch to verilog, then run the old yosys-arachnepnr toolchain that comes with the BeagleWireSoftware18:42
programmerjake<ghostmansd> "it's somewhat strange that..." <- that's because python treats 0.1 as binary floating-point, causing rounding. rounding is a big no-no when dealing with money, so I made a Money class that does the arithmetic correctly.18:43
ghostmansd[m]But you get the stuff like "ghostmansd=0.1" as string anyway19:08
ghostmansd[m]Why force anyone to wrap it into "ghostmansd="0.1""?19:09
ghostmansd[m]A guess: do you use eval?...19:09
programmerjakebecause the toml parser library and the toml spec treat it as binary floating-point19:10
programmerjakeno, i do not use eval19:10

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!