@Di4na@hachyderm.io avatar

Di4na

@Di4na@hachyderm.io

SRE. Elixir Dev. Learner in Resiliency. French.
All Opinions are my own. And i have a lot.

Co-Founder and President Haruspex.dev

dom. He/him.

Blog: Softwaremaxims.com

This profile is from a federated server and may be incomplete. Browse more on the original instance.

thephd, to random
@thephd@pony.social avatar

So, after a full day and a lot of back and forth, a few very important things have come to light. I'll try to summarize them here because I've been posting primarily in the Bad Places® rather than the Good Places™, so that nobody's lost.

  • It looks like somebody from inside the Rust Project, but not with the consensus of all leadership, tried to downgrade my talk (or perhaps have it outright retracted) because they did not like the direction the compile-time work I was doing. (Learned from: https://www.jntrnr.com/why-i-left-rust/)
  • The decision to downgrade my talk, once it was known I was going to talk about Compile-Time Reflection in Rust, came anywhere from 4 to 8 days before I was actually told this past Friday and had to release the blog post stepping away. However, as evidenced by much of the public statements from existing, ex, and now-ex Rust Project members, the decision to unilaterally downgrade was not known to many of them until they read my post.
  • Downgrading the keynote was NEVER voted on like inviting me to do the keynote in the first place.

All in all, this reeks of someone trying to run-around the consensus of the Rust Project because they don't like Shepherd Oasis's or my work (detailed here: https://soasis.org/posts/a-mirror-for-rust-a-plan-for-generic-compile-time-introspection-in-rust/).

I don't know how to handle this going forward. The Rust Project has effectively ultimate commit rights to rustc and all of the projects our work would touch are under the control of the organization that did this. Even if we do the work, they could effectively unwind and undo a lot of our work, or indefinitely block it with an endless slew of "reasonable concerns" from ranking project members who seem to have problems but don't want to communicate them except by taking potshots at the status of my now-gone RustConf talk.

I don't know what to do. I'm pretty lost, it's Sunday, and I have a shitload of things I still need to do, not including this whole trainwreck.

Di4na,
@Di4na@hachyderm.io avatar

@sgrif @barometz @thephd all i will say is that hugs are available of consented for everyone that hurts on this. These situations are never easy for the people used by people actions.

And working to help make it better is also hard, if good, work. Big support to you for this.

Di4na, to random
@Di4na@hachyderm.io avatar

I rarely agree with the conversation of "lack of reciprocity" for FOSS maintainers. Reciprocity itself does not make sense to talk about in this situation imho.

But I do think that there is a double standard. Here is my example for today. Do better corp.

Just. Give your employees frigging security tokens.

https://www.softwaremaxims.com/blog/2fa-community-participation

Di4na,
@Di4na@hachyderm.io avatar

@kurtseifried What is not said in there is that I am replacing my yubikey because my employer is switching SSO, and only allowing FIDO2 on the new one. No more U2F or FIDO.

But still totally allowing push notification 2FA. So I cannot use my safe security token anymore, but the unsafe push notifications can still be used. Oh and no there is no program really to give security tokens.

I feel so safe now! (not)

Di4na,
@Di4na@hachyderm.io avatar

@kurtseifried I do not disagree but

  1. That vendor definitely have good one. It is okta.
  2. See my double standards point. If corporate ask FOSS to do it, they could impose it to their other vendors too...
Di4na,
@Di4na@hachyderm.io avatar

@joshbressers @kurtseifried i think the difference is that opensource actually is "invisible" and cannot push back. So by definition we are always the one that have to adapt.

Same as my answer for my employer was... To buy my own yubikey. I could adapt and could not push back. They could not.

Di4na,
@Di4na@hachyderm.io avatar

@mdwyer i mean your previous employer could have given some as swag for personal use...

Di4na,
@Di4na@hachyderm.io avatar

@stevel if you trust touchid and the ilk. And indeed if you do not close the lid.

Both things i do not (it is not about privacy but it working well with good recall. Too many bad experiences with it)

Di4na, to opensource
@Di4na@hachyderm.io avatar

I keep having to explain to people that building powerful developer tooling (from programming languages to ide plugins) is mainly hampered by the broken economics of the field.

So I wrote down a summary of these economics from my pov, hoping it helps inform the discussion and spurn some actions. Let me know how wrong I am ;)

This is not exactly about , but it is adjacent to it.

https://www.softwaremaxims.com/blog/economics-developer-tools

danluu, to random
@danluu@mastodon.social avatar

I often see people site Brooks's comments that development can't be sped up by adding people and, in fact, adding people slows things down but, in practice, I've generally seen adding people speed things up and, in a 20 year career across multiple fields, I'm not sure I've ever seen an example of adding people slowing things down whether or not the project was late.

Is my experience an outlier here or is Brooks just mostly off base, as in https://www.patreon.com/posts/46629220 and https://danluu.com/essential-complexity/?

Di4na,
@Di4na@hachyderm.io avatar

@arclight @danluu it gets even worse when you realise Boehm basically hallucinated his numbers.

I have seen numerous projects being made later by adding more people, but that is usually a symptom not a cause.

benoit, to random French
@benoit@ruby.social avatar

I was quite impressed with the impact of specifying the order column (when there is limit) in your index in

> An important special case is ORDER BY in combination with LIMIT n: an explicit sort will have to process all the data to identify the first n rows, but if there is an index matching the ORDER BY, the first n rows can be retrieved directly, without scanning the remainder at all.

https://www.postgresql.org/docs/current/indexes-ordering.html

Di4na,
@Di4na@hachyderm.io avatar

@clementd @benoit additional fun one. How many APM providers out there have a query system that works like this and have a forced limit due to pagination.

Spoiler: yes the big ones you think of do.

toplesstopics, to random
@toplesstopics@eldritch.cafe avatar

hastily drawn page once again mocking etc for their policies towards a photo or video of a woman without a shirt (or wearing a strapless dress) cut off at the shoulders as "implied nudity," but not a woman wearing a teeny tiny bikini and g-string. and of course, not banning topless people who "look male," even with full nipples fully on display. make it make sense 🙃

Di4na,
@Di4na@hachyderm.io avatar

@toplesstopics urgh yes. All i would add is that the one that decide are usually Apple and card payments network... MasterCard in particular.

arclight, to random

I have code that builds under linux
I have code that builds under cygwin
I have code that builds under cygwin with the Intel compilers (C/C++/Fortran)
I have code that builds under cygwin with the Intel compilers (C/C++/Fortran) and has debug info
I need code that builds in debug mode on Visual Studio 2017 using GNU Makefiles so I can generate a .sln file for somebody who wants to trace through the code on Windows.

Because I am in hell.

Di4na,
@Di4na@hachyderm.io avatar

@arclight the way we handle it in erlang is that we compile for windows on wsl
Cross compile from wsl linux to windows, with the windows dynamic lib loaded from the windows side of the fs.

The ability to use gdb on it is left as an exercice to the reader

Di4na,
@Di4na@hachyderm.io avatar

@arclight So you also need to beat their blocking devices, I mean sounds like a day in Y

jasongorman, to random
@jasongorman@mastodon.cloud avatar

We spend roughly 10x as much time reading code as we do writing it. A tool or technique that makes you twice as "productive" at writing code at best makes you 5% more productive over all. Making your code easier to understand will have 10x the impact. But that doesn't sell tools or put developers out of work, so you won't be reading about it in Forbes.

Di4na,
@Di4na@hachyderm.io avatar

@jasongorman I may have been writing about this stuff a bit lately... https://hachyderm.io/@Di4na/110431524694102577

Di4na,
@Di4na@hachyderm.io avatar

@jasongorman @flq I do not think we have to. We have a pretty good research into this the past like... 4 or 5 decades? In the ML community in particular but not only.

Ofc the problem is that bringing that to mass use needs a lot more "UX" work that we definitely do not fund.

But stuff like Rust definitely makes the C stuff more readable. Swift also made things more readable compared to ObjC

fasterthanlime, to random
@fasterthanlime@hachyderm.io avatar

is Revolut any good?

Di4na,
@Di4na@hachyderm.io avatar

@fasterthanlime i mean. Good enough that their exec teams keep leaving, they cannot provide financial statements and their UK auditor keep slamming them for lying in their financial statements.

You know. Small stuff.

frescosecco, to random
@frescosecco@mastodon.social avatar

I think that NASA's OpenMCT could be a great Observer/Mission control frontend for Erlang projects.
https://github.com/nasa/openmct

Di4na,
@Di4na@hachyderm.io avatar

@frescosecco having spent a lot of time in this space, i understand the attraction, it is nearly never the right answer

Di4na,
@Di4na@hachyderm.io avatar

@frescosecco yes, these kind if dashboards or mission control. Not in term of effort/benefits but more in term of being adapted to the needs of the operators. For the kind of software systems we tend to operate, the information needs tend to be far more ad hoc.

Closer to what @mononcqc point towards in https://ferd.ca/operable-software.html or what Cantrill call Zebras in this talk https://youtu.be/fE2KDzZaxvE

Mission Control is not really that. For good reasons! Different context.

arclight, to random

Today was not great. Desk-poundingly bad frustration over (intentionally? at least knowingly) poorly-written procedures.

Hint to any sentient human still writing mundane professional documents: directional and imperative language is necessary when writing a procedure as are complete sentences. Declarative language is not suitable for procedures, which are intended to describe actors taking actions; leave declarative language in the higher-tier program and plan descriptions.

If you cannot read the text of a procedure and convert it to a flowchart, you are not helping anyone. If you're consultants, you get paid regardless so wild. It isn't your ass getting hauled to prison.

I hate this fucking project and I want it to be over. Technically, it's fine, people are decent, but the process needs to be put down like Ol' Yeller.

Di4na,
@Di4na@hachyderm.io avatar

@arclight I call this the "well I am getting close to retirement anyway and we had clogged the whole system so that no one younger comes in, so why get better? There is no pressure, we control it all.

textfiles, to random
@textfiles@mastodon.archive.org avatar

I acquired (with help) a killer audio cassette digitizing rig, and then IMMEDIATELY had to spend over 2 months digitizing 90+ cassettes from the 1999 Game Developers Conference (via a slow-and-steady workflow that got the job done without being too disruptive to any other work I was doing).

So, here we are, it's done. Go enjoy 70+ hours of presentations about all aspects of game making and producing, in 1999.

https://archive.org/details/1999_Game_Developers_Conference_Audio

Di4na,
@Di4na@hachyderm.io avatar

@textfiles Omg there are so many good titles in this aaaaah

regehr, to random
@regehr@mastodon.social avatar

one of my back-burner projects for the last year or two has been a "tree search guide" which is a dropin replacement for a PRNG that is intended for use by generative fuzzers:

https://github.com/regehr/guided-tree-search

the insight is that a PRNG is only one of many ways that one might want to drive such a fuzzer. a different way is a breadth-first-like guide that I wrote, that gives about 4x as many distinct inputs, compared to a regular PRNG, for a silly example fuzzer that I wrote

Di4na,
@Di4na@hachyderm.io avatar

@regehr I suppose you know about Lineage Driven Fault Injection work by Peter Alvaro, Chris Meiklejohn and al?

thephd, to random
@thephd@pony.social avatar

All The Blog Posts I Want To Write Are Angry:
A Software Engineer's Dilemma

Di4na,
@Di4na@hachyderm.io avatar

@thephd I accepted it. I will be polite. I will be nice to people. But I will let the anger flow through me.

It is something I learned from safety science. Where the good scholars are all angry about the current state of affairs and the pain it inflicts on practicioners and users.

neil, to random

deleted_by_author

    •
    Di4na,
    @Di4na@hachyderm.io avatar

    @neil while i do not disagree, we have pretty good correlation now between the reduction in alcohol consumption and the reduction of unwanted sex acts ...

    Di4na, to opensource
    @Di4na@hachyderm.io avatar

    As a maintainer of OpenSource libraries and packages, there is something that kept feeling off in the whole Software Supply Chain discourse. I think this comes down to something simple.

    I am not a Supplier.
    You can read more explanation there https://www.softwaremaxims.com/blog/not-a-supplier

    Di4na,
    @Di4na@hachyderm.io avatar

    @stoff I am really sorry that our needs to sleep, eat and spend time with our family is heavily undermining your ability to get free advanced and niche engineering to your front door.

    And that it irks you that our sense of ethics stops when it means we are all abused. I am sure there is a world where people with higher sense of duty do the work instead of us... Oh wait we have that. They are burning out!

    Iirc the last stats 40% of maintainers have quitted doing so or are close to it.

    Di4na,
    @Di4na@hachyderm.io avatar

    @stoff Maybe next time we can talk of the real world instead of your imagined wonderland of fluffy unicorns jumping on rainbows?

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • provamag3
  • kavyap
  • DreamBathrooms
  • mdbf
  • magazineikmin
  • GTA5RPClips
  • khanakhh
  • Youngstown
  • everett
  • slotface
  • rosin
  • thenastyranch
  • Durango
  • ngwrru68w68
  • JUstTest
  • InstantRegret
  • cubers
  • ethstaker
  • cisconetworking
  • tacticalgear
  • osvaldo12
  • normalnudes
  • Leos
  • modclub
  • megavids
  • tester
  • anitta
  • lostlight
  • All magazines
    •