https://blog.expensify.com/2018/01/08/scaling-sqlite-to-4m-qps-on-a-single-server/ boring tech is the best

bash q: this wrapper

my_rebase(){
  git rebase "@"
}

is https://clojurians-log.clojureverse.org/ still working? it seems to stop sometime November. Is there another archive?

LOL, tried to throw an exception to test something, but then the throwing failed and I got a different exception!

(throw "Test exception. To be removed.") ;; ClassCastException :-D 

hello! Does anyone know of a lisp/scheme that works with tensorflow?

@pablore a lisp besides clojure?

clojure also

I'm working on this: https://github.com/bpiel/guildsman

oh baby

👶

how did you do it?

the bindings I mean

JNI

I took the official TF JNI bindings and made some small modifications, partially helped by the TF-scala project

genius

yes, but I can't take credit

I took A LOT of inspiration from this guy: https://github.com/eaplatanios/tensorflow_scala

as far as the JNI/C++ stuff

Didnt know there where java bindings for TF

how complete is your project?

that's a tricky question to answer. I intended to get an alpha out in Nov. I'm really trying to release something before the Jan is up. But, I also presented some working demos in Oct at Conj: https://www.youtube.com/watch?v=8_HOB62rpvw

Nice. Gonna try it later

anything specific you're trying to do? I could attempt to answer whether or not it'd be easy/hard/possible

OH also, if you're open to writing c++ and contributing to TF directly, note this: https://github.com/bpiel/guildsman#you-can-help

that links to part of the readme describing how you can help guildsman by making specific types of contributions to TF

also.. if you want to bring this conversation ON-topic somewhere (now or ever), there's #tensorflow

ok.. I've got to catch a train to work. Let me know if you have any questions/thoughts/etc

@pablore @bpiel: there's also 1. mxnet and 2. jcuda if you're willing to deviate from TF a bit

well i know c++ and am interested in digging into ai stuff but machine learning seems to have a very steep learning curve and i dont have much time to devote to it yet 😞

@bpiel I wanna get into NLP. Still dont know if TF is the answer

@pablore TF is an answer. I don't know if it's THE answer

I haven't done any NLP with TF

@jgh I think the steepness of the learning curve depends on the individuals comfort level with the related math

However, I can say that making the contributions to TF I mentioned earlier does NOT require much knowledge of math/ml

just python & c++

https://www.reddit.com/r/programming/comments/7klny2/on_lessons_the_common_lisp_community_could_learn/drfwzp6/

@vemv, "$@", not "@"

...erk, missed the thread hanging off the original question.

Am I the only one who find it really difficult to find gigsin Clojure coming from different technologies. Over the years I worked with ruby, python, golang, angular, react at a seniority level and before jumping to each of those I was not familiar with the tech in question. I feel like people hiring in the clojure community put unnecessary barriers. I may be wrong so I'm curious of the reasoning why no one is willing to hire a professional familiar with other stacks who's really interested in Clojure. Basically the only community where I apply to job I get answer such as "You don't fit the profile" we're looking for - it is really a turn off

@lewix anecdotally I’ve seen a fair number of hires and none of them knew clojure before coming on board (current job is 100% clojure, jvm clj on the backend and cljs browser frontend) - but we rarely need to add more people; I can’t answer for any of the people that failed to hire you of course

@noisesmith where they white? just curious

(seems like a good discussion for #jobs-discuss BTW -- since we've been talking about exactly this recently)

two years ago it really frustrated me so i gave up on clojure and went to work with something else. Recently i tried again and same thing.

@lewix yes, we are in one of the whitest cities in the country which could be a factor

The only feedback they gave was 'don't fit the profile', with no details on what that profile might be?

@fellshard right

A lot of companies hiring for Clojure seem to not have bandwidth / budget / time to train new-to-Clojure folks so that narrows the field a lot. Some companies are willing to hire new-to-Clojure folks and train them, but it's a smaller pool.

@fellshard: Legal/HR probably won't allow them to say anything

And you've also got to realize that the desire to do Clojure is high and widespread so candidates completely outstrip available positions.

Even when we were hiring for other tech roles, because folks knew we did Clojure, we got a lot of people applying to those other roles who really wanted to get into the Clojure work instead...

@seancorfield that is a good point but I wish they said that they found better candidates as opposed to say I do not fit the profile.

I received an answer very quickly less than a day and still the same thing - you do not fit the profile

That's just "HR speak" in my experience -- companies are often very limited in what they can tell unsuccessful candidates (assuming they tell them anything and don't just ghost them).

I mean a this point I have no desire to apply, I'm just venting haha

Hiring / job searching is an area I have very little experience in, so a great deal of what I hear is hearsay. That said, being turned down without any feedback at all is... aggravating, to say the least.

Multiple shops?

yes

I should collect all the emails, and just write a blog post about it

We have a JIRA workflow that both HR and hiring teams use -- and that ensures there's a "contact unsuccessful candidate" step for every candidate.

our team did hiring at my last job. company was big enough to have big HR department. we didn't handle rejects, HR did. so there wouldn't be any meaningful feedback from that

@fellshard: Anything the company tells you is grounds for discriminatino lawsuit on your part -- therefore, in a really shitty game theory way, the company is best off not telling you anything.

that was the policy at previous job for sure

They gain absolutely nothing by telling you why you get rejected, but open themselves up to all types of shitty lawsuits.

In theory I could sue them for saying that I do not fit the profile then?

Ah, 'murica, home of the frivolous lawsuit... 😞

you can sue them for anything

(Not saying that you are a litigious jerk, but that such people exists, and legal/HR is covering their own ass.)

you probably won't win. but HR is lawsuit mitigation as well as staying legal

and it costs money to be sued, even if it's an obvious losing case

oh well, thanks a lot of guys. I appreciate your input. I have to head back to work

Hope you find a better opportunity soon.

I won't look for clojure opportunities anymore. I value my time more than anything, and it's not worth it. I think I'll keep on playing on the side with it though 😛

do what you need. but i can't imagine the two places that you have applied for clojure jobs give you an accurate snapshot of the clojure jobs available as a whole

Actually, have you tried making a few commits to cider (or some other open source project) ?

^ that's how i got my job

I bet if you can say something like: "cider is the most popular clojure emacs dev setup; I contributed the following features to cider"

it should bubble your resumee up to the top

don't know about that. but its certainly a conversation in the interview

also, most CIDER development is in elisp lol

For better or worse, OSS contrib does seem to be a vital factor in Clojure discussions

@dpsutton: true, but ppl who hack elisp for fun tend to be top notch programmers

in the stack overflow survey recently, vim users outperformed emacs users in job interviews. i don't put much stock in how good someone is as a programmer based on what languages they use. except forth. those people are geniuses

I'm not sure if this debate is going anywhere productive, but when someone tells me they have made elisp contribs, my thought process is: 1. this person understands scheme / recursion / lambdas / ... CHECK 2. this person cares enough about their craft to build their own tools ... CHECK 3. this person cares enough about their tools to use a dynamically scoped pile of shit .... CHECK 4. this person has the confidence / courage to build new things rather than a cookie-cutter view of the world as it is .... CHECK but maybe I'm reading too much into programming elisp

> 3. this person cares enough about their tools to use a dynamically scoped pile of shit .... CHECK hahaha perfect

it shows grit + endurance in the face of annoyance 🙂

I'll remind everyone once again that a lot of good programmers cannot contribute to OSS, either because of restrictions imposed by their employee, or because their lifestyle doesn't allow time to work on software-as-a-hobby outside office hours (for reasons that are perfectly valid). Leaning on "candidate has OSS experience" may bias you against better candidates that don't/can't.

(remember: you just talked about irrational and unfair discrimination above!)

afaik, "has open source contribution" is neither a protected class nor something determined at birth, it's more akin to discriminating on skills

@qqq Demographic analysis shows that a lot of protected classes do not contribute to OSS so, yeah, it's a good indicator of discriminating against protected classes.

(and it's partly because the OSS community is often harsh on such folks)

@seancorfield does that take into account that protected classes often dont participate at all in the tech industry?

Not saying it should be that way, but rather simply that it is, again for better or worse.

The percentages in OSS show that even proportionally those folks are way under-represented in OSS so, yes, it takes that into account.

ok, gotcha. That’s interesting to know.

I wonder why it would be worse in OSS than in the industry at large.

Heh, if you're genuinely interested, I can probably point you to all sorts of intersectional material on the topic...

Differing incentives. In a sense, OSS contribution is a luxury afforded by having time and energy to spare.

@seancorfield: genuine, non rhetoric question: do you know any legal precedence for situations where "we want OSSS contrib" has been prosecuted as "discrimnation vs protected class" ?

and a contract that doesn't forbid it

@qqq See above comments about how hard that would be to prove as discrimination...

@seancorfield i think that would be an educational read if you have material at hand, otherwise i can probably use the googles

...besides, it's not what I actually said: I said that by favoring OSS contributions you are likely to bias yourself against good people who can't participate, without even knowing it.

And it may not be active discrimination, but rather accidental or unintentional; conflating the two categories would not be helpful to the industry, either.

Better to coach than to attack.

@jgh Here's a reasonable place to start https://adainitiative.org/2014/01/24/guest-post-whats-wrong-with-meritocracy-in-open-source-software/

(I'm hoping folks are familiar with the Ada Initiative?)

seems they aren’t active anymore

great name though

Yup, other organizations carry on their work.

thanks for that post, good read.