#django

Excellent question. Especially since it was public information. I don't know.

Some folks would have wanted him fired for inventing Javascript. :-)

Every project has politics. I would rather that a project's politics attracted good contributors -- and good users! -- rather than repel either.

But I'm just a (hopefully good) Django user. :-)

To me I think it interferes with what needs to be done

How so?

What needs to be done that's not being done because of this change?

django admin prepopulated fields for slug, it seems to automatically remove common words like “the”, “a”. is there a way to turn that off?

For example the time spent changing all instances of the master/slave terminology could of been spent fixing bugs(not saying there are any, just an example) or improving features.

HowardwLo: It's baked into the JS, so.. Not without replacing it, no.

HWJohn: Eh, false dichotomy. Whoever was involved in the master/slave change may not have otherwise been spending time on Django at all.

HWJohn: the time was spent by a volunteer who otherwise wouldn't have contributed ... yeah.

The time taken by reviewers, sure.

And now in the channel perhaps.

Heh.

But that seems to be small beer.

But then again, engaging people in discussion is always good.

Well I wasn't talking about django in general

I mean universally

What does .s() do in Celery?

And FWIW, Django's core developers, probably moreso than most OSS projects, do place a high premium on cultivating the Django community's culture, not just its codebase.

mattmcc: I wonder how much that attitude has contributed to Django's growth and development.

Synthead: It's a shorthand for creating a task signature

Synthead: http://celery.readthedocs.org/en/latest/usergui...

Anyway, thanks for the conversation, have a good night.

Culture is fine with me, but when we get to the point where people are upset over words that aren't really harmful or start saying "it's too masculine"(seen this a lot), there's an issue that's not cultural.

mattmcc: I actually heard a mention of adrian on a totally unrelated-to-django podcast last night.

I hate to be the one to change the topic :), but I would like to ask when in using CBVs, the examples I have seen often do something like get_context_data(self, **kwargs): context = super(ClassNameView, self).get_context_data(**kwargs)

Why is that syntax used in a CBV?

I'm not sure I understsand the purpose

acmehandle: That keeps the behaviour from the parent class(es) intact.

HWJohn: On the other hand, what if the industry actually was less "masculine" across the board? http://money.cnn.com/2014/05/29/technology/goog...

You'll see that quite a lot in python code: it's the only way to ensure in the case of multiple inheritance (which includes mixins) that all parents' versions of a method get called.

From the parent class being the DetailView or UpdateView?

acmehandle: Yep, and any other mixins. Even if you don't know the parent class(es).

mattmcc: the higher males than females could amount to being just a higher amount of males in the industry

So then why not perform that on every method in the ClassNameView?

Why only certain methods?

acmehandle: Which means you can use it _in_ a mixin, to ensure that the other versions are called.

It's not too often I see a girl who can code

acmehandle: If you don't do it, it means you are _replacing_, rather than extending the behaviour of that method.

acmehandle: (which, in a lot of cases, is the desired behaviour)

HWJohn: Oh, there's no question of that. The question is why, and what's the opportunity cost of an industry that seems to broadly deter women from entering it?

The desired behaviour being the extended behaviour?

When I studied CS, there were more guys named "Matt" than there were females in the course.

mattmcc: I don't see them detering women from the industry, it's just that they aren't as interested in the industry. Yet extreme feminists in the IT industry blame this on the male privilege

when it's actually just there aren't too many females interested in the first place

Perhaps because those considering the industry see an environment they aren't comfortable entering..

HWJohn: if that's your experience, perhaps you're one of the people dterring women from IT

It's not like I would tell them to go away

I imagine the gender difference is quite a bit larger than you'd see in most scientific fields, for example.

HWJohn: it's not about that

I honestly don't care what people want to do

Yeah, I always found it strange that there werent more women in programming.

acmehandle: Especially since the first programmers were all women.

mattmcc: pardon my ignorance; the docs are confusing. I have two tasks I need to run synchronously (one after another). The two jobs should still run in the background like .delay() would. How do I do this?

Exactly

And programming is like the least 'hunter' activity

Well almost.

Looking for bugs in code is sort of a hunting activity

true programming is a creative act

Synthead: So you're probably looking for a chain. http://celery.readthedocs.org/en/latest/usergui...

essentially, creative problem solving, combined with expressing it in a constrained language

we are poets with a limited audience

I could be wrong but I believe that programming requires the least amount of bravado so I find it strange that feminists would make any association with male priveledge

Working with limited tools

frankly

mattmcc: in the first example, "add" is an example task, right? not celery syntax?

Synthead: Right.

Synthead: Scroll down to the example that uses chain(.., ..)

mattmcc: so I have to use signatures in chains?

Synthead: Yeah, because you're not actually executing the task yet, you're setting up the sequence to be executed.

Hence 'canvas'. Laying out the pieces.

I think femnist try to make that claim to get attention for themselves. If anything I think women why away from programming because men wont give them enough attention.

It is an isolated profession.

mattmcc: oh, I see... if you put the tasks in a function like chain(), it would run them as normal and pass them to chain, which isn't what we want yet

I can't really believe that this is a profession better suited to one gender or the other.

mattmcc: s/pass them/pass the return value/

No, I think this profession is the most gender neutral one out there

acmehandle: That sounds like a remark a Mad Men character would make, frankly.

Probably.

acmehandle: the work itself is, sure... but not the industry...

I am not a good programmer because of my naughty bits :p

gender norms and stereotypes are retarded

thats true Funky. The amount of attention and money lavished on startups is obscene

So anyway. class based views.

Synthead: Right.

mattmcc: can I do something like that = That.s(...); those = Those.s(...); chain(that, those) ?

Sure.

mattmcc: ah that's great. I'll try it. Thanks a lot for all this help; I greatly appreciate it

The profession may be gender neutral but one has to let go of their own ingrained gender stereotypes

mattmcc: especially Celery ... what a nightmare :p

If I don't care about the result state of celery tasks (that is, I don't care about what they return), but I *do* want to use flower, and I *don't* want rabbitmq to peg one of my cores and eat up memory, is there anything I should be setting in the celery/rabbitmq configs?

karanlyons: You _might_ want to look at event snapshots.

Not something I've played with much.

I'd have to monkey patch that into flower, I'd think.

What aspect of the tasks do you care about for monitoring? :)

Right now rabbitmq is eating 3gb of memory and pegging a core at 100%. I have 263,561 completed tasks in the last 20 minutes or so.

If I"m extending a method in a CBV do I have to follow the signature as the original whan when I super it?

I don't really care about anything about those tasks beyond the metadata, and I've never really spent any time configuring rabbitmq. I honestly don't even know what the eviction strategy for completed tasks is.

acmehandle: That's usually the best way to prevent unusual side effects.

acmehandle: That's true of subclassing in general, too.

Although, as long as what you pass to super(MyClass, self).<method> matches, you can change stuff.

I do this to have a mixin that will grab a user passed to a form instantiation and store it as form.user, for instance.

mattmcc: do I have to run .get() to run the signatures in the chain?

Synthead: You have to tell the chain to run, yeah.

mattmcc: oh, I see, they're doing chain( ... )() in the example.

I could set ignore_result=True on my task, but ideally what I'd have is the task metadata getting rolled up for flower, but the task results being trashed.

I'm pretty sure that's a contradiction.

Yeah, without flower being rewritten I don't think it's possible.

The problem is that I run a *ton* of tasks, which is fine if I'm just evicting them when they complete.

What you may want instead is to build some progress reporting into your tasks, separate from Celery's AsyncResult, that you can then aggregate and discard at your leisure.

Problem is that I'm then building another system on top of the one they're paying me for. Wish I had other developers.