SunPy Developer Discussion, 1 Jun 2015

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: SunPy Developer Discussion

Description

PRs and 0.6

A

And out is on: why be careful what you say exactly.

B

Watch your language, oh.

C

This dude hello, okay, Stacey.

D

A

Stuart how's our progress throw points. Events, oh.

B

Well, it's not doing too bad, actually I think look doing. Nothing was mostly there I mean. Are we gonna? We need to decide what we're going to change the names of the name, tuple thing in map and.

A

People don't seem to feel all that strongly about needing to change it as made people offered alternatives, but it kind of like it could be this or it could be that there's no clamoring that I'm hearing I'm.

C

Clamoring for change, I, don't really care which way we decide, but I definitely want to get rid of those stupid dictionaries. Oh no.

D

The nation earrings are gone. Oh good.

C

Okay, question.

A

Is whether we're changing the names it should not be X of Y, if I just mean, let's do er I just saying the issue has basically had no.

C

Nobody cares, don't know, as you gaze I think X to Y is fine. No, it's not we're.

A

Gonna be really.

E

Really suggesting.

A

They're calling an x-axis and y-axis I really don't understand what that's better for your perspective, because.

C

It makes it less like.

A

That it's not legal.

C

To hpc anymore yeah, but.

F

X and y are plots, are.

G

Like way more generic than anything like that, I mean how can we be offended by that? Come on you're, looking you're looking for a reason to be offended by that, because exok.

B

Exom, why are you like stupid TV? It's all right, I mean he's not the end of the world. It just kind of it's just a little too.

E

Only if your sense very.

B

Dense too yeah I mean I don't mind if people want to keep it as is then we'll keep it it is, but I think it's probably be accessed my act if.

G

You draw a plot and a whiteboard. What do you name the FD exit like randomly for anything I? Don't.

H

Know are you talking about polar coordinate system of this like.

B

My point, my point is that undefined my point is that it should be of to play putting axis there. It makes it obvious that you're talking about the axis if the plot, as opposed to the axis of the coordinate system, but especially, doesn't really.

I

Okay, I know k.

A

Believe it is well like I, said no discussion on the email and over email. Basically, no new discussion on the issue. I'm, not convinced that there's left kind of thrown over this I.

B

Just it just seems weird to me to do Matt X to get something in Helia, graphic goodness, which you would happen which would happen if the map was in native Helia graphic as a phobic. You.

A

Could just do map bracket 0?

A

B

You could, but if we're going to say what you could just do, that we will just make it a two-point only allow by everywhere I.

G

Would have thought if I ski-doo the graphic, then you should somewhere somehow explicitly mention that it's X or Y geographic, rather than getting offended by a nonspecific X&Y label. But.

B

The point is that map should not be directly linked to whatever coordinate system of maps in.

B

Its fine, fine, northing and using the argument bet.

H

Nobody, like 'I, wonder need to or x1 oh.

C

My god, that's terrible! Well,.

H

C

H

I mean, though, the DS notes like x and y, is just it's purely heritage. Wait yeah,.

C

Take a heritage.

H

C

Union lightning, c'mere ya,.

C

Must Lester's just throw it all away all of history, but.

B

C

Issue with this, it.

B

Whoever decided that he appear projected was Cartesian, coordinate, system and decided to call the axis fecking x and y is ya biggest lie ever that's dumb, not.

C

B

Coordinate system they should not be named x and y is in fact a spherical coordinate system. They should be David longitude latitude. You don't should swivel what.

E

Their level we should do it understand.

G

B

Just cook very painless,.

F

B

A

B

Have this problem I disagree.

A

Because we don't know we could have maps that are in Cartesian coordinate systems. What.

B

We do we have, if my point is it shouldn't, be coordinate system and you shouldn't that the names of the attribute should not be implying anything about the coordinate system in which the data is the held, but if you call them wet and lon, you would be right. Oh sure, we.

I

Shouldn't call him that and Eliza okay.

A

Its wine mixer.

C

A

C

Abstract the easiest abstractions of people would understand right away about the axes in the plots that they're using right there shorthand they're, not perfect, but they're understandable, and we can move on.

C

We call them anything else than four we're forcing people to learn what we're doing, and you know we don't want to put those kinds of obstacles in people's way. No.

B

To my mind, saying x-axis and y-axis is more explicit than just x and y. By suppose it doesn't necessarily remove the ambiguity I am taking offense to.

C

So what are we doing? How we're moving have.

E

You got bigger.

G

Problems myself, I well yes,.

C

E

We don't that's great.

C

Yeah we've lost Albert he's rudely taking a call with.

E

Someone calling to complain about the EXP exes.

C

um How Stewart, how long do you think how much time do I have to fix him the plotting step for to get it into 0.6, or should we ignore that completely fixing wore their plot fixes.

B

Well, you mean.

C

We're making cloth.

B

And peak to the right thing: Tribble.

C

Yeah- and there is a few changes that I think are important in terms of the way we automatically set the color bars so.

B

I was having a conversation with Larry manager, / hangouts the other day about normalizing things. He was getting very grumpy about normalizing things, but that's a grumpy. He was saying that it with grateful stuff the way we know all those things that learn, I, think the normalization stalking, probably after white Wow,.

D

C

D

C

The single single thing I wanted to accomplish for sure was that we show all of it is from you know, minus whatever 2 plus whatever the other thing that we need to do is.

B

Make sure it's easy for people to change because, like he does bite skating, he uses the ideal white skating routines to scale it and looked at 25 40 f. So he can feel competences yeah.

C

Yeah I mean I change the api by moving it all into a like what parameters, dictionary and.

B

The other thing yeah that's a good thing as well. The other thing that I remember from when you did that last time he says now the ass for pie, visualization of normalization stuff. Remember I linked you to that route sometime after that all requests.

C

They went in again so.

B

This a stroke why I've got a load of new coke about doing.

H

This kind of thing is it the least yeah in the main branch said yeahh two more point: Oh.

B

H

Think this is a kind of estate discussion of the more important question is: when do we 10.6 out, then we want to know different I, favorite area, I think getting a as quickly as possible before we have another pail of g shock. Cord light completely agree, so I mean it may be nice to fix the normalization stuff, but I think we need to like I've been tellin people's. My texts coming out, so don't bother with 0.5 on something so always bother with bug fixes nobody I know, but we keep it.

H

Does that need to hold up 0.6 and take down okay. So no I don't think there is.

B

Otherwise, we'll never remove never were. There are two open pull requests for 0.6.

C

B

C

Mine, maybe mine, you.

A

Don't have an open for wordpress right now, right, I.

C

Know so that's why I same time and I can put it in you.

E

Just want to be mentioned and that's a charge. Steve.

F

B

D

B

I would I would propose that 1423 gets in as well, unless someone spots something majorly wrong with it I think it's done so.

H

You put the 0.6 label on it nope, but if I.

B

Have now there are now three open progress, milestone, 20 quid 6. Thank you to a ridiculously simple. We should merge them rubber, hose.

H

The documentation looking 20 more dollars probably do need some love yeah, so there may be a small window for you. Stephen.

B

Suggest we leave it because I'm not sure I'd rather shut off changes did not win six yeah. I hope it's on it, get a release. Candidate I'll get the documentation edited, and then we get a good running Salvador. Yes,.

C

Okay, that's fine! You give me a timeline if I can't satisfy then just move ahead.

C

B

C

B

Will probably be baking nor point six RC one sometime around midday, my time or friday, then.

C

Okay well I'll see, if I can make that not you know, don't worry about me. Just go ahead.

B

Okay, so can someone else review 1415, please.

B

By luvely online fix.

C

C

Here have your thumbs up yeah.

G

I'll give it a thumbs up.

G

Don't tell me, that's complained about this. One.

F

What are you doing.

E

Good, who is calling.

H

He may have a species fit in his office.

B

Okay, so 1423 the last remaining or reverse biased owners bill points that exceptionally contains some interesting changes and probably worthy of about some discussion so which one it's 1423. Okay,.

H

B

Adds the ability to pop in and user-defined function to the map key animator class, which is the thing that is generated by Matt? You've got Pete that allows you to define custom plotting routines on top of each frame of the map, view and.

B

Thereby allowing you to do a massive animation with an overlay grid or a limb or anything else, you've important problem.

H

Well, I think this is great, so this came out of Andy's desire to animate market with Lynn mended on it. So this is a generalization of that wait. Yeah.

B

Doesn't this doesn't do anything with map keep the floor which is possible, but Jesus Christ into it, and but in theory you can call Matthew venom, eight dot, there's a method to map your enemy. They gives you a map, la Plata, animation object which allows you to save it. Fine I'll.

H

Do this I'll k with us this afternoon and the.

B

Two other changes that come with this that I, actually really like, is the change in behavior to draw a grid and draw them in map. This makes draw real in return. The map wat lib patches object.

A

Instead of the Hatters art idea, though, instead.

B

Of the eighties object, the reason I had to change. That was because, when you define the custom plot function for peak anything, you've returned custom plot function, has it removed method called, and it is deleted at the start of the next step next frame and by allowing you to plot something for a frame, delete it and plot something else on top of it, because otherwise it was stacking all of the limbs for every frame.

B

You know in the map your rocket, so if you want to plot something that stays there for the whole time series, you don't return it and anything you return gets its remove, object, move, remove ever call, and it's deleted from the point I.

A

Know, okay, I haven't tried running this code, but just looking at the dock after ducks, how does it? How do you use splat function?

A

How do you watch sorry I'm, just looking at the dock stirring I haven't tried writing this curve. I'm just look at the doc string. You find a custom plotting function, my plot that draws the limb yeah, but how is that my plot actually called or oh.

B

I didn't add it two absolute an oversight and a look at the comment in the original, in the top of the plural, quit you passed in as I argument to a function.

I

To the state now, okay, that makes my second sir I was but magically cold magic.

A

Also am I miss reading this story, but these this is a wise. The doc string forepeak not have examples using a peek.

B

Yeah, what is not my fault has clearly been like that for a why.

I

C

um I have another question on another subject: great p for that yep, so the branch is gone. 0.6 branch is gone, no.

B

Are we still using it or no? Yes, I just I rebased, the null point, 6 branch, so basically I deleted it and rebuild it because it got horrific, be out of sync with master and it made it very difficult to merge. My mat pull request on 29.6 who's going to be really ugly, so I just did a bit of get cleaning, but it's still bet any spend it now allowed. You can now pull request from master no.

C

Point sticks so master is I mean whoever is a master. Goodbye is no pointing what.

D

No well yes, and no. What.

F

D

C

Is no difference between T is what I should be developing you you.

B

Can develop on to master and then I will pull it across the North mindset. Okay, that's what I wanted I think. The only difference between the two at the moment is the version number, but we're having the not win. Six branch allows me to do some build stuff a bit easier, then we'll be able to avoid.

B

We have got one thing that needs to be done from a release perspective, and that is, we need a WCS accys canta package before we do it release and which is being worked on on the astro pi side. So hopefully they'll get that done, for we ought to do a release and I have to bake my own.

B

Finish: I might bake my own anyway,.

B

C

Anyway, yeah I.

G

I, just pray for you mentioned some pointer in this meeting, but you go ahead. O.

A

Ok, so Jack you're gonna check out the math you've animator yeah.

H

Yeah I'll do that this has to mean okay.

A

I'll try to take a look at it too, but anyone else who wants to take a look at I should let's get this.

H

Moment they so the docs 40.6. Should we just go down the to change log and then see what the doc see in lessons that bear way to go for the docs.

B

Probably need a complaint rate for her to make sure that they're sane.

B

But yeah I mean.

A

B

A

Place to start it's just near to Steve. Look like he's frozen.

B

G

You still there, it's just Albert.

H

Cold-Hearted stunned by fun pie, I.

G

Just wanted to mention at some point during this meeting as with just what I kind of mentioned in the the email was away on Sheldon and just to hear if anyone had a new about this at all, basically there's the in the new version of our the other coming up. This will just to make people aware, there's going to be an ideal, Python and Python IDL bridge that they're writing, which means that, during a audio session, you can call Python code and vice versa. Now, officially the downfall OH ideal.

B

By the way, just saying this, all like yeah I, know ice.

G

Is the end this.

B

Is the last Horsemen of the Apocalypse that.

G

Is it I up dead? It's either yeah I mean right exactly so. It's like I'm not going to tell them how excited I am about this and why I am but anyway.

G

So the point is that for, like there's still a huge amount of ssw code that we haven't, you can't the capability which we can be producing some time at the moment or we haven't reproduced, but this would enable us to essentially call like instrument specific routines, which are only written idea in a Python session, so there's a discussion to be had as to whether we're interested in either using it or promoting it as some way of like hey.

G

This is a way to you can still use your instrument, specific software, an ideal but actually move over to to python and or if you want to use it as a solution that you can even some pie itself and I'm not sure that's a good idea, specifically at all, because then you're alive from.

B

G

Dead body and if no well right but I'm, just saying if to get functionality, the all the thing is that a zit discussion to be had- and the first question would just be- is anyone aware of this or interested in promoting or using this in any way? But it is.

I

G

C

It's not out yet right. It's in the next version of ideal, correct.

G

Its it's in the upcoming version, video well.

B

G

B

G

Yeah for the next, the next eight.

B

There's no difficulty to be deadly, not from some fine master did not dare not some pie, prime, but there's definitely. We could definitely write a good affiliated package which represents nice, Python rappers to some of the most important ideal functionality, coughs, ecchi, breath, cough and and other instrument calibration our routines. No, no! No!

B

H

B

H

So use IDL. If you want then write out the data and then read it in Tucson pie, I, don't think no.

C

That's not always so simple, no.

G

It's also big discouragement people tend to want to when they initially I. Think about this question. I, don't think I want to use Python or audio if I have to use both. Why don't I just use audio? It works. It's.

B

More anything, you know you that you can work like that. You can just open up open up an ideal prompt and do seki preps save file, or you could load up and I by the notebook read in a fits file using map pass. That map object into a Python function which underneath goes out and call sec. You prep and capacities you back a new map, which is a nice calibrated second map. That second workflow is much nicer and appears to be Python, even though it is.

G

It's what it's, what me that a user would want to do and if you, if you open up ideal and then go back to python, you have to be absolutely determined to use Python, which is not literally a bad thing. But a lot of people aren't so if you'd only have to interface with one language.

C

I wonder who the audience is for this: the people that are doing this that are like going out purposely the good thing, no, that they need to go to I, do and run like sexy, prep or something and then bring it out into into Python and some pie. They are already converted right. You know they're already on our side, I guess it would be nice for us to spend some time to make things a little bit easier for them.

G

So this is it, so, are you saying the people who would want to use sorry to go out to a separate audio session and then come back or already on our side? Is that we using yeah right exactly yeah I? Do.

C

It the you know, but I, want to do the rest of my analysis and Python. I just need to do a few things that I can't do in Python an ideal and then I'll come back to primal. I think the question is that if we provide these kind of tools, is that going to convince people there are not using Python now that they can just skip out on ID all and move to python and.

G

It still requires that they have audial 8.5 Nadia license, but it means that they don't have to. They only have to interface with one language right, but it would be a new language for them right right, but the point is, if they didn't have to jump out of that mean up. Oh heard this a couple times about saying: oh it really introducing Python. But but can you do you know? Second?

G

Well, you know whatever and it's a car well I mean if you can't do that, then I'm not going to vote, whereas if you could actually do everything by interfacing through python, at least, if not Sun pipe yeah.

C

G

Suddenly becomes a much smaller excuse for the not to bother, even though they do have an interest. You know, because.

H

G

H

What it says is like: why should I bother with Python at all, what, if I'm going to use second enemy? If I don't, I just stick with this idea. The.

B

Argument gets a lot thinner if you don't actually ever have to write a line of my dl coach yourself exactly.

H

If you're going to have today could anyway, then, why not just stay an idea.

B

I sure you're always going to get people with that argument, but I think that you remove one more barrier and I. Think I do have a good.

G

People would come across well, I, don't.

C

Have many anything.

G

It would be someone I think.

C

We could argue kind of both ways on this I. My personal view is it's probably not worth our time spending too much of our own personal. You know our own time diverting to stench recordings IDL in Python, I think the first thing we should do is we could maybe you know, create some guides or something I'm ipython notebook in the galleries be like look. You can do this this great and not. You know great lots of code just show like. Oh here you can.

C

This is the circuit command and we send it off the IDL and it comes back right. I think.

G

This is initially.

C

The most valuable.

G

I

G

Do I don't think, actually takes that much coke? It's essentially you just call the audio.

H

G

H

G

Comes back a dictionary.

H

G

H

Have to see is a.

G

H

Whatever comes back right would have to be passed by son pie and turned into some kind of son by object, and then at that point no we're supporting IDL night. No.

C

No, we'll just show them the code on how to convert that to like a weapon, but.

B

Then something's changed and I would I would argue that we should not put death time into this if someone gets this goes so that's quite interesting and wants to write some docs or even write a package, an affiliated package to perhaps some of this stuff. If someone else wants to put some time into that, then it's fine but I should like a take which writing all dev tighten too much so.

H

I mean I have a use case for this actually, with I would love to use imce and the resi called the spectral 15. So this village will be very useful for that. I'm not going to expect some party to support that now. Do that on moment, Amy yeah I find I, think I I, don't think. If.

G

I got history, I.

B

I think if any form of support should be in a affiliated package, we should we should not support this inside the core library yeah.

G

C

I think everyone agrees on that yeah.

G

But I also think that we shouldn't spend, like my view at the moment so far, is that we shouldn't be righted 22 much time, writing code supported, but I think it would be really useful to publicize this, because I think it is a selling point that it it's. This start this one argument or one of the other two gets it. Oh yeah, but you've done have sec. You don't have this in some kind. You know to counter that argument, saying: there's no reason to leave hyphen. If you want this yeah.

I

G

You want this functionality so advertising time, I! Think you should. We should seriously consider some advertising for them to some I. Think I mean.

B

G

In theory, sorry.

H

Cool, so for we need to see how this thing operates because, like if somebody's gonna run say, keep it and expect to get a fits file of the an array exactly the same way that you know if they expect idea, but at pizen, then I'm not sure. But that's no.

B

That shouldn't.

H

B

Shouldn't be what we expect the things like okay, so I have my license in theory. Okay, correct me! If I'm wrong, I haven't used ideal in a long time. No I like it that way, and but if you call AAA prep, you give it a file name, and you get back a header and a right right. You get back it's trapped in a right.

B

Is that right? Yes, yep, so I'm, assuming that this pie, that bridge, if it's done well, which you know hmm may not be a bad assumption- is coming out, IDL, awful and we'll get back to variables. When you call it you'll get back a non, quite array and a dictionary.

B

Now you could in theory, throw that into the map constructor, and that would be it. You've got a file name. You.

G

H

Little bit, I did.

G

See was that if you, if you put or if your ideal output, that you're putting into python is a structure that maybe should like the most sensible way to put that it might be numpy record in some cases, they'll put. It comes out as a dictionary, but I mean, if I just saw one brief. My.

B

Point is direction if the if the API looks something like this, where you have like a high SS w dot aia prep, you give it a file name.

G

We've lost assured the commute of yourself. I.

B

Was typing it? So if you, if you call something like this in the chap, what you call pious, SWA, I, a prep and a file button, you get back a header and data which are two outputs of this over something. If you do that, you can then just do map.

B

Hell is going on with my keyboard. You can just do math of theta.

B

You just do map later.

H

He'll, like that I, don't know, we'll have to wait and see so AIE, for example, has this funky index file, which I'm not sure is used in good else across the cost?

H

Well, all the lockheed based stuff I think uses this. Can index construction I mean.

B

Obviously there are long endurance is, but I think it's definitely what I'm. Definitely with that in that. If this thing exists and you can calibrate data from a Python API, even if it is doing nasty things of IDL, underneath it's still going to make life easier for some users and therefore it's something we should be documenting that we should be compromising yeah.

H

B

H

End up having to translate a whole bunch of translate the organization of her someday and ADL. As a has thought about, does technol problems then affiliated package fame, but not in some pay problem. I. Think.

G

We all agree that there.

B

Will be no pollution of some pipe with evil ydl code.

G

So for philosophical or you.

E

Know derisive language, it just makes everything a.

G

Lot harder, yeah, it's you know, it's just do.

B

It just for ease well jokes, aside: England's I mean for a start, you don't you don't want solarsoft and ID able to become a dependency of this on my library, I mean just just. That alone is an argument to never do that. Right, I mean it's activated, impractical to a whole new level. I, do wonder about building a condo package for so soft that.

G

Will wonder away with.

C

Time there another more point: I, don't.

B

Worry I would be.

C

um I'd to other bits of news ish, oh.

C

We did not get funding for some pie. Development through master, um essentially deemed some pie proposal to be irrelevant to the call, unfortunately, which I thought was a pretty harsh interpretation of what we are providing and then I also spoke with. um This is more like bored news, facebook. What was it Jesse? What's? It was their name Chris.

H

I

H

C

H

New York hella cool yea, Korean.

C

Fantastic Kelly um and that's I mean she spoke with someone at NSF and really didn't go anywhere me and.

C

Yeah, it was still a productive discussion with them, so they're kind of a little bit more on their radar, but they provided the same types of excuse that you would expected. Why? Don't you guys, you know funded through some other ways and in the normal way right projects can cleanliness blah blah blah anyway, I just thought I'd. Let you guys know.

C

H

G

I'll puts it home, which I say yeah.

A

Going back to PRS, if people hear me yes, yeah I came.

H

Up early I throws, but.

A

As long as you guys can hear me on the bead had a he mentioned on the response to the event that he wants people to look at 1002. We are the oldest PR. We have on our list, 1002.

B

So yes, he has has resuscitated this. Isn't it hey.

A

Hey on second Stewart you're, not getting an error when you don't specify plot function, no I'm getting it iterating over none type error, I'll, throw up the trace back somehow.

B

Oh no wait: I was on the wrong branch. Yeah yeah. Oh, that makes sense. Hang on a thing: I need to change the thing that is returning.

B

D

So cool about that anyway, 1002 is scraper.

A

Any thoughts anyone.

C

Why why do we need this yeah.

A

Well, that could be a question for w okay.

B

So I haven't, put I, haven't put much brain power into this recently, but I think what it does if I'm, remembering correctly I haven't even looked at this recently, is that it allows you to specify a like pattern and then get all the URLs that match that pattern. So if you to an ftp directory and when I want all the files that are in this directory that exists for this day of this month, right.

C

No, not you're answering what it does not why we need it all.

B

Right so it's useful for things like em say the goes like a where you want to get all the files for a day, so you specify a pattern and the scraper generates all the URLs that you can then download.

B

So it's useful for the uni down module. So when you're dealing with data in ftp and.

C

There are in modules that already do this.

B

Not the dahveed found clearly I.

H

A

B

Right and everything yeah, it's not it's, not a particularly complicated, so I mean for the event for the. What what we're talking about here, a couple of hundred lines of code I'm, not convinced that a new dependency would necessarily be worth it.

B

So also I think it's. This is specifically yeah. This is specifically date based looking at this, so.

H

Because this just a create a set of files- and it's.

B

Creates a set of URLs.

H

Self URLs, so it's just a big canoes string exercise.

B

Yes spring and you up knowing it only returns, URLs that actually exist on the server, so it uses beautifulsoup to query the server to work out what files exist and returns you from the five.

H

So there is a package called Skippy you're, a terrible game or scrape a crappy.

D

H

B

D

B

C

H

C

Yeah I mean I, mostly agree with you Stewart in terms of like new dependency. On the other hand, this is code that we would have to support right over time to fair scrappy looks.

B

Way way more complicated, oh yeah,.

C

It was a pretty general problem, it.

B

Is but it is, a postage is a reasonably specific implementation again that is specifically dealing with time strangers.

B

Yeah also Albert, if you don't jump some pie back in it will kill the broadcast.

B

What, if the some pie, G+ page, isn't in the hangout for a certain amount of time it would kill the broadcast okay. It's.

D

A big syke, refugio company and.

B

B

Mean I'd meet I. Obviously, I need to review this properly, but, like I've got nerve. I got no nothing against it. In principle, I think it will be quite useful and even including tests you own just about breaks three hundred lines of code. So it's not like we're taking on a monolith. I.

C

Would argue and I'm gonna put this as a comment that this code is not useful unless it's incorporated into the rest of the Sun pencoed like.

B

Well, I mean first of the google summer of code project. Doing the SI t data at the moment would will find this useful. That's.

C

Fine, but this should be used in uni down somewhere sure.

B

But it will be like it fit. Eight, it's on a source by source basis in you down, not globally in you del, because the sources of the issue, you need things like the error and goes and sit or whatever comes next generate the URLs that you need and then downloads. So it's a source by source problem now and obviously everything that we've written in philly down at the moment works by a virtue of the fact that we didn't have this when we wrote them.

B

So I mean I, don't I, don't really have a living against and against this in principle, I.

H

Can see that this would be no other than using an outside package if we have a small amount of code which does something extremely to kill, that was having it in maintaining it, as opposed to using something like scraping, I.

B

Mean I think it would take a reasonable amount of effort to implement some this with scrappy anyway, because you've got you've, got a deal with time and date unless there is something in there that a specific date and time.

B

Man, this is scotty's much more of a generic web scraper than it is.

H

H

Should use this for you love you.

B

Hello screaming yeah.

H

Well, although like so, we had to take the same solution of weight since we define how the packages are or have the details later, when you're downloading the detective doesn't let me just draw a custom scrape up to so yeah. We had the same problem in the same solution. I.

A

Think promise I fulfill them.

B

So I think that scrappy is more about following links to get more links. It's just it's a web is a web cooler like google bot, okay,.

H

So maybe like these, these solutions are too heavy.

B

Yeah, where is what that means done these? Basically a URL generation routine that never occurs through a few directories like okay,.

A

Yeah I mean what the v's think seems that I would say that scraping doesn't even seem like the right word for this yeah. It's just like a huge shotgun blast to find everything that exists with us that fit certain pattern. Yeah.

B

I mean these dice, but scraper is probably an apt description in that it is actually going and asking the websites what urls exist and then scraping them and return them to the user bout five. But yes, I see your point that it's exceedingly specific scope of its.

C

Grey, we talked about this non 0.6 issue long enough well,.

A

The veba wants us in by because I'm.

B

Going to say, the G sub project could probably use this. It well.

C

What's stopping them from using it, but.

B

I mean nothing and they were they are going to use it, but I mean if we can get it into bill. Point 6 and you'd be nice to get into the concepts I will attempt to put in a branch or something I mean it's fine. Are we? But if people can review it before the end of the week, we can get into Northman.

D

Tips, there's nothing scheduled ever a myriad of it slide apart. So exactly barbour.

B

Take a look at the other time, I'm going to remove the working progress label them to sleep with you yep. That sounds good. There are still lots of pomp or request for review on it. How old things actually review? No see this one shouldn't be review anymore, because that's probably in fact over the closest progress, because I am never going to find the time to finish it go away own the landscape by 0, 1.

B

Half page finger verification tests, sometimes finished I.

A

Mean it's it's usable, so what I should have been finished? It's a question of whether anyone inside from me and partial you have looked at it. Ahaha well,.

B

I review did it, we won so this was in now. We should get this into 2.6 and.

B

Puzzle visit I thought this. You gone.

A

The main question, the main reason that I have any hesitation on this really is that the actual process for updating a library is a little bit weird yeah.

B

But that can be fixed. It's not like we're using it very much of the moment, maybe nice, to have the framework in place for your 16. That's right.

A

B

A

Always actually it.

B

Doesn't really matter because it's not really user-facing anyway,.

B

It would be nice to merge this soon, be it before or after your point. 6 doesn't actually make the difference, but it would be good to get it in there's.

A

Really no reason to wait on it because it doesn't hurt anything to have it in there. No.

A

Okay, I actually have a phone call two minutes, so I have an overwhelming urge to come to bed. Everyone look at PRS, let's target 0.6 by the end of this week, so anything that's review can go in review too. Well, I can go ahead and anything. That's not ready. We'll just have to wait till next point release. If it's not api changing or you know 0.7, hopefully we can do that on more timely basis should be nice. Okay, you guys can keep talking if you wish, but I really hang up thanks, guys kiss.

A

Thank you cool.

D

H

You later yep all right thanks.

C

Great thanks, everyone.

H