loony rants

[bubby]

Here are some fun numbers to compare and ponder.

1/4# of D97 has about 1675 feet on it - per BCY site - $23.00 1/8# avg
1/4# of B50 has about 1075 feet on it - per BCY site - $13.00 1/8# avg

D97 needs 6-10 strands - 8 is most common
B55 needs no less than 12 and most often 14-16 strands

Which costs more per string?

On ebay you can generally get d97 for $14 a roll, figure up the price now

[loon]

The graduation ceremony is tomorrow. School is over. I'm getting my B.S. in Computer Science.
I left the maple board laying hidden behind a building for move-out, since there was no space. Hopefully I can get it back.

[Hawkdancer]

Loon,
Congratulations on the B.S.!  Degree, that is !   Should be a very good career field to support the archery addiction .  Puts you in the geek squad, but it pays better than most fields. We may have to consult you professionally😂👹
Hawkdancer

[loon]

Thank you! I'd love to help out.

[EdwardS]

I was forcibly retired after 16 years in IT.  If you have questions, I'll help as I can.  Been out of the game a few years, but I still build some masterpieces.

[loon]

That sucks.
I've had some burn-out (startup..), but CS still seems to make more sense to me than any other careers. And some fields/languages are more enjoyable to me. I think I want to look further into F# and computer security, though I know the latter is very tough. I've gone to a few CTFs and could do a lot better, but haven't given up yet. I think you could find a job if you got good at security..?
What sort of stuff do you build?

[EdwardS]

Let me suitably geek out here...

The setup I'm using to post this is a x79 i7 with 64 GB of RAM, a Titan Black and a full watercooling setup.  For storage it runs a SSD RAID10.  My newer computer is a Bulldozer, 16GB of RAM and a 1070.  I've still got a quad-socket Opteron setup, thinking about turning it into a coffee table.  I spent more than twenty years doing stuff like that.  Kinda hard to quit now, but the archery business gives me new stuff to work on.

[loon]

Let me suitably geek out here...

The setup I'm using to post this is a x79 i7 with 64 GB of RAM, a Titan Black and a full watercooling setup.  For storage it runs a SSD RAID10.  My newer computer is a Bulldozer, 16GB of RAM and a 1070.  I've still got a quad-socket Opteron setup, thinking about turning it into a coffee table.  I spent more than twenty years doing stuff like that.  Kinda hard to quit now, but the archery business gives me new stuff to work on.

Cool. My most powerful machine is just an i5 4460, 8GB of DDR3, a GTX 960, air cooled, an SSD w/ the OS and an HDD. Though my Thinkpad T530 has an i7-3630QM, and also 8GB of RAM. I was being foolish and thinking that it'd be okay to build standing in carpet (to avoid being yelled at by putting it on the dinner table...) if I just kept touching the case, now I get some graphics artifacts. Could've been worse. So now I got an anti-static wrist band and swore to never touch the circuits while on carpet again.


Today, I ruined what I thought was a very nice 3/8", 36" long poplar dowel. I got tired of reducing it with a rasp and tried using a knife. From now on I'll only use the knife for the area nearest to the nock. I can't seem to control using my knife on wood. It just always goes too deep and makes nicks. So the previously stiff shaft ended up being floppy like a wet noodle. And I should've figured out where and how much to taper and used calipers, instead of trying to just do it until it 'looked good'.

Considering making a tapering jig, but I barely have enough space for more jigs in here.

The shaft on drill thing is pretty slow, it only seems to work well for smoothing the shaft a bit more after using a rasp. Then maybe a card scraper could be used, but I don't seem to be good with scrapers either, always leaving nicks and parts of it going too deep at the end.

[Hawkdancer]

I drilled a piece of hard maple with a 3/8" Forster bit and used a gooseneck scrap to get a fit.  I will try a thumb plane on the thicker shafts to get them to size.  Also have to find an 11/32nd bit for the next down size.  Gentle taps with a small mallet help a bit, too, on the maple, not the shaft!   Easier on the fingers that way
Hawkdancer

[loon]

What do you do with that maple piece, though? Just a piece of board with a hole drilled on it?

edit to avoid bump: cool thanks, I see. so for bamboo shafts

[Hawkdancer]

Right, I drilled about an inch or so from the end of the board and use the hole to size the shaft, also, seems to help burnish the wood a bit.  It takes a bit of work, especially around the nodes, but it works, I think.  Got to sinew wrap and cut the nocks yet to see how uniform the set shoots, working on a half dozen.  Spine is pretty heavy right now, but my other set of POC are spined heavy also.
I will mark the hole size before I drill a smaller hole(11/32).   Don't think I want to go any smaller.  I start the shaft and keep scraping, forcing,  making with pencil, tapping loose scrape some more, force
Keep after it 'till it fits through all the way. 
Hawkdancer

[loon]

Should I get a hatchet for reducing width profile? (ie in bamboo as well as wood) Maybe a single bevel one. Thing is i already have trouble controlling my knife with a symmetric scandi grind. and I'm not very good at sharpening. I'll probably just use my ryoba saw and shinto rasp for now.

todo:
spine tester
blunt bareshaft bamboo and sitka spruce arrows
red oak long pyramid
maple D bow - edit 7/6/2017 yeah, we got the maple board back... still haven't started. need to back with with something eventually too.


That garage is like a toaster oven... but I can deal with it

Also want to make a chair. And maybe a portable bench/'horse' with a vise i can take outside

[loon]

Code: [Select]

from bs4 import BeautifulSoup
import requests

ua = "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36"
url = 'http://i984.photobucket.com/albums/ae321/isaacscr/Misc/HPIM5242.jpg'

headers = {'Upgrade-Insecure-Requests': '1', 'User-Agent': ua, 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
           'DNT':'1'}

requests.get(url, headers=headers)

IT WORKS?!?!? Before putting in all those headers, it just gave me that blasted "give us your money" image. This is the first step. Making it redirect to the actual PhotoBucket photo page. Then hopefully somehow being able to actually pull the image. Then I'll build a dictionary of image links to other image links so that all the PB links can be replaced.

I think I just need to get the cookie it gives me and send it back while requesting the same thing again.

I... started making a spine tester. Haven't been very productive.. the garage isn't a very pleasant place.. It's a big board, the supports for the arrow are pieces of wood glued on with TB3.

I guess I should taper one of the Sitka spruce shafts from Wayne, and put a wooden blunt head on it for short range practice. All it takes is one arrow with no fletchings to be able to practice well...



edit: This saves the proper image from a given PhotoBucket URL. BeautifulSoup isn't even being used yet. It works... next, just write some code to extract all the PhotoBucket paths from a forum thread or whatever, and repeat this process. I should probably add diagnostic error checking stuff as well. This would probably be easier with a browser macro, but meh. Slower.

Code: [Select]

from bs4 import BeautifulSoup
import requests
from io import BytesIO
from urllib.parse import urlparse
from os.path import splitext, basename

s = requests.Session()

ua = "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36"
url = 'http://i984.photobucket.com/albums/ae321/isaacscr/Misc/HPIM5242.jpg'

# haha i'm totally Chrome
headers = {'Upgrade-Insecure-Requests': '1',
           'User-Agent': ua,
           'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
           'DNT':'1'}

s.headers.update(headers)

req1 = s.get(url)
s.headers.update({'referer': req1.url})
img_req = s.get(url.replace('http', 'https'))

img_data = BytesIO(img_req.content)

img_url_path = urlparse(img_req.url).path
img_filename = basename(img_url_path)

with open(img_filename, 'wb') as out:
    out.write(img_data.read())

...
photobucket is a slow piece of trash, so it times out sometimes :\
never liked photobucket...

[loon]

getting late

more pieces of the puzzle

Code: [Select]

from bs4 import BeautifulSoup
import requests

pa_session = requests.Session()
PA_THREAD = "http://www.primitivearcher.com/smf/index.php/topic,27206.0.html"

thread_page = BeautifulSoup(pa_session.get(PA_THREAD).content, "html.parser")

thread_pageno = int(thread_page.select(".pagelinks strong")[0].string)
print(thread_pageno)


print(thread_page.select(".pagelinks")[0].find_all("a")) # can increment thread_pageno and try to find if it's here to keep going through pages

pb_imgs = [img.get("src") for img in thread_page.find_all("img") if "photobucket.com" in img.get("src")]
print(pb_imgs)

will continue tomorrow...

my spine tester's coming together. need a weight and a wire coathanger for a hook etc

[loon]

so close. with this i can just make a loop that keeps fetching images rather easily for all pages in a thread.

Code: [Select]

thread_pagelinks = thread_page.select(".pagelinks")[0]
thread_pageno = int(thread_pagelinks.find("strong").text)
next_page_link = thread_pagelinks.find("a", text = str(thread_pageno+1))
# if next_page_link: repeat..
filename of saved image should be a hash of the URL, and i will store... probably pairs of photobucket image URL to new URL, after images are uploaded

What DB do the PA forums use? MySQL? Postgres? I think I'd need a query somewhat like

UPDATE posts SET text = replace(text, "http://photobucket.com/blah/blah.jpg", "http://example.com/blergonagr.jpg");

on postgres for every single image that is to be replaced. Should do a DB backup beforehand... maybe there's a better way to do it with less queries or prepared queries or something


This would actually be easier for me and cleaner if I was given a list of all photobucket img links, which could be acquired from a database query as well...