_thebrain_

joined 2 years ago
 

What I am listening to seems to ebb and flow with the time of year. Early September seems to be violent Femmes for whatever reason.

[–] _thebrain_@sh.itjust.works 1 points 1 week ago

Granted the forms of mfa available without a hardware key are far from perfect. But they are better then no extra authentication method.

A good plan violently executed now is better than a perfect plan executed next week

[–] _thebrain_@sh.itjust.works 1 points 1 week ago

As have I... But it's so dang handy. From what I know there is no alternative that is quite as good

[–] _thebrain_@sh.itjust.works 5 points 1 week ago (5 children)

jdupes: it's great software. The author left GitHub not because of Microsoft, but because he refused to implement 2fa on his account, which GitHub made mandatory.

[–] _thebrain_@sh.itjust.works 16 points 2 weeks ago (6 children)

In my limited experience the speed a rust complied executable runs is highly dependent on compiler options. By default (from what I remember), rust includes a ton of debug info in the resulting program. With the correct compiler flags you can strip all that out and programs run very close to c speeds.

[–] _thebrain_@sh.itjust.works 11 points 2 weeks ago (1 children)

That is a lot of effort to say that you don't care....

 
[–] _thebrain_@sh.itjust.works 8 points 3 weeks ago (1 children)

The number of .gov email address found in janitor section of the DB dump when the site was hacked not too long ago perhaps.

[–] _thebrain_@sh.itjust.works 4 points 4 weeks ago

The only thing I figured out before the main characters (so far) has been

Tap for spoilerThat the alien, Jules Vern was disguised as the fra during the mesal meetings.

There was a ton of foreshadowing but I'm still pretty proud of myself .

[–] _thebrain_@sh.itjust.works 3 points 4 weeks ago

Maybe your right and I am coming at it from the wrong angle.

But several side discussions in the book have prompted major philosophical discussions with my daughter... Primarily the one about 3 being a prime number. Is that just how he universe works? Or it only prime because of the way we define a prime number? I think it was like a half paragraph afterthought in the book but it is an interesting thought experiment.

[–] _thebrain_@sh.itjust.works 9 points 4 weeks ago

It's funny... I started Cryptonomicon thinking it was going to be daunting... But I tore through it in about 3 days it was just so captivating. I think I have reread it 3 or 4 times.

I can tell you now, no matter how Anathem ends, satisfyingly or not, I doubt I will ever pick it up again.

I tell almost everyone, scifi fans or not, they should read the cryptonomicon tho.

 

About 15 years ago I read though everything Neal Stephenson had written. It started with snow crash, then the cryptonomicon, then less popular works like zodiac and the diamond age. In general I loved them all. I stopped reading his books right before anathem came out...

But on a vacation last year I picked up a copy on my e-reader and started it. I enjoyed the concept, I loved the mat-as-philosophy, despite the difficulty getting into it. Anyway, vacation came to an end, I was about 40% through it, and I just stopped, picked up a couple other books, and moved on. No real reason, but a combination of how slow it moved, the extraneous details that seem like they could have been left out, etc.

Well, vacation this year hit so I picked it back up. After about the 50 percent point the book totally changed, as if I was reading an entirely different story. I'm trying to leave out spoilers, but now I am about 80% through and I'm having trouble with what the characters are doing in the book based on the history provided in the beginning.

Tap for spoilerLike, these monks who shunned technology are suddenly flying space suits and plotting the takeover of an alien ship.

Don't get me wrong... I follow the story and the plot tracks well, but there was so much character development in the beginning and suddenly what they are doing doesn't track with any of that development.

The best summery I read by someone on reddit said "what do you think about Anathem? I think it is about 200 pages too long". I'm going to push through and finish it. My e-reader says I have about 7 hours to go and I think I am enjoying it. I just was wondering how others feel about it.

[–] _thebrain_@sh.itjust.works 2 points 1 month ago* (last edited 1 month ago) (1 children)

His Divine shadow.

Edit: spelling

 
 
 
 

This is my 400mm 110v silicone heat made on my bed. Suddenly nothing would stick toy bed any more, regardless of temp, offset, and adhesion assistance agents. I finally looked at the center of the bed was puffy. I was able to slide a ruler in between the aluminum bed and the pad. It took a bit of elbow grease to pull off the edges. Dunno how close to disaster I was but it's a little too close for comfort. It never overheated in all my use and the thermal fuse is still in tact but the scorch mark is still disturbing.

 

This post is a follow up to this post here Quality issues where I was having tolerance problems with lead me to discover inconsistent extrusion in base mode.

It has been quite a ride and I used a lot of the suggestions given in the last thread, and also discovered several other problems which lead to a major rebuild/redesign of the corexy x and y axis, bed motion, and extruder. Here is what I found:

The first thing I noticed were the shafts in the x/y stepper motors appears to be bent just slightly. If I watched the pulley on the motor it appeared to wobble. I replaced both motors with some spares I had laying around from various projects. The x and y gantries both move on linear rails, and I ended up totally reprinting the carriage adapters as well.

Second, I had 4 points leveling on the bed, which I know is quite controversial. I found that the z screw coupler in one corner has become stretched out and wasn't really adjusting height correctly when I ran my 4 point script. I modified the bed so two steppers are connected in series and moved the pivot point between the 2 old pivot points. The bed moves on 4 linear rails and 4 smooth rods with z-screws between the rod and the rail. I got rid of the spring type couplers and replaced them with solid couplers. Any wobble in the screw is taken up by wobble nuts.

Third, and probably most important, I realized the zesty nimble 2 extruder was suffering plastic fatigue. If your not familiar, the zesty nimble is 85%-90% sintered nylon. There is one hobbed gear and a "breech" that holds the filament against the gear. The breech clips onto ears on the main body of the nimble. One of the ears, over time, has gotten a bit weak and the breach only consistently clips into one side causing less pressure to be put on the filament against the extension mechanism. I could have inverted the mount, reprinted it, and flipped the nimble as you can use it either left handed or right handed, and bought some time, but zestytech has gone out of business, so I decided it was time for a change. I picked up a shiny new orbiter 2.5, and printed a new hot end carriage, a hero me, and totally rebuild the print head.

I spent a couple of hours updating and doing basic tuning on the config and am much happier with the results. I still need some more tuning, need to work on wire management, and come up with a better wire support instead of the old number direct drive cable. I may eventually reuse the number as a second print head extruder but that is a ways off.

Here are some of my first prints from the new setup.

I still need a bunch more tuning: extruder stepper vref tuning, retraction, print temperature, feedrate, etc, but I feel like I am finally getting somewhere.

 

I have a printer I have basically built. It is a tronxy frame (and corexy motion), but over the years I have added linear rails, a duet 3d control board, quad zscrew independent leveling, a zesty nimble extruder, and other things.

My printer was down for 6 months or so, a new kitten we got ate some wires and it took me a bit to get the motivation to rebuild it. It turned out just a couple of things got unplugged and I was quickly back in business.

Ever since starting printing again I noticed that my tolerance has been off and it seemed to be over extruding quite a bit. I tuned my steps per mm, and driver power on the extruder motor to no avail. Eventually I replaced the motor and the nozzle (which was perfectly in check but if I was going that far I figured why not). It seems to have solved the problem... Sort of.

I have been using the prusa procedure to test and tune extrusion multiplier: print a 40x40x40 cube in vase mode, and measure the wall thickness. My extrusion width is .45mm, and until today I was getting a width of .52-.54mm. the replacement parts have cured this, if I measure in the right spots with my micrometer I get .45 exactly. But I have these bands. If I measure the high spots on the bands I get the same .52+ mm.

I looked at some of my old test boxes:

I have the same bands but different patterns.

A little googling and someone suggested (for a similar problem) that belt tendon was unequal. From what I can feel my tension seems to be the same.

I'm printing a tolerance test now to see if my issues are fixed, but I only feel 50% confident. Does anyone have any advice as to what might be going wrong? My belts have been on the printer since upgrading to linear rails and could probably use changing... But hopefully someone else has an idea?

 
 
view more: next ›