[X] My Assistant

Decreases filesize by maintaining low bitrates during low movement and increasing it during intensive ones that move a lot.


Worked really well during the time one piece went from SD to HD.

This post has been edited by Pillowgirl: Apr 30 2021, 00:32

(IMG:[i.redd.it] https://i.redd.it/7285bvqz43s01.png)
I hate whenever I have a Javascript thing I'm struggling with and then all I get is JQuery answers.
Fuck React and Vue, though, pure Javascript is all I need (though a Lisp would be nicer).

That said I haven't had to look up any answers recently, except for with 'promises' and other asynchronous stuff

What's the point of having optic fiber when the servers you download from apparently are still on bad adsl upload speed (IMG:[invalid] style_emoticons/default/mellow.gif)

What's the point of having optic fiber when the servers you download from apparently are still on bad adsl upload speed (IMG:[invalid] style_emoticons/default/mellow.gif)

Lower ping?

I once had a 3gigabit connection, it was pointless for most downloads.

I'm happy with 1mbit, that's around 115kb/s.

I'm happy with 1mbit, that's around 115kb/s.

*shakes pillowgirl*

Is that ADSL 1?


I just hope where i am moving has at least ADSL 2+

oof

*shakes pillowgirl*

Is that ADSL 1?
I just hope where i am moving has at least ADSL 2+

I'm saying i'm happy with that, not that i have it.

Someone linked me to an incomplete patch to get V8 (Node.JS and Chromium's JS engine) building on 32-bit PowerPC (big-endian).

So now I'm wrestling with that and trying to fix the places where it's broken as I go along. The outlook isn't good, but we'll see if it finishes building this time around. After applying the patches I've only had to manually change one header so far, which makes me nervous.

Most likely, if it does finish building successfully, it'll just try to run some illegal instruction and die or segfault and die when I run it.

It doesn't help that the patch set is from the Chromium project, but I'm trying to build Node.JS which uses its own fork of V8 that doesn't follow the upstream very closely. Also the patches against upstream chrome have been reviewed, but never got applied (they were submitted a year ago), so unless something changes they're dead in the water and will never propagate to Node via normal means.

My guess is whoever it was who hired the guy who wrote the patches let him go, or something.

This post has been edited by dragontamer8740: May 3 2021, 02:40

I'm saying i'm happy with that, not that i have it.

I think I'd like at least 250KB/s download.

64kb/s to 100kb/s was basically painful when I was stuck on it for a while.

nvm

This post has been edited by Necromusume: May 3 2021, 09:36

Both plastic switch coupling in the station monitor scope broke. The shafts are grounded where they exit the RF pickup section, and again at the front panel, so just ordered brass replacement couplers from amazon. Old ones were an overengineered plastic ones that can take up some misalignment and dampens torque- probally just to make sure the shafts don't bind in the front panel bushings any.

Speaking of amazon, got a couple of cheap BNC jumpers to cut up, and they were made of cheese. I needed a BNC to RCA and RCA to RCA cable, both with 50 ohm coax. This jumper had coper clad aluminum braid that refused to solder, jacket and dielectric that melted like nothing, 75 ohm BNCs, and the ground wasn't even soldered to the shell under the molding. Threw out the one I was trying to use, and the unwrapped one. Not sure what I expected for a $13 2-pack of 10' jumpers, but cmon.

Also improvised a hand mic, as I can't source the OE one. Got a 500 to 50K matching transformer from a cheap dynamic mic (PA, not radio style), an old dynamic hand mic I liked the case of, and already had a good cord from a 3rd one (bought just for the plug cause that was cheaper *shrug*). Used a plug and the matching transformer from the first, the case and element of the 2nd, and heavy duty coil cord of the 3rd. The phono plug from the first mic + a chunk of it's cord went on a morse key I have, for I ever learn morse code.

This post has been edited by Wayward_Vagabond: May 4 2021, 18:07

Both plastic switch coupling in the station monitor scope broke. The shafts are grounded where they exit the RF pickup section, and again at the front panel, so just ordered brass replacement couplers from amazon. Old ones were an overengineered plastic ones that can take up some misalignment and dampens torque- probally just to make sure the shafts don't bind in the front panel bushings any.

Maybe I'm misunderstanding, but couldn't rubber grommets work, too?

Speaking of amazon, got a couple of cheap BNC jumpers to cut up, and they were made of cheese. I needed a BNC to RCA and RCA to RCA cable, both with 50 ohm coax. This jumper had coper clad aluminum braid that refused to solder, jacket and dielectric that melted like nothing, 75 ohm BNCs, and the ground wasn't even soldered to the shell under the molding. Threw out the one I was trying to use, and the unwrapped one. Not sure what I expected for a $13 2-pack of 10' jumpers, but cmon.

Looks like you got exactly what you didn't ask for.
Did they advertise this as a 50 ohm cable? Lol.
That being said I'm like 99% sure all my video gear is actually using 50 ohm connectors/cables, even though it's pro gear. I get the feeling the industry does not care.

Are you just determining it's 75 ohms because there's no dielectric present? Apparently that's not a reliable indicator. Wish I knew how to actually do it, though.

Nice job with the mic, btw, sounds like a fun project.

This post has been edited by dragontamer8740: May 5 2021, 00:27

Yeah, 75z based on no dielectric. Coax is RG-58, which is 50 ohm.
RCA connectors don't really have an impeadance, and it's just for low power IF lines, so wasn't concerned about the mismatch for that application- but the coax refusing to solder into the RCAs wasn't a thing I could work around.

Shaftt couplings connect a long aluminum shaft to the steel one built into the rotary switch. The long aluminum shaft runs to the front panel, and out a hole in the casting.

the coax refusing to solder into the RCAs wasn't a thing I could work around.

You already tried to sand them down and flux them, I assume?
I've soldered to coin cell batteries (don't do this) with that technique.

Shaftt couplings connect a long aluminum shaft to the steel one built into the rotary switch. The long aluminum shaft runs to the front panel, and out a hole in the casting.

Ah, I forgot; I did see that in the picture. My mistake.

Recently accidentally ran into system config screens for the mouse, and ended up going back to monochrome cursors (no more spinning blue circle)

More flux didn't help. The wire would take a little bit of tin, then no more. I think the thin copper plating was burning/flaking off and exposing the aluminim.
The RCA connectors readily accepted solder, and were attached to some solid center RG-58 I already had.

Doing another misguided and stupid thing and trying someone's old patch set to build the IBM J9 JDK for 32-bit PowerPC. I'd do nearly anything to get away from the pure-C "Zero" OpenJDK JVM, especially since IBM's J9 for Java 6 works so well by comparison.

It's not an official platform anymore, though. We'll see what happens if I try this. Assuming I can even get the build started.
Update: build started. Since the build requires an existing JVM to bootstrap from, I have to use the no-optimizations JDK8 that I already had installed; it'll probably be a while.

This post has been edited by dragontamer8740: May 6 2021, 05:28

Got farther with the OpenJDK build before it died, i think i know how to fix it this time so maybe it'll get there eventually.

Also I compiled Joerg Schily's "schily tools" today just to try the CD burner software, discovered he also has a bourne shell forked from OpenSolaris. It looks pretty interesting.

I was happily surprised to find that I was able to build all of his tools using David Korn's 'nmake' from an old AT&T AST release (where ksh93 is from), instead of using Schilys own 'smake' program. Go figure. Apparently smake isn't even derived from the Korn source code, but it is a 'workalike.'

I built the korn shell and various other AT&T AST tools a few years back; that's why I had nmake floating around.

Schily tools also has its own 'tar' variant and stuff. Kind of interesting.

This post has been edited by dragontamer8740: May 7 2021, 08:07

Thought:
The "change size of the mouse cursor" thing of Windows may be surprisingly useful when I read 100% text things for afar with a 300% browser zoom.