Hacker Newsnew | past | comments | ask | show | jobs | submit | _x5tx's commentslogin

I’m currently in a trip in the south of Brazil, and I can’t open the 2600.com website since I’ve been here. I’m not sure if it is some kind of government filter or just 2600’s firewall blocking Brazilian ips.


Honest question here about jails. I like docker because the tooling makes it super easy to get anything running in no time. How is jails tooling? Let’s say I need to get an app running, a database and a redis for the app. Is that as easy as docker compose?


I wonder if it would be possible to run MiSTer emulators on it.


It would be a significant amount of work. The MiSTer project is specifically targeted at the DE10-Nano so it's designed for that FPGA and that peripheral set. In particular MiSTer relies on the onboard arm to run support software, you'd need to replace that on this (perhaps with a RISC-V softcore you could build the MiSTer software for but lots of software work to do to make that work).

It may also simply not be possible. The DE10-Nano has DDR memory for example but most MiSTer cores need extra add-on SDRAM. The reason being the latency on the DDR is just too high for the accurate emulate most cores are aiming for. So SDRAM is required. On this board you don't have that option. Though perhaps you can build your own DDR controller, with the DE10-Nano you have to use the hard controller built into the FPGA. Perhaps another controller specifically optimised for the latency needs of MiSTer cores could work. Each core would need potentially significant porting work to use this new setup as they get direct access to the SDRAM pins in MiSTer. You'd need to trace back to where they're actually generating memory accesses and then plug that into whatever new DDR controller you had.


In principle, yes, but connecting a display or a controller to the FPGA would be difficult.


If it's on a PCIe card, maybe you can plug it into a motherboard that has USB and HDMI output?


There's also a small amount of IO pins on a high density header. I built an adapter that converts the LVDS voltage levels to TMDS so you can use it with HDMI or DVI - https://github.com/teknoman117/ACORN-CLE-DVI

Side note - this is the smallest pitch component I've ever soldered. Used a stencil and a hotplate since these are 0.4mm QFNs.

Edit - I really need to update the pictures. I forgot to twist the differential pairs before taking them.


The PCIe link definitely has plenty of bandwidth to stream uncompressed video to the host system—and if that host supports P2P DMA, it's probably even possible to push straight to VRAM without a round-trip through the host CPU's RAM. If there's enough room left on the FPGA, it could implement upscaling and CRT-emulating filters to provide a 4k stream to the host.

The downside is that you'd need custom drivers and software on the host system to redirect input events to the FPGA and handle the video feed it produces.


Taking the thinkpad apart is part of flashing the bios, which usually is just clipping an SOIC8 clip to the bios chip and using CH341a SPI usb programmer to flash it. I did the same to my thinkpad x220 to flash it with coreboot.


Thanks for that. I was still wondering why the flashing couldn't be achieved on a running system as-per the usual method of updating the proprietary manufacturer's BIOS. I found this key quote on the Coreboot web site:

There are various protection schemes that make it impossible to modify or replace a firmware from a running system. coreboot allows to disable these mechanisms, making it possible to overwrite (or update) the firmware from a running system.

Usually you must use the external method once to install a retrofitted coreboot and then you can use the internal method for future updates.

https://doc.coreboot.org/tutorial/flashing_firmware/index.ht...


how come it always be an incentive to miners when the price is not profitable anymore?


The mining difficult adjusts. I don't know enough to say what it would take to drop so low that a single asic somehow was too overpowered to be economical, but generally there is automatic matching between the supply and demand of mining power.

As mentioned, 50% attacks are possible at some point (they may even be now). There is not really an incentive for that. I would be curious to see some analysis of how likely they become as the hash rate drops, and match that against the hash rates supported to profitably mine for different prices.


that’s great! I’ve been a long time Fedora user but last year I decided to try NixOS and I couldn’t be more surprised, everything works well and it was easy to install and learn. I’m glad I can try it in my m1 macbook pro!


Ever since I heard about AdGuard Home I never looked back. I have a raspberry pi 4 serving all dns requests from every electronic at my house for 215 days of uptime now without any issues. Browsing the web seems faster and cleaner ever since.


Well, all dns except that over ipv6, DOH, DOT, software using a separate resolver hard-coded to a specific ip... The amount of work to force dns to your dns server these days is nuts.

I use a pi-hole and these days I have to:

- block dns advertisement of my ISP's dns server for ipv6 (which, there's no UI for that on my router so i have to edit the config by hand) - maintain a list of dns over https ips and block them in my firewall (since mitming this traffic is a huge pain) - force all traffic to port 853 to my pi and setup a masquerade for that traffic so the client doesn't know - force all traffic to port 53 to my pi and setup a masquerade for that traffic so the client doesn't know


But at the same time software got slower and slower to the point that every processor upgrade just feels the same as the previous one. Sure there are some speedups, but I feel software could be way faster than it is today. There is little reason for a command line only linux to be slow in an older machine, and yet it is slugish.


That’s an impressive amount of progress already. Congrats on the work. I’m waiting apple’s next silicon to jump to arm, and hopefully by then I’m going to be able to run linux on it.


Natively, it's gonna take a long time, I'd wager. Virtualization works pretty well today, however.


Pretty much all the things except the GPU are working in Corellium's quick'n'dirty "demo" Linux port. And you see how amazingly quickly the GPU driver is coming along… So not that long of a time after all (I'm surprised too).


I'm positively surprised as well. But working well for day to day use… I don't know. But it's not my field, just more or less educated guess.


I went through the same issues 6 years ago, had to create the entire billing system from scratch, and there are so many corner cases I had to deal with, specially with plan changes on recurring subscriptions, had to deal with world wide timezones for the invoice creations and lots of other stuff. I just wish there was a pluggable, robust and flexible solution to use instead of reinventing the wheel.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: