So I’ve been tracking the iPhone 15 rumors and I think the best looking renders so far are the ones that keep the antenna bands as a straightened-flattened ribbon wrapping around the phone with the back glass being curved itself at the edges. It will then look like the two different pieces have come together to make up the overall phone. This, along with the flush camera lenses, which dates back to the beautiful iPhone 4S design, would help make it look a lot more stylish as well. I will greatly miss the silencing physical dip switch if Apple decides to remove it but let’s see what they end up doing! 🙂
I didn’t post the source code for this one as it was more of a highly specialized project that I thought applies for my particular use case and requirements.
So one evening recently I got an email alert for a big spike in traffic going to this blog. I don’t know why it happened and I don’t see my post view counts getting that high either so it was a strange event. Thankfully Cloudflare’s service was able to handle the total requests being made as I don’t have much visibility on the WordPress side of things. My blog setup here is a bit simplified unfortunately!
Well, my grandpa had a saying that I remember til this day, he said “son, nothing good ever happens at 4am”. So last night, at 4:15am, 2 folks stole my license plate off my car and they could be out there committing crimes in my name!
So I usually try to save my money where possible and I don’t celebrate my birthday but I wanted get an updated gaming console so that I could continue to play Fortnite when I have some free down time. Way back in the day, my grandpa gave me an original Nintendo (which I sadly didn’t think of keeping over time) and eventually I got the Nintendo 64 when it came out as well. I then purchased the original Xbox (which had the Halo title on it) and that was one of my first Microsoft products that I enjoyed owning. After that, I switched over to the Sony side and purchased the Playstation 2 – and then the Playstation 3 – and then the Playstation 4 (so that I could keep playing the different GTA versions).
For my birthday this Jan 2023, after spending years playing with the original PS4, it was getting a little old and hot and dusty so I purchased myself a gift of the Playstation 5 (to be able to play the Fortnite battle royals). However, once I turned it on, configured it, updated it, downloaded the game and launched it – the entire system powered off. I tried it again and the same thing happened at the same spot in the game. I also tried to download a free title that came with the system and it also shut off, right before the game play starts. I recorded a long boring video of it because I couldn’t believe it!
This new unit left me in such disbelief that I went back to Microsoft and purchased an Xbox Series X instead! I have to say that as a long time Playstation fan, I was pretty impressed with the build quality of this gaming console. It has a good weight to it, it’s less bulky, has proper ventilation holes near the back bottom for cool air and the main fan near the top has nice big blades which quietly moves the air upwards and outwards – it’s properly engineered and designed – it reminds me of a little chimney stack! My only improvement would be if they could just round out the corners with a small radius to make them a little more user friendly and less sharp feeling/looking 🙂
The PS5 is fairly big and bulky, the fan blades are way too small sized and it doesn’t really have proper ventilation holes or airflow direction – and according to my online searches, other people have also been facing power / heat issues with it. The only downside to the Xbox is that I highly prefer the PS5 controller over it as the bumper/trigger buttons are less bulky, the joy-sticks are symmetrically aligned, the D-Pad is in the proper location and the overall feeling/comfort is much nicer while holding it. I was even able to buy a little add-on attachment for the Xbox that allows me to still continue to use the PS5 controller which is amazing!
The ultimate gaming console setup for me in 2023 is the: Microsoft Xbox Series X ++ Sony Playstation 5 Controller
Well, this year has been a long and crazy one for me and I wanted to upgrade the style of my first mechanical keyboard. Even though it’s not my fav one to type on, it is still a memorable part of my collection and it’s a great WFH keeb. I bought a wood case which raises it up a little but it’s much more solid, better sounding and nicer styling compared to the cheaper, thinner, and hollow plastic case that comes with ducky keyboards!
This was my first mechanical keyboard build and they have made it a pretty smooth process (took me half a day to lube the stabs). This is the KBD67 ANSI hot-swap and I put in a mixture of everything into it. It has holy panda switches for the alphas, halo true switches for all else, and ducky plus drop keycaps. It has a pretty good typing feeling, feedback, and sound to it – it’s a little lower in pitch compared to the Drop Alt’s open top metal frame (I think the plastic and padding underneath help deepen the sound). It’s a great overall keyboard to have as part of the collection! 🙂
So nginx has a stream proxy module that you can use for transparent SSL/TLS relaying/forwarding, however, it is only capable of reading the SNI hostname upon the initial handshake of the connection. In addition, the destination IP address is replaced because of the firewall redirect pointing to the proxy server. I wrote a small modification that can be compiled into nginx which allows you to run a script that can pull the missing destination IP address from a given state connection table in a firewall, for example pfctl or iptables.
[error]: no host in upstream ":443", client: 192.168.X.Y, server: 0.0.0.0:3129, …
This code mod above will allow you to run a shell script of your choosing if nginx cannot get the hostname or address of a connection requesting to be proxied. You can then look up the destination IP address based on source IP + port combo from the connection state mapping table of the firewall. The result is a much more stable proxying experience for HTTPS connections without needing to wait for the SNI or hostname of the initial handshake!
So I was a little late to the game on mechanical keyboards but I tried my best to catch up in 2022 and so I could try to make it into 2023. The first one I got was the Ducky-Mini with Cherry MX Brown switches which offers a bit of a quieter/smoother version of a tactile switch or a rougher/scratchier version of a linear. This is a good wfh keyboard as it’s not as annoying to type on while in the middle of meetings and it offers more feeling during key travel compared to a linear in addition to lesser sound compared to a clicky. I then picked up a Drop-Alt and put Holy Panda switches in it and the feeling and sound matched more closely to the tactile experience which other reviewers were talking about. It’s a solid and stable keyboard and the key press feedback feeling and sound reminds me more of an olden analogue type writer. It’s a much more enjoyable keyboard to type on — a true tactile experience. I lastly purchased a Matias-Mini with Alps White switches that are both clicky and tactile. This is a fun keyboard to type on for personal projects as it has the same characteristics of the Drop keyboard but with a slightly louder and higher pitched sound to it. It’s a great keyboard for all Mac enthusiasts!
The BSD PacketFilter firewall has an extra scrub option which is, “reassemble tcp”. I was researching and exploring the different types of fragmented-packets/segmented-streams of data that could be forwarded within a network that may have a smaller MTU link in the middle of the routing path. I am still reading about what this option does on a streamed session and if Linux has anything similar to it…
Note: nftables has user-land hooks via nfqueue
# nft insert rule ip mangle FORWARD ip daddr 22.214.171.124 tcp dport 53 counter queue num 1
from scapy.layers.inet import IP
from netfilterqueue import NetfilterQueue
pay = pkt.get_payload()
ipf = IP(pay)
nfqueue = NetfilterQueue()
Edit: I found that it was a bit complicated trying to understand when optimized network stacks (software or hardware) will combine multiple TCP segments into bigger IP packet payloads and that trying to perform reassembly at that higher level was a bit challenging/difficult. I came up with a way to solve the occasional web site having slow upload speeds for large files by running an nginx transparent reverse proxy server for HTTP/HTTPS instead!
rdr on en0 inet proto udp from any to any port 53 -> 127.0.0.1 port 3127
rdr on en0 inet proto tcp from any to any port 80 -> 127.0.0.1 port 3128
rdr on en0 inet proto tcp from any to any port 443 -> 127.0.0.1 port 3129
Well, 2022 was a long and crazy year for me (health wise as well) but I’m trying to get back on track and focus and concentrate on new things in life! I’ve been finding that the older I get, the more I think about the olden golden days of being a kid and growing up in an analogue world, before the invention of the internet and computers. Times were simpler but more basic and it’s kinda crazy to think about how much progress and change exists nowadays. I remember a time when my specific search term in Google actually showed me the exact thing I was looking for haha 🙂 Anyway, I’ll try to keep this blog going as a personal hobby of interesting things that I find out along the way in life. Just random tech craziness if you will!
So, I was searching around for the Linux equivalent of the BSD PF firewall rule “scrub-all no-df” and I was expecting that there would be an “iptables mangle prerouting” version – but – no luck! I just wanted to be able to clear this specific bit in the IPv4 packet header with a Linux router so that it would be able to fragment the packets along the inner routes that may have a lower MTU set. Since I couldn’t find anything I made one myself 🙂
Edit: In addition to this Linux kernel nftables module, I found out that there is another one running for IPv4 packet defragmentation which should also be similar to the scrub rule “fragment reassemble”. According to the documentation, if you have any stateful connection tracking rules for layer-4 (for example, postrouting nat rules or prerouting udp ports) then the fragmented packets should get reassembled again. I am looking into the scrub rule “reassemble tcp“…
So I had this Mac Mini that I was originally using as a WiFi relay bridge (layer 2) but since then I was only using it as an rsync backup server. I decided to also turn it into a VPN tunnelling router (layer 3) for my home network setup by using OpenVPN on MacOS (connected to a Debian Linux server). I was testing out its performance and it was holding up pretty good and stable but I wanted to give WireGuard a try as well, just to see and compare. Upon initial configuration, I noticed that WireGuard was defaulting to a lower MTU (1420) compared to what I had set in OpenVPN (1450) and that some of my connections to websites were unstable/slow/hanging. Additionally, with OpenVPN, I was able to set some extra settings like MSS & fragmentation limits and I couldn’t find the equivalent of those with WireGuard. However, I do like the overall simplicity of the WG config and setup process!
I actually lowered the WG MTU (1410) but this could potentially cause larger size packets to fragment. I was searching around and found that it was possible for one to clamp the MSS values on forwarded/routed packets with an iptables forward/mangle rule. Since the Mac is the VPN client and has the BSD PF firewall, I also set a scrub rule as well:
# bsd pf scrub all no-df random-id max-mss 1330
# nix nf iptables -t mangle -A POSTROUTING -p tcp --tcp-flags SYN,RST SYN -j TCPMSS --set-mss 1330
# iptables-translate ftw!
I was also trying to read more about the “don’t fragment” bit that could get set on packets along the way (sometimes when the PMTUD setting is enabled) and the pf-scrub rule can help to clear that bit (no-df) where I couldn’t yet find the equivalent on the Linux side. I also tried setting these sysctl values as well:
# mac sysctl net.inet.tcp.path_mtu_discovery=0
# nix sysctl net.ipv4.ip_no_pmtu_disc=1
I’m still playing around with this setup but it’s been an interesting networking experiment so far in terms of seeing how your web connections handle and react to being encapsulated automatically by a network router…
Edit: I have written a Linux kernel module to do the equivalent of “scrub-all no-df“
Well it’s been a tough number of years for me as all of my personal belongings have been packed up in boxes the entire time, however, I have finally been able to move into a new space and a new place.
As a new decorative piece to hang on the wall, I was able to get a large size, glass print of the MacOS flying Apple desktop background wallpaper I created from Fracture Me. It helps to add a little more fun and colour inside and they have pretty good quality printers (quick service as well).
So after many years of using and learning iptables, one of my favourite firewalls, I had to translate my command line ruleset into a new format/syntax. It wasn’t too bad but with the added power and flexibility of nftables, it can be harder to find the order/priority of the rules for a given filter hook, for example, INPUT. However, I really appreciate how the new firewall incorporates some previous features that I used to have to install via iptables modules like ipset and hashlimit, which provides the rate limiting functionality!