Threads for jummo

  1. 2

    With the nature of Excel it can be used in different ways to achieve the same goal. Because of this the user should know the indendent way to use software from the developer. This might be possible with a tutorial or a video like You Suck at Excel with Joel Spolsky.

    1. 1

      Am I the only one who finds watching that video incredibly painful? There’s so many kludgy approaches to things you could avoid with a program driven more by explicit (typed) commands. The select-copy-paste values pattern to switch from formulas to values is particularly egregious but I think this general approach shows up in a lot of other places. For little 15 entry sheets like this it’s not a big deal, but if I’m dealing with tens or hundreds of thousands of rows the problems start becoming obvious.

      I’m aware that its easier to just ingest sheets into Access (or Base or whatever) or to read the values into a program and manipulate them that way (or at a REPL) but Microsoft (quite unintentionally) developed this relatively accessible data manipulation tool full of anti-patterns that rapidly became close to the lowest common denominator in the field.

      I do appreciate the caution against vlookup: I once worked in an office with ~200 people where some nitwit had decided to create what was effectively a personnel, training, and travel database run entirely off of vlookup and undocumented macros for ingesting CSV files from the official HR databases. The results were a mess.

      1. 3

        I’ve been thinking about Excel being a terrible interface to a great tool that’s just extremely useful in practice (a little bit like git) for quite some time. And I really wanted something more structured.

        In case anyone’s in the same boat, I’ve checked out quite a few projects and settled on Grist (https://www.getgrist.com/) - being able to keep the tables and the formulas, but also keep it in a reasonable format and drop down to python as needed is just perfect access/excel combination for me.

        There’s lots of other programmable docs options out there like airtable and notion, but they never resonated with me.

        1. 1

          Thanks, that is interesting and incorporates many of the features I find desirable for this. I will have to noodle about with it.

    1. 3

      Great Post! Thank you very much! I can recommend to configure the router to reduce Bufferbloat and test with https://www.waveform.com/tools/bufferbloat.

      1. 4

        http://futel.net/ tries to reenable abandoned payphones.

        1. 4

          Cool idea! But no RSS feed for the website.🙁

          1. 8

            Wow.

            It’s disappointing how slow the roll-out of fast fibre broadband in the UK has been. A couple of years ago I was living in central London, with 67Mb/s the top download speed I could get. Of course, this was asymmetric, so the upload speed was even worse. After several months, Hyperoptic wired the building up with fibre and I could get a symmetric gigabit connection, which was fantastic.

            Then I had to move slightly further out, still in the London area and still with a fibre connection, but now the best I can get is asymmetric 550Mb/s down / 35Mb/s up. Yes, this is still really fast, but… it’s so much worse than it could be!

            1. 6

              In Cambridge, I have BT’s FTTP package, which is 900 Mb/s down, 110 Mb/s up. I don’t know why they do the asymmetric thing, I’d more happily pay for a 500 Mb/s symmetric link. CitiFibre is rolling out a parallel fibre network, though it doesn’t seem reliable. Folks I know using it frequently report downtime and often have poor quality in video calls and so I suspect that they’re not getting anything like their promised 1000Mb/s symmetric bandwidth.

              Generally, with 900Mb/s downloads, the bottleneck is elsewhere. A lot of servers top out at 200-400 Mb/s, so with wired GigE I can make a second fast connection somewhere else but can’t get more speed from a single download location. With 802.11ac, the WiFi is more often a bottleneck than the wired connection. I don’t have any 11ax hardware yet, in theory it should move the bottleneck back.

              Upgrading the wiring in my house to handle more than GigE is probably a lot of effort, so I doubt that I’d get much benefit from a faster connection - I only upgraded the switches from 100Mb/s to 1Gb/s a couple of years ago after GigE equipment prices dropped to the dirt-cheap price that I paid for the 100Mb/s hardware I’ve had for 10-15 years. 10GigE switches seem to cost about 50 times as much as 1GigE ones, so I’m in no hurry to upgrade.

              I remember the upgrade from 2400baud to 14.4 Kb/s and then to 28.8 Kb/s as big jumps that made it possible to load images on web pages by default most of the time. The jump to a 512 Kb/s cable modem in a shared house was a huge improvement. First, because it was always on, but it meant that downloading entire videos or Linux ISOs was feasible (though with some traffic shaping at the router, especially so that someone using BitTorrent didn’t saturate the upstream and prevent ACK packets getting through for everything else. I learned to use PF / ALTQ on OpenBSD from one my housemates solving that problem). I was living with geeks and so when the 1Mb/s option came along we jumped on it and had enough spare bandwidth that we could listen to decent-quality Internet radio. I did set up a repeater for Radio Paradise so that we weren’t using half of the bandwidth to all download the same stream though.

              I think the provider (NTL, later Virgin Media) upgraded us to 5 Mb/s and then 10 Mb/s on the same price. That was, again, a big jump because we didn’t need to restrict usage at all. I stayed on the 10 Mb/s connection (by then living by myself) as it went from the most expensive package to the cheapest, and as the cheapest connection went from 10-20-30 Mb/s. Streaming video came out around then and Virgin Media did some annoying rate limiting, which meant if you watched an hour of HD video at peak times you’d be throttled for a few hours. They stopped that after a year or so.

              I think I stayed on 30 Mb/s until moving here. I moved from the cheapest FTTP offering to the most expensive during lockdown when working from home and wanting to make sure the Internet wasn’t a bottleneck (again, mostly for upstream) but 99% of the time I don’t notice the difference. We can play cloud games on Xbox game pass and stream HD video at the same time, but I think you could do that on the 56 Mb/s connection too. Backing up from my NAS to the cloud is faster and downloading games from Game Pass or gog.com is faster (a lot faster from gog.com), but I increasingly don’t install games locally given how good the streaming option is (Game Pass pops up a thing saying ‘Install this game for the best experience’, but I don’t consider worse graphics and longer loading times from my Xbox One S versus the Xbox Series X in the cloud to be the best experience).

              Maybe 3D AR things will drive up the demand again, but since we passed 50Mb/s we’ve been well in diminshing-returns land, unless you have a large family that all wants to watch different HD films at the same time.

              1. 4

                I don’t know why they do the asymmetric thing,

                Sometimes the underlying infrastructure is asymmetrical, e.g. with GPON. But mostly, I guess, the big end user ISP optimize their network for incoming traffic from big content provider.

                1. 1

                  I suspect it’s also to discourage people from using residential connections to operate servers.

                  1. 2

                    It’s usually because residential users tend to consume content rather than produce content. Offering a symmetrical 200 Mbit connection is generally less useful than a 300/100 Mbit connection. This also lets ISPs cost cut more as they try and use available channels for downlink rather than uplink. There’s limits to how far this goes as you definitely don’t want to saturate your uplink while trying to consume content, but that’s typically why.

                    1. 2

                      This is exactly right. People enshrine into technologies and solutions the approaches people are currently taking. This means asymmetry was an engineering shortcut to maximize the usefulness of the technology for what people actually needed.

                      And then the rest of us upload images to the cloud and actually get around to saturating that upload, dreaming of a world with symmetric links.

                    2. 1

                      Also the reason why you can’t get static IPv6 prefixes at most provider.

                2. 3

                  Hello from the North of England! I’m jealous; there are certainly benefits to moving away from London (I lived there for 15 years) but when it comes to internet speeds the saying “it’s grim up north” certainly rings true!

                  Speedtest.net reports 25 Mb/s download, 5 Mb/s upload, and 29 ms ping times for my current connection. And that’s a fantastic improvement since I moved 5 months ago: at my old house a few miles away the fastest connection money could buy was 19 Mb/s down, and just over 1 Mb/s upload. I work from home, and Zoom calls can be rough when others in the house are playing online games.

                  Edit: fixed MB/s -> Mb/s (oops)

                  1. 2

                    I still only get 28Mb/s down in zone 3 of London. Our infrastructure is generally awful.

                    By the way you should know there’s a big difference between “Mb” and “MB”.

                    1. 1

                      I’m not sure where Speedtest.net’s edge is, but 29 ms ping times can be killer for video calls depending on the latency to Zoom’s closest video edge. Is the 29 ms over WiFi?

                      1. 1

                        That’s interesting. Yes, it’s over Wi-Fi. I don’t own any computers with a physical network port in any more, but I can try to see if I can do a Speedtest from the router. If I get better ping times from that I’ll try to stretch a cable via the loft to my office and buy a usb-c network dongle.

                        1. 2

                          On my home network, speed tests tend to read a latency of 35ms ping under load on WiFi. Latency stays lower when I’m using Ethernet (and I’ve corroborated similar numbers using iPerf.) Zoom performance is way better on my home network with an ethernet connection even if I’m the only one using it (many fewer stutters or freezes). When both my partner and I are using Zoom over Wifi, the experience is pretty terrible unless one of us gets on Ethernet (since it’s easy to have frames collide on Wifi, causing retries and latency on the RTP “connections” Zoom uses to send video).

                          1. 1

                            pinging your gateway may also give you a an approximate picture of how much latency your Wi-Fi leg is contributing to your score, but with less effort.

                            1. 1

                              Thanks, that’s a great idea. Running mtr from my laptop to the domain of my ISP yields this for the first two hops:

                                                                     Packets               Pings
                               Host                                Loss%   Snt   Last   Avg  Best  Wrst StDev
                               1. 192.168.1.1                       0.0%    67   26.6   7.5   1.8 124.2  15.9
                               2. fritz.box                         0.0%    67    3.2   5.2   2.5  21.0   3.6
                              

                              192.168.1.1 is a TP-link mesh-networking thing that’s plugged into fritz.box (my ADSL router) with a short cat-5 cable.

                              Walking through the ADSL router’s options looking for a speed-test option it looks like it too supports mesh, so I will try to make it the primary. That might let me discard a hop some of the time? I can see the router itself from half my house, but tend to connect to the mesh. (It has a cooler network name ;-) )

                              1. 1

                                Does your ADSL router have an AP as well? If not then this is standard. Your packet first goes to the AP which then pushes your packet to the router and then to the upstream ISP router.

                                Try running an mtr to a remote and see how much time is spent getting to your AP.

                      2. 2

                        Honestly the state of broadband in the capital was extremely dire 8 years ago. It doesn’t surprise me that you’re not having a good time but I am impressed you’re getting those speeds.

                        I was on 16Mbit and it would die every night. 3 places in wildly different areas had the same awful oversubscribed ADSL thing. I even ranted about it at length: http://blog.dijit.sh/the-true-state-of-london-broadband

                      1. 7

                        Really cool project, and great to see homegrown home automation as opposed to it all running on a “cloud” which can disappear at any time.

                        1. 8

                          Maybe it’s my age showing, but I’m reluctant to deploy home-grown automation solutions because they inevitably break or others in the house get frustrated with them.

                          My garage monitoring solution is an off-the-shelf Z-wave garage door tilt sensor that sends updates to my Z-wave network, which are picked up by Indigo running on a Mac Mini. This starts a timer and sends a Pushover notification to my wife and me if the garage door is left open for more than 10 minutes.

                          1. 3

                            Their is always a risk with external services being not available anymore, like Insteon Shutdown. But I really dislike “face-lifts” or behaviour/feature changes for apps. It’s annoying and most often this change doesn’t have a benefit for me. Something like a rearrangement of products in your favourite supermarket. I try to use simple solution to avoid the not working frustration. For my Hue lights I just used two modes (bright or dimmed) which I can set with a text message to my XMPP bot. I also use this bot to get the current outside temperature or to send me a reminder via email.

                            1. 3

                              Maybe it’s my age showing, but I’m reluctant to deploy home-grown automation solutions because they inevitably break or others in the house get frustrated with them.

                              This was front of mind when building this project. I tried hard to reduce the surface area of breakage in this project by stripping the system image down a lot. Time will tell I suppose :)

                              1. 2

                                I’m with you there. I have the same setup, but with node-red. This particular sensor is on Zigbee, but the door controller is on Z-wave. I want local control, but for technical ownership of these problems I want as close to zero as possible.

                            1. 2

                              Not going to touch this. Especially since it’s in CVS

                              😂

                              1. -1

                                Not going to touch this. Especially since it’s in CVS

                                😂

                                1. 1

                                  Does anybody know a similar improve version for Microsoft Windows?

                                  1. 3

                                    There seems to be a windows port of mtr. I’ve long given up traceroute in favor of mtr.

                                    Plus, with mtr (as root) you can use TCP SYNs to probe the network. Useful for weird (read: “bad”) network equipment and also useful to be quickly blacklisted by the server on the other side (and they’re right to do so).

                                  1. 4

                                    Looks very good but too expensive in my opinion. The Vortex Race3 is also programmable with good keys and cost half of it.

                                    1. 4

                                      Yes, but,

                                      • Low profile milled aluminum case
                                      • Box switches
                                      • RGB backlighting
                                      • N-key rollover
                                      • Hotswap switch sockets
                                      • PBT dyesub caps in XDA profile
                                      • Semi-ergo 75% layout
                                      • QMK firmware
                                      • 2 port USB 3.2 gen 2 USB-C hub
                                      1. 2

                                        Note that the Vortex Race3 does seem to be a very similar keyboard, with a low profile milled aluminum case, RGB backlighting, Semi-ergo 75% layout and DSA profile caps (similar to XDA). It comes with Cherry MX switches though. I’m wondering if part of the cost increase (besides the USB hub) are those hotswap switch sockets with box switches. I’m not sure from a reliability standpoint if that’s even preferable to MX switches. I think the only big differentiating factor to most people is that USB hub. Would you pay double to price to get that though?

                                        1. 2

                                          It’s all about the price to essential features ratio. A Ford Fiesta drives very well and a Volkswagen Polo is much more expensive and … ehm … drives, too.

                                      1. 1

                                        I have a Fritzbox 5530 with two Mikrotik Switches connected via Single Mode BiDi to each other and cAP AC from Mikrotik as access points. I had a old telephone cable with a very some hole between the basement and first floor and with Single Mode LC/LC cable from fs.com it was not necessary to make the hole bigger to replace it.

                                        1. 3

                                          A Fritz!Box with fiber modem (currently 500MBit synchronous), switch and 802.ac WiFi access point (provided by our ISP, XS4ALL) near the fuse box. Ethernet to the living room and my desk. Most devices are hooked up through Ethernet. We also have a Fritz!Box 802.11ac repeater, hooked up to Ethernet as an additional AP in the living room. Fiber + ethernet has been super-reliable (only a brief downtime once since we moved in here 2.5 years ago). Never had any issues with WiFi through the Fritz devices either.

                                          I haven’t researched AVM (the makers of Fritz! devices) in detail. But they seem like a decent/reliable privately-owned German company that do not do ‘growth hacking’ and other shenanigans. Also, they seem to provide firmware updates for quite a while. I could be wrong, but my impression as a long-term customer has been pretty good.

                                          1. 2

                                            Fritz! here too - very good so far. The first Fritz!Box came with the ISP connection. I bought two more second hand to use as mesh repeaters.

                                            1. 2

                                              Hi Daniel,

                                              You’re right they are from Berlin, Germany and had already ISDN cards years ago. It’s quite common brand here in Germany and very good but expensive.

                                              What model is yours? I have a 5530 with SFP for GPON but my ISP has a ONT provided by them which I can’t replace. This ONT is connected via cooper to the Fritzbox.

                                              1. 1

                                                5490, the fiber cable is directly connected to the Fritz!Box.

                                            1. 3

                                              A Vortex Race 3 which I adore, as it’s the perfect mix of laptop layout and comfortable key travel.

                                              1. 2

                                                I can confrim. I really like the ability to remap keys directly without any applicatino or change in the operating system (e.g. remap Ctrl and Caps Lock).

                                              1. 1

                                                Puh, I try to keep it simple. Just Excel for my tasks and things I want to adress to other during regular meetings. A calendar named Reminder for time critical things. Flat text files in <10 folders with simple markdown as my note storage. Ah, and a regular notebook for in person meeting with others.

                                                1. 2

                                                  https://imgur.com/a/DvFUyMR

                                                  Mac Mini from 2009 with Dell 4K monitor and Vortex Race 3/BTC 5100 und Evolument Vertical Mouse.

                                                  See also https://www.pahem.de/usesthis.html

                                                  1. 1

                                                    awesome! didn’t knew the Mac mini from 2009 can support a 4k monitor :O

                                                    1. 2

                                                      It’s 1080p, but enough for me.

                                                    1. 1

                                                      I used also a DNS Blacklist but point the DNS to a local IP address and the host just reject the TCP connection with TCP-RST. I think this is faster than redirecting to a webserver which sends back e.g. a white image 1x1 pixel, because for the TCP-RST just two packets and for the webserver at least three (or more if TLS is in use). But is this true? Is the client faster because it cache the image and don’t even make a request at all? And if this approach with NXDOMAIN even faster, because it cache the response from DNS locally?

                                                      1. 3

                                                        Yes, using NXDOMAIN is a lot faster then 0.0.0.0 or 127.0.0.1 in the /etc/hosts file.

                                                        Some pages that I opened before (with 0.0.0.0 in /etc/hosts) loaded 10+ seconds. Now after change to NXDOMAIN with unbound(8) its 2-3 seconds and its loaded.

                                                        It really makes a difference for some pages.