- main - earth.li

OT: Debugging a weird web quirk
by Mark Rogers 03 Dec '21

03 Dec '21

I host a wordpress multi-site install for personal use. My Dad has a site on it, and he's come to me with a weird issue that I can't even figure out how to start debugging! If you visit the site in Chrome: https://www.lyveden.org.uk/ .. and visit a page with images, eg "Garden Revealed", you can click on the images to see them full size. However, when you click "back", the browser reverts briefly to the page before showing the image again, and if you click back a second time you are returned o the previous page (the one prior to the one you opened the image from). Dad has reproduced this on three PCs (two Ubuntu, one Windows). I couldn't initially reproduce it until I tried in a private tab, which might be a clue. I tried using Chrome's developer tools to see what was going on but nothing stood out and I was quickly out of my depth! I know it's not Linux related (highly doubt it's a server issue!) so it's OT, but does anyone have any clues? -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0344 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

3 4

Capturing a web request (server-side)
by Mark Rogers 01 Dec '21

01 Dec '21

Background: I have some hardware which can make a REST request when an input changes. I'm trying to send that request to a ThingsBoard server but it's generating a mostly useless error (it indicates that the JSON won't decode but won't tell me what's being received). If I direct it to a PHP script on a different server which just logs the POST data, everything looks OK and if I copy+past that same data into a commandline curl request to the TB server it works fine. So whatever is failing is more subtle. So what I want to do is set up a temporary server-side script that will log everything that is received in an HTTP request (URL, POST/GET data, headers, etc). I can then point the hardware at it, and then also the working curl commandline, and see what is different. Any suggestions as to how to do this? Or how better to describe what I' trying to do so I can Google it more effectively? -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0344 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

2 2

Mail search that can find an exact date
by Chris Green 29 Nov '21

29 Nov '21

I'm looking for an E-Mail search command that will find an *exact* date. I.e. I know the E-Mail I'm looking for has:- Date: 19 Jan 2010 15:30:24 However none of the mail search programs I've tried seems to be able to handle this, they all want to search for mails in a date range, usually with no time at all specified. I've only tried mairix and grepmail so far but it looks as if notmuch will have the same issues. I know I could simply use grep but that doesn't "know about" E-Mails so although it can find the specific E-Mail it can't show me the E-Mail body which is what I'm after. -- Chris Green

2 3

Best for Twitter?
by Phil Thane 26 Nov '21

26 Nov '21

Hi The only social media I use is Twitter. In the past I've used Choqok as a desktop client but it doesn't play video links you have to click a link to open in a browser. In itself that's not really an issue but it's not always obvious there is a video it could just be a photo and you can see those in Choqok so there's no need to click a link. The result is you either waste time clicking liks tha reveal nothing new, or miss a video. Cawbird does, or rather did, play videos in a popout window but following a distro change/upgrade now doesn't as a rule. I'll post examples of links that do play and don't play in Cawbird and an error message for the latter. It seems to be a GST api problem but I'm not smart enough to fix it, can anyone help? This is an actual post on the Twitter website with a video that _does_ play in Firefox https://twitter.com/RosieisaHolt/status/1464170045861515277 Error message if I try to play in Cawbird: gstvideodecoder.c(1138): gst_video_decoder_sink_event_default (): /GstPlayBin:video/GstURIDecodeBin:uridecodebin2/GstDecodeBin:decodebin2/avdec_mpeg4:avdec_mpeg4-3: no valid frames found Address if I chose 'open in browser' from Cawbird - which also doesn't work https://video.twimg.com/ext_tw_video/1464169933110235137/pu/pl/rFhk7k0W508d… This one DOES play in Cawbird and this is the address if I chose 'open in browser' and that works too https://video.twimg.com/tweet_video/FFHoG74X0AItrRe.mp4 Those last two addresses are very similar in format except one is actually a link to an mp4 and the other to a container. Is that relevant? -- Phil Thane www.pthane.co.uk Tweet @pthane 01767 449759 07582 750607

2 1

Accessing a web page via another server
by Mark Rogers 17 Nov '21

17 Nov '21

My Pi is on my local network (headless and somewhat physically inaccessible) at 192.168.1.200 I have a 3G USB stick connected to my Pi, which presents itself as an ethernet device (eth1); the Pi gets an IP address of 192.168.8.100 and the USB stick has a configuration website on 192.168.8.1 How can I access this website from my laptop (192.168.1.100) via SSH? I tried SSH tunnelling but it's not working, at least in part because the web page links to resources at http://192.168.8.1 which aren't accessible. I have tried w3m on the Pi itself, but the web page has loads of javascript so doesn't work. Presumably I can set up a proxy somehow? Other suggestions? -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0344 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

4 16

Replacement for Pinta?
by Steve Mynott 10 Nov '21

10 Nov '21

Has anyone any suggestions for a very simple image editor? I just want something to add text to screenshots which supports multiple layers. I've used Pinta in the past but since they changed the icons in the tools menu & used dark mode it takes me too long to edit since I have to search for "what does the text tool look like"? And the snap version doesn't seem stable. -- Steve Mynott <steve.mynott(a)gmail.com> rsa3072/629FBB91565E591955B5876A79CEFAA4450EBD50

2 1

Does anyone here use sshuttle?
by Chris Green 09 Nov '21

09 Nov '21

I've been trying to use sshuttle but I simply can't get anywhere. I have it installed and can run it but it just outputs loads of waffle and leaves me with no command prompt so I can't see how I can possibly do anything. I'm probably doing something stupid or not doing something vital but I can't find any tutorials that actually say 'type this' and then you will have a working sshuttle VPN. So, does anyone here use it, or has anyone here played with it? -- Chris Green

2 2

OT: Mail forwarding and junk folders
by Mark Rogers 21 Oct '21

21 Oct '21

Several family members have domains (mostly at 123-Reg) and use email forwarding for xxx@<their domain> to their personal addresses (eg at Gmail, Virgin, BT, whatever). Increasingly they're finding that mail sent *from* their domain addresses is ending in recipient junk folders. How they send varies: my dad used Thunderbird on Ubuntu (so maybe not entirely OT!), my wife sends via Gmail, others I'm not sure but probably thinks like Outlook/Outlook Express (or whatever it is these days). They'll be using the SMTP settings of their ISP or personal mail provider (eg my wife will be sending via Gmail using her own domain set up as an alias). Since I have several people with issues - and especially since one of them is my wife! - I want to take a fresh look at all of this and get things right. I'm guessing that this is going to involve SPF, DKIM, DMARC stuff that I don't really understand but have succeeded in setting up before. It feels like something that must be a sufficiently common issue that there would be plenty of help out there but it's one of those issues where I can't see the wood for the trees in any search I try, so I'm hoping someone here might have some pointers? The more that can be done server-side the better as that reduces the need for me to help lots of people configure lots of different clients. If I need to move domains from 123-Reg that's not out of the question - indeed I have plans to do that to somewhere I can control DNS, mail forwarding, etc via APIs anyway. -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0344 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

4 8

Backups and encrypted disks
by Peter 13 Oct '21

13 Oct '21

I'm thinking of moving to an encrypted home partition when setting up a new computer. Its a completely new machine. I'm a bit puzzled how to proceed. I've looked at tutorials on how to use fscrypt and encryptfs but they are quite complicated and probably beyond my ability to carry out error free. So finally it seems the simplest approved way is to create an encrypted home partition on the new machine, then take a backup of home and use it to restore to that partition. This would have the great merit of being very safe, the only thing one could mess up would be the new partition, or worst case the backup. I can do a partition backup easily enough to an external hard drive using clonezilla. So I boot the new machine, connect the external with honme on it via usb... and then what? Do I just copy all the files? Will that preserve the desktop and the links on it? Or is there some way of restoring the partition itself to the encrypted partition? Should have started out with full device encryption of course, but that was more years ago than I like to think! Peter

2 1

Any iptables experts here, I need something explained
by Chris Green 11 Oct '21

11 Oct '21

I have a script I wrote a while ago to enable my laptop to become a WiFi hotspot which connects via a proxy to give me the ability to be 'in the UK' when I'm overseas. The script uses redsocks and some iptables stuff so that HTTP and HTTPS connections to the laptop get forwarded across to the proxy server. So, ssh provides the tunnel to a server in the UK:- ssh -fTnN -D 1080 chris(a)cheddar.halon.org.uk ... and redsocks connects all socks5 requests on port 12345 through to port 1080 to connect to the remote server (portion of redsocks configuration) :- redsocks { /* `local_ip' defaults to 127.0.0.1 for security reasons, * use 0.0.0.0 if you want to listen on every interface. * `local_*' are used as port to redirect to. */ local_ip = 0.0.0.0; local_port = 12345; // `ip' and `port' are IP and tcp-port of proxy-server // You can also use hostname instead of IP, only one (random) // address of multihomed host will be used. ip = 127.0.0.1; port = 1080; // known types: socks4, socks5, http-connect, http-relay type = socks5; } What I can't understand is the iptables bits at the end of the script (I presumably copied it from somewhere!):- sudo iptables -F sudo iptables -t nat -F sudo iptables -t nat -A PREROUTING -s 10.42.0.0/24 -p tcp -j REDIRECT --to-ports 12345 The first line I understand (I think), it clears out any existing iptables settings. Then the second line appears to do the same thing, does the second line actually do anything rather than a clear out of part of what the first line has already done? The third line is what confuses me, it seems to me as if, firstly it's the wrong way round and secondly there should be more needed to make it all work as intended. The intent of the whole thing is that WiFi clients (e.g. my kindle lookalike, or my phone) can connect to the laptop (which is acting as WiFi server) and have their HTTP/HTTPS requests sent through to the proxy. It reads to me as if requests *from* the 10.42.0.0/24 network get redirected but that's all wrong. The LAN with the laptop on it is 192.168.0.0/16 (or is it 192.168.0.0/8) and that line is only supposed to do things to 'a packet that creates a new connection'. However it did all work OK when I first tried it and probably still does if I get it configured right. What I'm confused about is that 10.42.0.0/24 IP, where does that come from? Is it the local LAN's or something else? -- Chris Green

1 0