![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
(previous post)
God knows how I managed to get zimmer working on my previous computer, but I can't seem to replicate that success here. Fuck it: I'll just use wget. The resulting files are more broadly compatible anyway, 7-zip is very good at compressing collections of HTML files down, and I haven't been personally compiling any wikis for which the ability to use the results without first decompressing them is important. The reject flags will need some tuning, but I can handle that.
(Also I'm going to switch these from monthly to quarterly: I don't think I really need them to follow the live versions *that* closely, and I suspect wget is going to be even more time-consuming than the ~20 - 30 minutes/wiki that zimmer was. Yeah, it's just compute/bandwidth time and not active-monitoring time, but as I write this part of the draft I'm waiting for a one-off (or maybe annual) game-guide wiki scrape to complete so I can boot over to Windows and finish installing the Windows stuff.)
---
The tumblr-utils cron job quietly failed, apparently because Ubuntu 20.04 is moving away from Python 2. Installed the python2 package and replaced all instances of "python" in the command with "python2": manual invocation seems to be working now, and we'll see how the cron job goes.
[a couple days later]
Yep, that seems to be ticking along just fine.
---
(I just *heard* WinFF complete the latest batch of lifelogger files by the way the fan slowed way down. An advantage to only having four threads (and thus having WinFF take up a larger proportion of [the CPU usage the laptop is designed to withstand]), perhaps. Even when it's running, though, this fan *is* significantly quieter.)
---
My list of which programs to install on a new computer is much more thorough now: it was a good start, but I found some things I'd overlooked. It's been helpful to be able to look at my previous computer and see how things are set up there, but one doesn't always get that opportunity when migrating and I'd like that to affect things as little as possible.
(For that matter, I plan for the list to eventually contain not just notes of which programs to seek out but, where feasible, *the programs themselves*. Clouds are useful, but I don't like to be dependent on them.)
---
The food-diary file takes much less time to save on this computer. The file is stored on the HDD, too, so it's not an SSD-vs-HDD thing (at least not when it comes to the file itself; I think the LibreOffice *software* is on the SSD).
---
Migrating Thunderbird to a new computer by sticking the old ".thunderbird" folder into the new home directory continues to Just Work. It doesn't even need to be a fully up-to-date copy as long as you're using a synchronise-with-server setup (rather than a download-and-delete setup†): it will catch up automatically.
---
Not migration-related, but: apparently GNOME SoundConverter *does* have a 64kbps setting. I think I was thinking of 32kbps, which I was using back at the start of my lifelogging.
WinFF can't handle my podcast files for some reason: rather than the usual progress display, it talks about "frames" as if I'd handed it a video or something, and it never seems to make any progress no matter how many frames it goes through. I think I'll just stick with my current methods.
---
The command-line replaygain software I was using doesn't seem to be in Synaptic anymore. Apparently music player Quod Libet has a replaygain function, as well as a metadata editor and podcatcher. I didn't like its podcatcher setup (too much clutter of old episodes, and I don't think there was an option for "check for new episodes only when I say so"), but the rest of it might replace the stuff I was previously using for those things. I'd been having trouble with my previous replaygain not seeming to react to the reference-loudness flag, so perhaps this is for the best.
---
Turns out the biggest storage-drives-with-USB-interfaces-installed we have lying around are 80 GB (also 17 years old, and therefore probably not very trustworthy even *if* 80 GB sufficed). I thought we had a 500 or two, but maybe not.
Dad mentioned something about "Mom's old computer", and I realised that was a good point. Once the hermit-crab shuffle completes I'll try using the TV's prosthetic brain as a backup drive, at least in the interim until I can get something more dedicated.
(I'll have to encrypt them, of course: there's no question of whether to rely on the "token authentication" of having access to the drive if I'm putting them on an explicitly communal computer. P7Zip has an encryption option: I expect that'll do.)
(Not airgapped, admittedly, but it's something. Perhaps I'll keep them on the secondary partition, add an extra layer of remove.)
---
I haven't used Scuttlebutt much, but I think I *will* go ahead and save a copy of my client data. Best to reserve the option of picking up where I left off, if I ever end up returning to it.
---
†The lies-to-children way of putting this is "IMAP vs POP", but apparently both protocols allow for both setups (though with different defaults). ↩
God knows how I managed to get zimmer working on my previous computer, but I can't seem to replicate that success here. Fuck it: I'll just use wget. The resulting files are more broadly compatible anyway, 7-zip is very good at compressing collections of HTML files down, and I haven't been personally compiling any wikis for which the ability to use the results without first decompressing them is important. The reject flags will need some tuning, but I can handle that.
(Also I'm going to switch these from monthly to quarterly: I don't think I really need them to follow the live versions *that* closely, and I suspect wget is going to be even more time-consuming than the ~20 - 30 minutes/wiki that zimmer was. Yeah, it's just compute/bandwidth time and not active-monitoring time, but as I write this part of the draft I'm waiting for a one-off (or maybe annual) game-guide wiki scrape to complete so I can boot over to Windows and finish installing the Windows stuff.)
---
The tumblr-utils cron job quietly failed, apparently because Ubuntu 20.04 is moving away from Python 2. Installed the python2 package and replaced all instances of "python" in the command with "python2": manual invocation seems to be working now, and we'll see how the cron job goes.
[a couple days later]
Yep, that seems to be ticking along just fine.
---
(I just *heard* WinFF complete the latest batch of lifelogger files by the way the fan slowed way down. An advantage to only having four threads (and thus having WinFF take up a larger proportion of [the CPU usage the laptop is designed to withstand]), perhaps. Even when it's running, though, this fan *is* significantly quieter.)
---
My list of which programs to install on a new computer is much more thorough now: it was a good start, but I found some things I'd overlooked. It's been helpful to be able to look at my previous computer and see how things are set up there, but one doesn't always get that opportunity when migrating and I'd like that to affect things as little as possible.
(For that matter, I plan for the list to eventually contain not just notes of which programs to seek out but, where feasible, *the programs themselves*. Clouds are useful, but I don't like to be dependent on them.)
---
The food-diary file takes much less time to save on this computer. The file is stored on the HDD, too, so it's not an SSD-vs-HDD thing (at least not when it comes to the file itself; I think the LibreOffice *software* is on the SSD).
---
Migrating Thunderbird to a new computer by sticking the old ".thunderbird" folder into the new home directory continues to Just Work. It doesn't even need to be a fully up-to-date copy as long as you're using a synchronise-with-server setup (rather than a download-and-delete setup†): it will catch up automatically.
---
Not migration-related, but: apparently GNOME SoundConverter *does* have a 64kbps setting. I think I was thinking of 32kbps, which I was using back at the start of my lifelogging.
WinFF can't handle my podcast files for some reason: rather than the usual progress display, it talks about "frames" as if I'd handed it a video or something, and it never seems to make any progress no matter how many frames it goes through. I think I'll just stick with my current methods.
---
The command-line replaygain software I was using doesn't seem to be in Synaptic anymore. Apparently music player Quod Libet has a replaygain function, as well as a metadata editor and podcatcher. I didn't like its podcatcher setup (too much clutter of old episodes, and I don't think there was an option for "check for new episodes only when I say so"), but the rest of it might replace the stuff I was previously using for those things. I'd been having trouble with my previous replaygain not seeming to react to the reference-loudness flag, so perhaps this is for the best.
---
Turns out the biggest storage-drives-with-USB-interfaces-installed we have lying around are 80 GB (also 17 years old, and therefore probably not very trustworthy even *if* 80 GB sufficed). I thought we had a 500 or two, but maybe not.
Dad mentioned something about "Mom's old computer", and I realised that was a good point. Once the hermit-crab shuffle completes I'll try using the TV's prosthetic brain as a backup drive, at least in the interim until I can get something more dedicated.
(I'll have to encrypt them, of course: there's no question of whether to rely on the "token authentication" of having access to the drive if I'm putting them on an explicitly communal computer. P7Zip has an encryption option: I expect that'll do.)
(Not airgapped, admittedly, but it's something. Perhaps I'll keep them on the secondary partition, add an extra layer of remove.)
---
I haven't used Scuttlebutt much, but I think I *will* go ahead and save a copy of my client data. Best to reserve the option of picking up where I left off, if I ever end up returning to it.
---
†The lies-to-children way of putting this is "IMAP vs POP", but apparently both protocols allow for both setups (though with different defaults). ↩
