Looks like a good tip. Found this:

https://web.archive.org/web/20251112132449/https://wiki.debian.org/InstallingDebianOn/Marvell

The specs say it’s Marvell 6281 1.2GHz… not sure what that is but the search leads to some deeper specs for a RISC CPU that is both 32bit and 16bit. The info is dicey because “6281” does not appear in those docs. In any case, I suspect you’re right, that it’s not 64 bit.

Why are you buying copies of Windows?

I’m not. Hence the problem.

If you’re dealing with old windows boxes anyway, the OEM license is still on it. Further, you can just install the Trial versions of windows for the short time you need to do your hardware eval. No purchase needed.

I would be most tempted to do a multi-boot.

I did not buy any of the hardware. I rescue PCs from curbside dumps and have not bought a PC in over 20 years. The machines all have Windows stickers (mostly XP, win7, and vista). The s/n on all the stickers is rubbed off. All of them. Those stickers really do not hold up over time or abuse. Most of the dumped hardware has a drive w/Windows already installed, but that’s typically in a rough state… someone else’s data, someone else’s language (not English), and sometimes passworded. So if I am going to install Windows on it, it won’t be legit.

This is on hardware that came out of the trash, on your isolated DMZ network. What possible information could it spew that could hurt you?

Let’s not confuse the PC with the peripherals. It all came from the trash, the PCs and also whatever it is that I am testing. Some of the salvaged PCs have popups saying that the Windows copy installed is illegit (just as it would be if I install a new copy of Windows). I do not trust whatever Windows would send over the wire generally, legit copy or not.

I don’t know much about the trial versions you mention, but I think that’s a new option since win10, no? Can I get a trial version of XP or win7, for example? I kinda doubt it. I suspect there are likely only blackmarket copies of those versions at this point.

I understand not wanting old/unpatched Windows OS sharing the network with your other systems, but why not set up a NAT with a DMZ that would allow the Windows machines you’re testing to reach out to the internet to download whatever drivers it needs without those Windows machines being able to access your primary network?

I am not enthusiastic about buying 3+ different retail versions of Windows in order to put them safely online without worrying about whatever anti-piracy signals they send, along with whatever else those dodgy black boxes spew. It’s not inbound attacks that would concern me b/c they are mostly powered off test boxes anyway. I could setup egress firewalls that block everything outbound (as I do for printers), but then I would need to mitm my own machine and detect where the drivers and legit tools try to connect so I could put holes in the f/w.

Thanks for the tip. I installed it and first thing I tried was to save the weasyprint documentation on their website to a PDF (and I was surprised they did not create their own docs as a PDF). The result looks bad with text overlapping text. When their own website docs do not save well into a PDF, that seems like quite a signal for where they are.

Anyway, I’m glad to have a 2nd option which could be viable in some cases… something to cling to when wkhtmltopdf ultimately dies.

Oh, right! Indeed, I forgot there was a language specifier for the post itself. That fixed it. Thanks!

What a piece of shit software.

thanks for the tip, but it seems to make no difference. I changed it from browser default to english. Is that the right setting? The screenshot does not render for me.

I simply tried to add the word “needed” to the end of the title, and it said “language not allowed”.

Please report any spam you find to the mods by hitting the “create report” flag under the ellipse (“…”), and explain why it is spam (please read the rules and cite the broken rule so the mod does not have to guess). They will remove it if it is non-conformant or irrelevent to the topic.

Indeed minipage would be a decent workaround in this scenario. It’s effectively very similar to what I settled on (paracol). I just updated the OP to reflect this.

With both the minipage approach and the paracol approach, we give up the ability to have text flow from one column to the other. But for the doc I am working on at the moment, it’s perhaps best to control the flow anyway.

That would work if dates are not reused. But if you have a block of \DTMsavedate variables in the preamble and then refer to those dates throughout the doc by the variable name you assign, the spreadsheet would be more trouble than it’s worth because you would have to copy-paste all the dates into the spreadsheet, choose the new format, copy them back, and risk the update anomaly in the event that you revise a date in the preamble. Could be useful for some situations though. But I guess I would still rather replicate \DTMsetdatestyle{default}\DTMsetup{datesep=/} in every cell that needs it.

Ah, right, I think I meant elinks. It has been a while, but I recall that JS support was essential. I first heard of it working for someone else which inspired me to try it. And it worked for me. I don’t recall the extent of my tests (it was a while ago). If reg and login are a problem, you could perhaps do those things with cURL or something and hand off the logged in cookie to elinks.

All hacks aside, emacs is likely the best option since the pkg was designed specifically for using Lemmy. But that’s speculative since i have not tried it.

I have no idea about X11, which I assume is what you meant by X. GIMP crashes Wayland. Or Wayland crashes GIMP and takes itself down with it. GIMP can launch on wayland and carry out /some/ tasks, but some functions and windows in GIMP are sure to trigger a crash. Usually GIMP alone goes down, but sometimes the whole wayland session goes down with it.

(edit) I should also mention that GIMP probably has no Wayland support. It’s most likely using Xwayland in my case. Xwayland is an attempt to give Wayland the x11 hooks that older software rely on (as x11 is a dying technology IIRC). The Libreoffice suite of tools have the same problem and crash on Wayland, but there is an obscure hack that fixes Libreoffice.

I double-checked, and yes: A table can span an infinite amount of pages.

That’s useful. It also just occurred to me that I probably want 1 table per paragraph anyway to keep the alignment. If my collaborator turns out not to be LaTeX-literate, I will try libreoffice w/tables.

May I ask you: You don’t seem to use a GUI at all, and it seems like you have never seen a GUI. I only know one other person who lives like that, and they are 100% blind. Is that the case with you? If you are using Lemmy without a GUI, how?! I’d like them to be able to browse Lemmy!

I am not blind. I just have a strong bias for TUIs, for scripting, control, and performance on old hardware (as I don’t do spy chips, which are post-2008 intel CPUs or post-2013 AMD CPUs). I created the !text_ui@lemmy.sdf.org community because of this preference.

I still today use a GUI browser for Lemmy. But I have investigated and there are options:

• NeonModemOverdrive is a TUI app for Lemmy but at least for me it was too buggy to be useable.
• Someone wrote an emacs app. I would like to install that but ATM it demands a newer version of emacs than I have.
• The Links text browser (not to be confused with lynx) has some JavaScript support and it works with the Lemmy stock client. I don’t recall off the top of my head why I did not make regular use of that. It’s obviously a hack to get a TUI but the Lemmy UI was not designed for it so it’s not as fast in the ergonomical sense as we expect a well designed TUI to be. Perhaps it would be practical for your friend. OTOH, the emacs option is probably better.

I don’t go out of my way to send email to Gmail accounts.

Does that mean you allow Google to block you? Or that Google is permissive for you and does not require a CAPTCHA?

I personally will not go out of my way in the slightest. Running a mail server from a residential IP with no spf/dmarc/dkim ensures that MS always refuses my email and Google refuses it /most/ of the time. And I am fine with that. To be clear, I am not asking to change that. My question is political. I want to know what MS is putting small email operators through, when they are willing to send to MS recipients.

I would like to hear from a server admin who has had to solve a CAPTCHA in order to get mail from their server accepted.

Right, but the effort is in the manual entry. I am starting with a document coded in LaTeX which produces a 2-column PDF. So using latex2rtf will be my attempt at an automatic conversion but I imagine it will be a disaster. So from there, alternatively, I have a lot of copying and pasting to do. A table does not feel right, but I guess the first thing I have to look into is whether a table can span many pages. If each page must have a separate table, then what happens as the table grows? I anticipate a lot of pain, but nonetheless I’ll see what I can do.

Journals and newspapers commonly do multiple columns though not for different languages, so the right column is a continuation of the left. I suppose a table might be a hack around that, assuming L/O even supports 2 columns of text in the traditional sense in the first place.

I do the same and I have traced unreported leaks to the source that way, but it is useless when the breach notice comes not by email but to the physical address.

The thread is about legal obligations and recourse. Consumers should have to do anything sneaky to be protected.

I might try to use the latex2rtf tool then convert the rtf to libreoffice. But I am expecting disaster with 2 columns. A table sounds like high effort but it may be the only option in this situation.

Perhaps that explains the unpopularity of gammu and gnokii. I boycott Microsoft, Google, Facebook, etc, so whatever most people use is outside of my realm. I use XMPP with the very few others who use it (and those are ppl whose arm I managed to twist). For everyone else, SMS is the only option. No one really boycotts the tech giants, so I’m out of touch with the norms, which probably boils down to whatsapp.

Literally just print as PDF on Windows, you can pick if you want greyscale or coloured…

That’s a non-starter for the same reason I mention: color backgrounds are dithered and colored text becomes various shades of gray, not black. There are lots of dithering techniques to experiment with and some might yield black text, but if I am producing a PDF I would not want to impose that kind of expertise and experimentation on the end user.

I am asking how to produce a PDF that has a mono mode and a color mode – and whether that technology even exists. If the PDF is rendered on the screen, it might have color backgrounds. But when printed to a mono device, the color backgrounds should be stripped out (or not, depending on my specification).

If you want to produce a nice PDF from a markdown document, I recommend the knitr package in RStudio with R, and writing an .Rmd file such that you can just place all your graphs/code/text in it as you write, including LaTeX stuff.

I am not starting from a markdown doc, but I would if that made a difference. I’ve never used RStudio. Does that produce a PDF that has two representations, mono and color?

Ah, thanks… but you would see it because (like me) you are on sdf.org. I believe we are in a cached version of !budgetaudiophile@lemmyonline.com and only sdf.org people will see this.