Thanks a bunch! Currently I'm weighing SeaweedFS vs RustFS, and if I just wanted S3-compatible API, I'd probably go with RustFS although it's not so mature yet, but with WebDAV and some other stuff, and favoring stability and reliability, looks like I'm leaning more towards SeaweedFS.
TFA does not contain the word "copy", but copy-paste (shared clipboard) support stands out strongly to me as being "different" in Apple Silicon virtualization at this time.
TFA does mention the word "clipboard", but the containing sentence, while perhaps technically correct, seems a bit misleading, as follows: "As implemented in macOS (both as guest and host), there are also extensions to support keyboard and pointing devices, a shared clipboard, and high-performance graphics with Metal and GPU support." As I understand it, even if those extensions "exist", what good are they if they are not adopted?:
- If you virtualize macOS within macOS on Apple Silicon using UTM, you cannot copy paste between systems reliably (bidirectional shared clipboard is very, very fragile; "can" work a little but is essentially fully broken/unreliable).
- If you virtualize macOS within macOS on Apple Silicon using Parallels (often considered a best-choice solution), you cannot copy paste between systems at all (bidirectional shared clipboard is an explicit non-feature at this time).
Thus, if you want bidirectional clipboard, on a macOS host, you'll have to run a *nix (seems to work) or maybe Win (I haven't tried) guest OS.
I would have tried VMWare Fusion (free for personal use), but after jumping through all the signup and download navigation hoops, I couldn't even get the download link to un-gray itself out for me. Is bidirectional macOS to macOS clipboard implemented there? IDK and I cannot tell.
I use VMWare Fusion, (version 13.6.4 because 25H2 has a bad scrolling bug that makes it unusable for me) and it does implement bi-directional copy/paste, and you can also drag files in and out of the VM (there's a few tricks required to make it work sometimes) but I've only tested this for Windows 10 and Linux as a guest, not macOS as a guest.
Re: Downloading, I have the same issue. What I do is note the note the SHA2 for the file on the official download site, then find a copy somewhere else on the internet [0], and verify the SHA2 of that file matches the one on the Broadcom web site.
I used to run a macos vm under proxmox, and I just used screen sharing to remote in from my apple desktop. Copy/paste of even complicated stuff works fine between the apple desktop and the vm desktop. Also drag and drop files, etc
> SMB support in macOS remains slow and buggy to this day. I tried all combinations of server-side settings and obscure plist tweaks to make SMB navigation and search work as fast as they do on my Linux machine out of box before giving up. It is very obviously not a priority for their services revenue
That's where my thoughts went, too. I can make SMB "better" but not "great" usually, but it's annoying to have to look up and apply, and still have things not optimal. Just in case, IIRC I find this the most useful:
> Is there any way to download it? The reason someone might want to download it is for use as training data.
Another reason would be to able to keep running/using it even if the main site were to go down for whatever reason eventually; or, to operate a mirror of it, for redundancy (linking back to the original, of course).
Very, very cool. Hats off. I've considered attempting a more limited form of this for years.
For those who don't know, the 1911 Britannica is heralded for several reasons (and rightly criticized for regrettable others), but the most well-known is that it was the last encyclopedia before The Great War, and hence had a good amount of steam/optimism coming from the first and second industrial revolutions and the "Progressive Era", not sullied yet by thoughts of "the war to end all wars".
A question/idea for nice-to-haves, most respectfully. I don't know if it would be feasible. It's probably perfect as it is, simply linking to the image-page in unobtrusive text for each section. But I would love an option (emphasis on option) to see the text side by side with the page images. That parallel view would load all of the page images on the same page as the full article text. That way, I could "confirm" or "fact check" the faithfulness of the OCR, and also see the beautiful printing, at once, without opening each page separately and managing the images/windows myself. Most likely, I would use the site to jump to the articles, and read them mainly as images, only switching to the text form to verify what something said, or to copy-paste cleanly, etc. (As it is, initially, I thought I read the original images were available, but had to visit the page three (3!) times before finding where the side-links to them were.) Maybe thumbnails could be a middle-ground option (again, optional) for salience.
Thanks — really appreciate that, and glad it worked well for a random article.
That’s a great suggestion. A side-by-side text + page view would be very nice for exactly the reasons you mention (verifying the text and seeing the original layout). I haven’t built that yet, but I’ve considered it.
Also helpful to hear that the links to the scans weren’t immediately obvious — I should probably make them a bit clearer. This may also not be obvious, but you can click the vol:page links in the left margin and go directly to the scan of whatever page you're reading.
> But I would love an option (emphasis on option) to see the text side by side with the page images. ... That way, I could "confirm" or "fact check" the faithfulness of the OCR.
You can already do that on Wikisource. For example, here's p. 658 from the entry on "Molecule":
Also OP: I noticed some fidelity issues in your version (at https://britannica11.org/article/18-0684-s2/molecule). For example parts of the math formula under the line that ends with "the molecules of other kinds" ([1]) are missing (compare [2]). Also, in your version fn. 1 of this article is attached to "as they have always done" ([3]) but it should actually be attached to "Atom" on p. 654 ([4]):
That's cool about the WikiSource parallel text+image page view, TIL. Thanks!
As an example flow (since it took a minute to figure out): we can start at https://en.wikisource.org/wiki/1911_Encyclopædia_Britannica then click to navigate/browse volume > section > topic to get to a text page, then click Source tab, then click a Page Number (maybe hunt around for the correct page number), and see the parallel view, text + image. With previous and next page buttons available, retaining the parallel text + image view.
5. Notice the left margin contains hyperlinks like [105] whwere 105 is the page number nd links directl;y to the side-by side view of page 105. Click the [105] link on the left (for example), to get to https://en.wikisource.org/wiki/Page%3AEB1911_-_Volume_02.djv... which shows the text-and-image side by side (for that page).
This flow avoids the hunting-for-the-right page step, by using the direct links.
There's also a side-by-side option now. On any article, clicking the little scan button above the double-navigation arrows in the right margin will open the scan at whatever page you're viewing, and it will scroll as you scroll the text.
> As the conversation context grows, the page becomes increasingly laggy, quickly reaching a point of extreme sluggishness.
It was that way for me a year ago, so not that new of a phenomenon.
When a context/chat/session gets too long, I start a New chat and continue where I left off. Mostly solves this IME, though it's annoying to have to do.
You can do that (side by side), but I prefer the unified (rendered AND editable) WYSIWYG view enabled by those 2 settings in combo. Whichever line has your cursor will typically show the raw markdown as you make edits, but the document as a whole is rendered.
Seems there are other options, too, e.g. Garage [5].
I only used MinIO briefly in the past, but it was good/worked, and I do plan to set up S3 blob storage again soon.
[0] https://github.com/seaweedfs/seaweedfs
[1] https://github.com/versity/versitygw
[2] https://blog.elest.io/rustfs-vs-seaweedfs-vs-garage-which-mi...
[3] https://news.ycombinator.com/item?id=47806348
[4] https://github.com/RustFS/RustFS
[5] https://git.deuxfleurs.fr/Deuxfleurs/garage
reply