21 Saturday Oct 2023
Posted in Spotted in the news
Instant AI audio transcription / translation, in a small offline box which can can “translate 15 different languages” in real-time. For $300, fully private and offline, no corporate subscription required. Crowdfunding now, with the back end freely available under an open licence at github.com/usefulsensors/useful-transformers.
25 Thursday May 2023
Posted in Spotted in the news
A new free book from a UCLA historian, Athena Unbound: Why and How Scholarly Knowledge Should Be Free for All. Partly a history (no mention of JURN, though), and partly another stab at ‘how to make OA work’ in the future. There’s also a podcast interview with the author, albeit revealing some rather interesting assumptions. Such as…
“ChatGPT as I understand it at the moment scrapes and feeds off of the crappy end of the Web … I don’t think it’s able to get past the paywalls and into the scholarly databases and into the journals, as far as I know. So insofar as that’s true, then all we’re getting is a garbage-in, garbage-out product from ChatGPT … good ChatGPT should be based on the stuff that right now the paywalls keep us out of.”
The idea that worthy content is only to be found behind a paywall will raise an eyebrow among many OA publishers and indexers. He also makes the even more questionable assumption that piracy no longer exists in non-academic content (movies, games, TV, software, comics, instructional videos etc). But those assumptions aside, his core points are thought provoking…
i) It certainly would be interesting if an AI could be trained purely on a critical mass of non-science / non-medical academic journal texts. On say… Sci-Hub’s PDFs, Semantic Scholar’s PDFs (which I’m assuming subsumes the DOAJ’s relatively small PDF holdings), and perhaps even all the PDFs that could theoretically be harvested after spidering JURN’s index URLs. So far as I’m aware, in the admittedly blisteringly fast development of AIs, there’s nothing like that just yet. Neither of those three give complete coverage of course. But even in a partial early form such an AI would be interesting to have.
ii) He also raises the question of copyright in the output of such journal-ingesting AIs. If the pure unaltered text product of an AI cannot be copyrighted, he suggest that many will come to prefer the AI’s potted answers over struggling with the actual (paid) articles from which it was hashed. I’d add that what they won’t prefer to do, most likely, is then to laboriously hand-check the AI’s factual claims, logic, references, etc that may trip them up in a follow-on use of the text. Also the errors of taste and historical knowledge that will likely occur with scholarly arts/humanities AIs, such as we already see in dumb taste-matching software on store sites — for instance assuming that Ziggy-era Bowie is the same as Eno-era Bowie and Tin Machine-era Bowie, or that if you like The Hobbit you will also enjoy The Silmarillion.
That said, Elon Musk and others are already reported to be working on fact-checking and check-able ‘citation finding’ AIs. Daisy-chained workflows between very different AIs will likely emerge, and doubtless there will even be AIs which can suggest and optimise such daisy-chains. Part of such chains will likely be AI modules which try to strip out “AI-ness” and also steganographic watermarking and suchlike, and attempt to add “human-ness” to the look and feel of the sale-able end product. Perhaps even filters for glaring “errors of taste” in matters relating to art and literature.
24 Wednesday May 2023
Posted in Spotted in the news
A new version of PDF Index Generator, the first in a year. It’s the best standalone desktop software for making back-of-the-book indexes from finished PDFs. New in version 3.3 (May 2023), among other changes…
* Can now run-in the sub-headers (rather than doing list-style sub-headings). Video.
* Multi-page indexing (e.g. 265-278) can now be truncated (as 265-78). Video.
* Even/odd pages can now have their own margins set. Video.
* “Added an Include query … to index capitalized words”.
* Database files are now much smaller.
* “Fixed footnotes as it was showing footnote number & normal page number too!”
That last one is especially important for footnoting scholars. The footnotes feature was introduced in 2.9 (February 2020).
The software is still working all the way back to Windows XP, and is still the same price.
17 Wednesday May 2023
Posted in Spotted in the news
Currently “small-value goods can be imported free of extra charges” into the EU. Small has meant that parcels valued under 150 Euros (about $160), don’t currently attract customs fees. German newspapers and others are now reporting this is set to be abolished… if an amendment to planned EU customs reforms, tabled by a French MEP, passes into law.
Hopefully this is just a bit of French anti-Amazon gesture-politics, and the amendment will be withdrawn or struck out. Before it can cause damage to the cross-border mail-order trade in books, journals, comics and BDs, small artworks and Etsy-like crafts, and suchlike.
But other moves are equally ominous, and suggest a wider aim among the EU’s MEPs. One reads of plans that would see all mail-order sellers forced to “charge customs duties and VAT [national sales tax] at the time of purchase”, and they would also have to register with a giant new EU Customs Authority and log all transactions and buyers. This seems likely to place a huge and disproportionate burden on small publishers and catalogue-based dealers, such as those selling paper books into the EU. Small-scale creatives are facing enough challenges (digital tax reporting, increased postage, rampant piracy, generative AIs, customers no longer buying due the cost-of-living, etc). They don’t need to be whacked with this is well.
Developing…
14 Saturday Jan 2023
Posted in Spotted in the news
The Index of Medieval Art Database will become perpetually ‘free to use’ from 1st July 2023 onward, for “researchers at all levels”. The largest database of such research, it is well-established and includes a “photographic archive” which offers iconographic clustering and links to referenced texts (e.g. Arthurian + Sir Lancelot pictures can be clustered together, and apparently there are also links to the story-texts related to each image). It also seems to includes carving, engraved items (spoons etc) and so on, and the definition of “art” appears to be as wide as you might expect or require for answering research questions (e.g. “plain as a pikestaff” — how plain and unadorned were medieval English ‘palmer’ pilgrimage staffs, exactly, compared to the staffs of officials and merchants?).
The Index of Medieval Art is currently public and free for initial searches, though “Subscription is required to view images” even for the thumbnails in the search results.
18 Sunday Dec 2022
Posted in Spotted in the news
A new Flickr Foundation, set to start hiring in early 2023. Among the current aims: “restore and then grow the Flickr Commons”; provide evidence about use and usefulness; and bring in new curators. If only they hadn’t locked me out of my Flickr account years ago (due to the Yahoo crash-and-burn), I’d be among them.
18 Sunday Dec 2022
Posted in Spotted in the news
There’s a new Democracy’s Library at Archive.org, a unified hub bringing together… “more than 700 collections from over 50 government organizations, archived by the Internet Archive since 2006.” And they’re collecting more from governments around the world. The collection has a search box, constrained to the collection. As you might expect many documents are a little dated though many are still practical, such as How to Catch a Cat.
22 Tuesday Nov 2022
Posted in JURN tips and tricks, Spotted in the news
URLlister 0.4.0 (March 2021) is Windows freeware. Manually click through a long list of URLs, as if with a TV remote-control. Each new click passes the next URL on the list into your Web browser, and loads it in a new tab. No more manual copy-paste needed, for slow and careful ‘eyeball’ manual checking of a list of URLs. More details on GitHub.
It may not be ideal to have lots of tabs opening and accumulating over time, and you may prefer not to have to manually close them. In which case the Tab Wrangler extension for Chrome-based Web browsers can handle that. It “automatically closes idle tabs after a designated time”.
See also the Chrome browser extension Load URLs At Interval, which moves through the URL list at a set timed interval. The URLs are loaded into the same tab, rather than a new tab opening for each.
03 Sunday Oct 2021
Posted in JURN tips and tricks, Spotted in the news
The browser problem I described yesterday is fixed.
As a test sample, consistently utterly un-reachable sites in Opera were…
www.majorgeeks.com
www.etools.ch/
www.davidrevoy.com/blog
All loaded perfectly fine and instantly in the Pale Moon Firefox-based browser.
One clear possible cause I found was LetsEncrypt changing its root site certificates, which are used by way too many (20%?) of the world’s smaller website servers…
DST Root CA X3 will expire on September 30, 2021. That means those older devices that don’t trust ISRG Root X1 will start getting certificate warnings when visiting sites that use Let’s Encrypt certificates.
The timing was right. The systems affected were right. The reason for the Chrome vs. Firefox strangeness was right…
Browsers (Chrome, Safari, Edge, Opera) generally trust the same root certificates as the operating system they are running on. Firefox is the exception: it has its own root store.
Thanks to ‘GGG’, who got it right. He had exactly the same problem as me, Chrome (Brave) not working, Firefox working fine. He traced the broken sites to their use of LetsEncrypt root SSL certificates. This led me to the server techie Gunter Born in Germany warning of the same problems a little in advance and describing them in detail. Apparently the certificates are free and thus are widely used by smaller sites. It’s the world’s largest certificate authority. Seriously. The world’s largest certificate authority suddenly revokes its 300-million+ key server certificates and effectively breaks 20% of the Web and… the media don’t tell anyone in advance? So far as I can see only a few gadget sites and some Indian sites gave a few hours warning.
Anyway, assuming rogue SSL certificates rather than iffy DNS servers was the actual problem, as now seemed very likely… how to fix it?
The solution: You need to manually add fresher certificates. Do as the Tech Journal explains in the new page for the DST Root CA X3 Certificate Expiration Problems and Fix. There Stephen Wagner has kindly dug up the links to the new fresh certificates.
The guy who saved the world.
You will need a Firefox or Pale Moon browser to get them, as LetsEncrypt’s problem is blocking LetsEncrypt from itself (durh…). Some Windows users will need to choose the .DER rather than the .PEM version of the certificates. Best to get them all and see which version your Windows recognises and adds an icon to.
Once downloaded you need to double-click them and for each one a Windows Certificate import Wizard will launch. Install it to the correct folder….
Don’t just accept the Windows defaults (could install anywhere…), but guide each certificate to its correct folder. isrgrootx1.der and isrg-root-x2.der go in the “Trusted Root…” and lets-encrypt-r3.der goes in the “Intermediate…”. Intermediate seems just as important as the others, so don’t skip it. There appears to be no need to delete the old defunct certificates, although browser access seemed to speed up a bit when I hard-deleted the Sept 2021 certs from “Trusted Root…” and “Intermediate…”.
Now when you close and re-launch your Chrome-based browser, and after a pause of perhaps 12-20 seconds for each previously blocked site, the problem should be fixed. It was for me. I assume the one-time pause is for the browser to re-cache the page.
I did not need to re-boot Windows for this fix to ‘take’. The Windows-savvy will be able to type MMC at the Windows Start menu and then load a Snap-in to see new certificates and their dates…
This is also the way you delete the old ones, which cannot be done via Settings | Security | Proxy in Chrome/Opera…
Update: According the Linux Addicts the problem briefly took out Amazon Web Services, Shopify and The Guardian. The Daily Swig adds Google Cloud, Microsoft Azure, and many others.
02 Saturday Oct 2021
Posted in JURN tips and tricks, Spotted in the news
A curious problem has developed persistently in recent days, for users of Chrome and Edge browser… but not for Firefox / Pale Moon. Evidently the problem is now shared by others as well as myself.
While browsing a site/page fails to respond to the Chrome browser, but springs instantly into action for Firefox or Pale Moon (based on Firefox). In the Chrome-based Opera you get…
This site can’t be reached. [URL] took too long to respond.
Doesn’t appear to affect the mega-sites like YouTube or WordPress. Sometimes there is a 20-30 second delay in reaching a mid-ranking site, and often nothing at all from smaller sites or certain known recalcitrant mid-ranking sites (e.g. Stack Overflow, GreasyFork). Slack also seems to be badly affected, though that doesn’t affect me…
The problem appears to be cross OS, as I’m on Windows and this other guy (linked above) is on Linux. I have the same symptoms as he has: Chrome often gives this error while Pale Moon (Firefox) is totally fine. The problem occurs even if you are using a DNS server other than that of your ISP. For instance in Opera, it’s possible to select from a number of DNS servers. They all exhibit the same problem. Other fixes tried include:
* Changing the Windows IVP4 DNS to another (9.9.9.9, 8.8.8.8, 1.1.1.1) makes no difference either.
* Running with all browser extensions and scripts off also makes no difference.
* Visiting the page in ‘Incognito mode’ makes no difference.
* Modem reset makes no difference.
* PC reboot makes no difference.
* I don’t have proxies configured.
My first guess was some iffy under-the-hood Chrome update, perhaps some new and imperfect query being made to the some local and rather sluggish and partial DNS cache. Linux-guy’s claimed solution thinks along these lines and he suggests flushing your local DNS, which on Windows is:
1. Start menu.
2. Run.
3. Run dialog box, type…
4. ipconfig /flushdns
5. Confirm. A DOS-box window should flash up for a microsecond, the DNS cache is flushed, and the Run box exits.
Works as described above, but this didn’t cure the problem for me.
Nor did clearing the internal Chrome DNS cache (who knew?) and restarting the browser…
chrome://net-internals/#dns
Then I downgraded the Opera browser, back to Opera 78.0.4093.147 (mid August 2021) with the help of the full offline installer. Still the same problem, and thus it can’t be due to some recently-updated Chrome component.
So… if its not in Windows and not in Chrome, and not due to extensions or other obvious problems… what on earth could it be? It must be some kind of interaction between any DNS server and a Chrome-based browser, even a slightly older one. A problem which Pale Moon/Firefox is not affected by, and which has only recently started in the last few days. It can vary between DNS servers, some loading one page and not the other and visa versa.
One odd thing is that if you click hard and long and quick enough to load such a jammed page, like 50 times, it will often eventually load. This is repeatable. It’s like there’s a ‘black hole’ somewhere along the route, for smaller and mid-ranking sites that need DNS lookup, and eventually the system will ‘get the message’ and use an alternative route. I wonder in DNS servers have been ‘split’ in three and now have different sub-databases for top, middle and lower-ranking sites? And that the low-ranking databases sometimes power down their disks when not being called? That might explain it. The disks could need time to power up. But surely they would be modern always-on SSD’s and not old mechanical hard-drives?
Why Firefox / Pale Moon is unaffected I have no idea. But it is. I’ve been unable to discover if it uses any special DNS routing. Only that Pale Moon has no support for ‘DNS over HTTPS’.
So the temporary solution is then:
1. Open the Pale Moon browser, which has no such problems, and keep it open.
2. Install Andy Portmen’s “Open in Pale Moon” extension in Opera or Chrome.
3. Pin “Open in Pale Moon” button to your bookmarks bar.
4. Launch any recalcitrant page in Pale Moon (Firefox). This browser is already open so it will load instantly, and the supposedly ‘un-findable’ page will also load instantly.
Sadly the above only works once Opera has actually received the “This site can’t be reached. [URL] took too long to respond.” message. If you pass the URL over to Pale Moon while the browser is still waiting (and waiting and waiting…) for a DNS server, you get nothing in Pale Moon. You can however go back and right-click on the original hyperlink and “Open in Pale Moon” that way.
You can also switch your RSS reader to open pages externally in Pale Moon / Firefox.
Update: This, at first glance, seems to explain the difference between the browsers…
1. “Chrome uses DNS prefetching to speed up website lookups”
2. DNS pre-fetch is off by default in Pale Moon… “DNS prefetching disabled by default to prevent router hangups”.
Checking the value on about:config / network.dns.disablePrefetch assures that it is indeed off in Pale Moon.
In Chrome/Opera this is now called “Preload pages for faster browsing and searching”, and again it is turned off for Opera. The uBlock Origin addon forces it off.
So, despite sounding plausible, the above can’t be the explanation for the problem.
Update: Browser problem fixed: it was LetsEncrypt’s expired root SSL certificates. Install the new ones. Firefox / Pale Moon uses its own SSL certificate store, which was why it was unaffected.