New version of the forum software coming soon

overscan (PaulMM)

overscan (PaulMM)

Administrator
Staff member
Joined
27 December 2005
Messages
16,454
Reaction score
19,154
No release date yet but the developers have started releasing information on the next version of Xenforo. Some big rewrites happening under the hood, with a lot of performance improvements.

As usual, I will be testing it out before upgrade with the moderators and inviting feedback on changes.
 
Well, those goddam NASA pdfs are (sometimes) bigger than the 15240 KB limit you mentionned.
 
isayyo2 said:
Any chance of the [ALL] option returning to merge multiple pages together? Really miss that feature.
Click to expand...
It isn't supported on Xenforo, no. It can take a long time to load and hog server resources when viewing topics with large numbers of posts, so the developers don't think its a good idea.
 
overscan (PaulMM) said:
It isn't supported on Xenforo, no. It can take a long time to load and hog server resources when viewing topics with large numbers of posts, so the developers don't think its a good idea.
Click to expand...

Some Xenforo instances do support changing the number of posts per page at the user level. But it might also be a resource hog.
 
TomS said:
Some Xenforo instances do support changing the number of posts per page at the user level. But it might also be a resource hog.
Click to expand...
Yes, I can alter the number of posts per page for all users, and there's a plugin to allow users to choose the number of posts per page, but this still has a ceiling value of I think 100 posts per page, so it isn't a substitute to the "all" button.
 
In regards to the larger attachments, I can switch to hosting attachments in a different way which is a bit cheaper (possibly slower though).

Given how we've lost access to some great resources like Nasa CRGIS disappearing forever or the NTRS cleanup, its worth considering.
 
overscan (PaulMM) said:
In regards to the larger attachments, I can switch to hosting attachments in a different way which is a bit cheaper (possibly slower though).

Given how we've lost access to some great resources like Nasa CRGIS disappearing forever or the NTRS cleanup, its worth considering.
Click to expand...

If there are spiders, bots, etc downloading attachments having a way to limit or throttle those would reduce the hosting burden
 
A continuing aggravation for me is linkrot amongst the documents pointed to on the forum. Much can be attributed to the ephemeral nature of resources on the web, but other cases are due to internal reorganizations within the subject website. NTRS and DTIC are notorious for this reorganization.

Although we can not tackle this problem from this side of the interface - most users here post only the link (which becomes unusable) without a document title. With a doc title, it is trivial to use the NTRS/DTIC search function to relocate the citation and retrieve the document using the new handle.

So - how to enforce a rule that poster list the document title along with the DTIC/NTRS/AIAA/SAE direct link? I don't know. Any ideas?
 
Or maybe a mouse<=>moose converter utility for the doc index handle for DTIC?
 
Yes absolutely this needs to be a thing. I do fix up broken links in my favourite topics but it's really important to have the document title it makes things so much easier.

Can't think of a way to enforce it though. Attaching the linked pdf tp the post is a brute force solution.
 
I've long tried to salvage NASA NTRS broken links. By finding back the document on the Internet archive. Alas, from time to time I run into the limit I mentionned earlier. There are good reasons why in 2009 some posters put the NTRS link rather than attaching the document: because it was too massive, and still is.

The good news: the old, broken NTRS links: it is possible to cut-and-paste them into the Internet Archive: and instantly retrieve the Pdf. I do hope Frank Wolf has kittens over this, because it means that his actions against the spying chineses served no purpose in the end.

So, starting from the (prolific) Triton and @Graham1973

It would be feasible to pick all the ntrs links they posted and get the pdf back - via the Internet Archive, vintage 2010. Most of the time, it works...
 
aim9xray said:
A continuing aggravation for me is linkrot amongst the documents pointed to on the forum. Much can be attributed to the ephemeral nature of resources on the web, but other cases are due to internal reorganizations within the subject website. NTRS and DTIC are notorious for this reorganization.

Although we can not tackle this problem from this side of the interface - most users here post only the link (which becomes unusable) without a document title. With a doc title, it is trivial to use the NTRS/DTIC search function to relocate the citation and retrieve the document using the new handle.

So - how to enforce a rule that poster list the document title along with the DTIC/NTRS/AIAA/SAE direct link? I don't know. Any ideas?
Click to expand...

DTIC links contain the accession document number, a unique identifier that is more useful than a title.
 
How about intercepting the link invocation for NTRS and DTIC on the SPF site and (with a script) translating it on the fly with a notification to the user? Make the computer do the work. Or go further, take the new link and do a one-time automatic modification to the post updating it with the revised link.

Parenthetically, I'll note that most older links are going to die as HTTP is increasingly deprecated for HTTPS...

Here's how NTRS has changed: For the report, "Control power requirements of VTOL aircraft, phase 1 study Final report, Apr. 1969 - Sep. 1970".

old: http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/1971006223_1971006223.pdf
new: https://ntrs.nasa.gov/api/citations/19710006223/downloads/19710006223.pdf

Even I could write some ugly C* code to parse for "http://ntrs", if found then extract the handle, add the extra "0" and plug that into the new target string.

template: sprintf ("https//ntrs.nasa.gov/api/citations/%s/downloads/%s.pdf\n", new_string, mod_handle, mod_handle);

(I forget the rules for forward slashes - but you get the idea.)

*fuglier F77 code available on request. Maybe those brain cells are located in the Wayback Machine..
 
"DTIC links contain the accession document number, a unique identifier that is more useful than a title."

Agreed. So why does it not work reliably?
 
aim9xray said:
How about intercepting the link invocation for NTRS and DTIC on the SPF site and (with a script) translating it on the fly with a notification to the user? Make the computer do the work. Or go further, take the new link and do a one-time automatic modification to the post updating it with the revised link.

Parenthetically, I'll note that most older links are going to die as HTTP is increasingly deprecated for HTTPS...

Here's how NTRS has changed: For the report, "Control power requirements of VTOL aircraft, phase 1 study Final report, Apr. 1969 - Sep. 1970".

old: http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/1971006223_1971006223.pdf
new: https://ntrs.nasa.gov/api/citations/19710006223/downloads/19710006223.pdf

Even I could write some ugly C* code to parse for "http://ntrs", if found then extract the handle, add the extra "0" and plug that into the new target string.

template: sprintf ("https//ntrs.nasa.gov/api/citations/%s/downloads/%s.pdf\n", new_string, mod_handle, mod_handle);

(I forget the rules for forward slashes - but you get the idea.)

*fuglier F77 code available on request. Maybe those brain cells are located in the Wayback Machine..
Click to expand...

It could be done easily as a bookmarklet script. It would take me maybe an hour or two to write but I don’t have that much time to give it.
 
True. Agree regarding the "Wolfed" files in Wayback. But why do it by hand when it can be automated for those files still available to us directly?
 
It would be trivial to automate this if I was running the DTIC server using a rewrite rule in their web server. This is how I redirected most of the links from the earlier forum software. No government entity or corporate gives two shits about preserving URLs, or they'd do it already.

It isn't as easy to do it to the links embedded in posts - I'd basically have to do this at the database level, and be very careful about the pattern matching to make sure I don't inadvertently break valid URLs. Worth a try when I do the test forum migration I think.
 
aim9xray said:
True. Agree regarding the "Wolfed" files in Wayback. But why do it by hand when it can be automated for those files still available to us directly?
Click to expand...

Because I'm not good at computer programming. It's like Klingon or Sanskrit, to me. :D:D:D:D:D

But I'm confident you talented computer scientists know how to pull out that miracle. So - be my guest. Do your magic ! That would be amazing, really.
 
Last edited:
Hi Paul,

Can you confirm that .jpg and other non .webp images will still be supported in the future ?

I can't download .webp images from anywhere.

( and upgrading Windows / browers is NOT an option.)

Regards,
Gerard
 
overscan (PaulMM) said:
For any geeks, some videos on forthcoming changes.

Key new features include built-in Dark mode support, WEBP images, automatic client resizing of large images before upload so you don't need to do it, and a number of performance enhancements.
Click to expand...
Like Dark Mode support, do NOT like WEBP images (linux user here).
 
I'm probably not going to enable WEBP support any time soon, but after perusing the list of supported browsers:
  • Google Chrome (desktop) 17+ (
  • Google Chrome for Android version 25+
  • Microsoft Edge 18+
  • Firefox 65+
  • Opera 11.10+
  • Native web browser, Android 4.0+ (ICS)
  • Safari 14+ (iOS 14+, macOS Big Sur+)
I am a little at a loss to understand how anyone has a system that can't support WEBP format - e.g. Google Chrome 17 came out in December 2012 and runs on Windows XP.

In regards to dark mode - there is a manual dark option now - simply click the brush icon bottom left and switch between SPF Light and SPF Dark 2.0
 
overscan (PaulMM) said:
I'm probably not going to enable WEBP support any time soon, but after perusing the list of supported browsers:
  • Google Chrome (desktop) 17+ (
  • Google Chrome for Android version 25+
  • Microsoft Edge 18+
  • Firefox 65+
  • Opera 11.10+
  • Native web browser, Android 4.0+ (ICS)
  • Safari 14+ (iOS 14+, macOS Big Sur+)
I am a little at a loss to understand how anyone has a system that can't support WEBP format - e.g. Google Chrome 17 came out in December 2012 and runs on Windows XP.

In regards to dark mode - there is a manual dark option now - simply click the brush icon bottom left and switch between SPF Light and SPF Dark 2.0
Click to expand...
Rolling between images is much simpler when they're not WEBPs. I can open the folder and click on one of the non-WEBP images, then arrow right or left to see the next one. But WEBPs I have to open each one individually.
 
Hi Paul,

overscan (PaulMM) said:
They can be, but older devices that don't support webp format then lose access to images.
Click to expand...

I have to admit that I'm not even sure what the forum engine does when webp support is enabled.

I've seen on other forums that a proportion of the documents uploaded in recent threads is served to me for download as webp files with "alphabet soup" names, so my suspicion is that they were uploaded in other formats and most likely with meaningful names, then converted and renamed by the forum engine.

However, that's just my suspicion. Maybe the alphabet soup names have some other origin, I don't know.

The reason I'm unhappy with this development is that I like to save interesting files from the forum to my system, and it makes my archive much more accessible if the files keep their original, meaningful names. (Actualy, retaining the original name is quite helpful even if it's not meaningful, as it avoids storing duplicates.)

As I don't know the forum engine's features, if there's a way to enable webp support without introducing a renaming process, that would of course be my preferred option.

Might be a complete non-issue, but I thought I'd mention it since at least, it's a plausible explanation for what I'm seeing in other forums that already have webp support enabled.

Regards,

Henning (HoHun)
 
Recent Chrome browser 'upgrades' seem to have increased the frequency that I get 'unformatted' web-pages and must reload/renew.

I suppose is due to increasing complexity of web-pages, zillions of cascaded style-sheets etc etc...

My browsing is, of course, subject to predation by Duty Cat. Herself has just returned from ~0235 patrol, leapt onto desk and announced that a snack would be welcomed. And her sis would like a snack, too...
 
Still using Firefox on android, linux and windows. No problems with 'unformatted' web-pages.
 

Similar threads

Back
Top Bottom