OEIS timeouts

[Sean A. Irvine]

Hi,

We are currently experiencing timeouts and other problems with both the OEIS and OeisWiki.

Russ is working on it.

You might want to postpone any substantial changes to submissions to avoid losing your work.

Sean.

[Russ Cox]

Hello all,

The OEIS server is back up and should be better than ever. 

I took the opportunity to future proof it against the particular problem it had

and to update the Mediawiki software to a more modern version.

One side effect of the latter may be a request to change your password 

if it is too short the next time you log in to the wiki.

Please mail ad...@oeis.org if you see anything that looks like server trouble. Thanks!

Best,

Russ

[Rémi Eismann]

Hi Russ,

LaTeX seems broken on the OEISWiki.

Best,

Rémi.

[Ray Chandler]

Server looks OK.  OEISdata did not happen this morning.  Not sure if it is related to your changes.  Thanks.

Sent from AT&T Yahoo Mail on Android

--
You received this message because you are subscribed to the Google Groups "SeqFan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to seqfan+un...@googlegroups.com.
To view this discussion visit

https://groups.google.com/d/msgid/seqfan/CADSkJJWPkBDe7aEgbe7Vua0%2Bc88WcBbonrHvBjDJYwOK7Kquuw%40mail.gmail.com

.

[Natalia L. Skirrow]

LaTeX now functions, but not as it originally did; the spacing and scaling are both way off, and integrals inside exponents now render as though they were done with \int\limits_{...}^{...} all the time.

I'm not sure which renderer the wiki has switched to, but MathJax and KaTeX would be very preferable (and conserve the property of rendering directly into HTML with selectable text instead of through SVG).

I have been working on a lot of pages in my userspace, and this has made a lot of them look less pleasant. For instance, the start of my page Stirling's approximations, the first equation

is much less intelligible than its original appearance (to which this is only an approximation, in MathJax)

the e is no longer meaningfully separated from the limits.

I store a lot of things on it to show to my professors, please try not to make breaking changes like this

[Russ Cox]

Hi Natalia,

On Mon, Oct 6, 2025 at 11:15 AM Natalia L. Skirrow <dronebetter...@gmail.com> wrote:

LaTeX now functions, but not as it originally did; the spacing and scaling are both way off, and integrals inside exponents now render as though they were done with \int\limits_{...}^{...} all the time. ...

The new Mediawiki has dropped support for the old image-based TeX rendering. The choices are MathML and MathJax. The default is MathML but you can change it in your user preferences at https://oeis.org/wiki/Special:Preferences#mw-prefsection-rendering. That said, I looked at your Stirling page with my preference set to MathJax and it still doesn't use compact integral bounds you showed in your alternative screenshot. I'm not sure whether that's still possible.

 

I store a lot of things on it to show to my professors, please try not to make breaking changes like this

Obviously we try not to make breaking changes in general, but the OEIS is not meant to be a personal notebook. For that I would suggest Jupyter notebooks or any of the other notebook software that supports LaTeX notation. The benefit of using one of those is you control its configuration, how often you update it, and so on.

Best,

Russ

[Natalia L. Skirrow]

I appreciate that I shouldn't expect support/warranty, but it is not just stylistic; further down in that very page, separate lines in the same equation (for Raabe's formula) intersect.

this is extremely likely to have caused errors in mainspace pages as well, I just haven't been looking for them.

[Gareth McCaughan]

On 07/10/2025 10:51, Natalia L. Skirrow wrote:
> I appreciate that I shouldn't expect support/warranty, but it is not
> just stylistic; further down in that very page, separate lines in the
> same equation (for Raabe's formula) intersect.

The following is entirely off-topic (w.r.t. this thread, at least) but I
find it amusing that Natalia has to refer to "the approximation that is
not Burnside's" -- since there is also a famous "lemma that is not
Burnside's" (the one about counting things up to symmetries).

Perhaps disappointingly, so far as I can tell the "Burnside problem" and
the related "free Burnside groups" really _were_ introduced by Burnside
himself.

--
g

[Jason Bard]

Appears to be down again.

[Jason Bard]

Never mind, perhaps it was a blip.

[K. Voskuyl]

There are still timeouts

Op ma 20 okt 2025 om 09:44 schreef Jason Bard <j4yb1...@gmail.com>:

--
You received this message because you are subscribed to the Google Groups "SeqFan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to seqfan+un...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/seqfan/c54c3cfe-bcb9-40e0-b794-ed06e550720bn%40googlegroups.com.

[Paolo Xausa]

I'm experiencing timeouts several times per day, with a Cloudflare message.

[Rémi Eismann]

There is a major Internet outage with AWS US-EAST-1 even Cloudflare is impacted but the situation seems to be slowly returning to normal: https://downdetector.com/

Apart from that, there has been a slowdown on OEISWiki since the update. oeis.org is still as fast as ever, though.

Rémi.

[Alex Violette]

Hey SeqFans,
I'm having problems with trying to get the OEIS website to load. Seems we're having another problem again.
-Alex V.

To view this discussion visit https://groups.google.com/d/msgid/seqfan/29009882-8b72-4e9e-b015-7dcdcc438b80n%40googlegroups.com.

[Russ Cox]

Hi all,

My apologies for the continual mini-outages. The upgrade two weeks ago seems to have made us outgrow the machine, and we are running out of memory. I am going to take it down tonight and move everything to a larger system, and then things should be boring again.

Best,

Russ