If I google `"libxmlb2" site:packages.ubuntu.com`, google returns no results. I've tried this with a couple of other package names, like cups and firefox. The robots.txt is permissive. The Debian package repo, packages.debian.org is showing up and duckduckgo has the ubuntu package repo indexed.

    dig +short -t txt ubuntu.com|grep goog
    "google-site-verification=8CcvY6fsgRbiFk2nyEaiA9rE-ykXOcWUPjJKlO6P4vk"

I would guess that someone may have used their search console interface to purge/disallow everything despite the robots.txt allowing it. Another possibility is that after purging the caches they may have blocked google. Only the admin that verified the ubuntu.com site in the Google search console would know for sure.

Was the reason for your attempt to search in this way that you got the “internal server error” message (that I just got) when you tried to use the website’s search feature?

Here's the Spolsky article, called "Strategy Letter V": https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/

Here's links to the whole series up to VI: https://lettersremain.com/joel-spolsky-on-strategy/

And the n150 supports AVX2: https://www.intel.com/content/www/us/en/products/sku/241636/...

The PSP has 32MB of RAM plus 4MB of embedded RAM, upgraded to 64MB of RAM in later models, the browser also is limited to only a subset of that. I wonder how old the anecdote is? It's hard to imagine any websites now working on that. It does have javascript, but a version shipped with Netfront browser from 2011 or so. https://en.wikipedia.org/wiki/PlayStation_Portable#Web_brows...

Nice to see that Astro supports non-js browsers. Perhaps website developers, especially ones needing to develop for government services, should use 1GB Raspberry Pis for testing, but that would still have modern javascript. I got a 1GB RPi 4 in about mid-2019 and was rudely reminded of how much memory Chrome used even back then. More than 1 tab open and it would be killed.

The Eden quote is here https://shkspr.mobi/blog/2021/01/the-unreasonable-effectiven... and it has some stats of gaming consoles using uk government sites.

I used PSP browser a lot in the mate 2000's, and don't remember having much trouble. It could even display pages that, at the time, ware pretty bloated, like the Flash-based Homestar Runner cartoons.

What's wrong with web server logs? That would give you the number of visitors, if not the pageviews on an SPA.

Related recent threads about react performance:

JavaScript-heavy approaches are not compatible with long-term performance goals 178 points by luu 3 months ago | 237 comments https://news.ycombinator.com/item?id=47029339 (should be called "Why React is not compatible with long-term performance goals")

Does anybody like React? (jsx.lol) 242 points by brazukadev 15 days ago https://news.ycombinator.com/item?id=48274077

I should have added this one from 2022:

https://news.ycombinator.com/item?id=30528473

SPAs Were a Mistake (gomakethings.com) 692 points March 2, 2022 | 620 comments

Uber was recently said to have a $1500 per head per month AI budget (1). Business use dwarfs consumer spend.

(1) https://news.ycombinator.com/item?id=48383056

I'm surprised that Go doesn't default to AVX2 support by now, considering that Haswell started shipping in mid 2013.

Speaking of Dr Lemire's suggestion of a V5 architecture level, would that make any sense given the fragmentation of AVX512? None on Intel consumer devices, but it is on the last few generations of AMD.

The fragmentation of AVX-512 is a legacy problem.

All the CPUs introduced after Ice Lake (Q3 2019), with the exception of Cooper Lake (Q2 2020; a server CPU with a modest installed base), which support any kind of AVX-512, support all the AVX-512 subsets of Ice Lake (which has very important additions over V4).

This includes all AMD Zen 4, Zen 5 and Zen 6 CPUs, which form the bulk of the non-server CPUs that support AVX-512. Thus 6 years have passed since the introduction of an AVX-512 CPU that is not compatible with Ice Lake (and 7 years since any such CPU that was in widespread use).

Both Intel and AMD have stated that from now on features will be added to AVX-512 (a.k.a. AVX10), not deleted, which will allow in the future the testing of the AVX10 version number to be sufficient for determining CPU capability in this domain.

It would make sense to define a V5 level that includes all instructions of Ice Lake and also a V6 level, corresponding to AVX10.1 (Intel Granite Rapids) or to AVX10.2 (Intel Diamond Rapids).

Here is a CPU from 2020 that does not support AVX nor AVX2 https://www.intel.com/content/www/us/en/products/sku/199288/... very low budget and probably not common but this is why one might choose not to require AVX2

Wow, that's interesting, apparently 10th gen Pentiums and Celerons don't have avx. The wonders of product segmentation. They apparently corrected course (and also gave consumers AVX-512) in the 11th gen: https://www.tomshardware.com/news/intels-latest-celeron-and-...

There's a related discussion on hn from 2020 about lack of AVX here https://news.ycombinator.com/item?id=24577069

> The world is out of fab capacity.

Can anyone expand on this point? I read an article saying that the big AI co's datacentre spend was a bunch of lies because they can't build datacentres at anywhere near the rate they want to.

From what I understand it’s mostly TSMC and the memory providers being out of capacity over the next few years.

So it’s not even about datacenters.

Here’s a Reuters article about TSMC: https://www.reuters.com/world/asia-pacific/broadcom-flags-su...

So this is actual committed contracts with all kinds of companies such as Apple, NVidia, AMD.

Also, the whole reason they can’t build data centers faster is precisely because of this.

> they can't build datacenters at anywhere near the rate they want to

That was because the supplies the datacentre needed were constrained - supply-constrained, not end-user demand constrained, so would be in agreement with the GP comment (and the article I read didn't imply anything about lying).

What quantisation do the creators intend this to be run at? They talk about 16GB of ram, so should it be run at 8 bit? People here are talking about using q4, but I would have thought a smaller model like this wouldn't perform well at such low bits per parameter. Edit, it looks like their bechmarks would have been done at 16 bit float, as the hugging face release is that size: https://huggingface.co/google/gemma-4-12B . Which is a little deceptive: they're advertising an 8 bit size will fit on 16GB laptops, while releasing a 16bit size.

I guess we have to wait for someone to produce perplexity curves at different Q's.

They haven't made one for this new model, but Unsloth has a comprehensive quant KLD map of Gemma 4 26B A4B here: https://3215535692-files.gitbook.io/~/files/v0/b/gitbook-x-p...