Panie Profesorze @NowakFar , proszę o wskazanie gdzie dokładnie w recenzowanej pracy znajduje się przez Pana zachwalany model?
Prof. dr hab. Artur Nowak-Far, recenzja z 6.07.2026.
s. 3: „Przyjęty model badawczy ma cechy oryginalności. Wynika to z uzupełnienia wątków analitycznych natury jakościowej (już obecne w nauce) wątkami, w których zastosowanie mają metody ilościowe - bardzo ważne dla przedstawionej w rozprawie analizy z zastosowaniem mierników, wskaźników i analiz z pewnymi cechami ekstrapolacyjnymi.”
„W konkretnym, deklarowanym przez Pana mgr. Jakuba Wiecha, modelu operacjonalizacyjnym badań (i analizy) przedstawionej w pracy chodzi zasadniczo o ustalenie wielkości luki efektywnościowej (dotyczącej polityki publicznej w dziedzinie funkcjonowania rynku gazu ziemnego) w momencie przystąpienia Polski do UE i wielkości tej luki w 2019 r.”
„Model zastosowany w pracy jest - w moim przekonaniu - prawidłowy.” S. 6: „(c) wypracowanie i przetestowanie modelu analitycznego, świetnie łączącego aspekt synchroniczny i diachroniczny”.
https://t.co/WJ34TH3tan
Matt Mullenweg has accused CFO Mark Davies of conspiring with board members. Meanwhile, BlackRock's latest filing has valued Automattic at 83% below its 2021 peak.
🔗 https://t.co/u60wK1lrvU
Five of our current clients are at least 100 years old.
One is over 200.
And I am counting active contracts only, not our past clients.
There is something different about working with manufacturing businesses this old. They carry legacy, but not in the lazy "legacy systems" sense. They have proof that they know how to stay in business: they are still here 100 years later.
Nassim Taleb popularised the Lindy effect: the longer something without a natural expiry date has survived, the longer it can be expected to keep going.
It is no guarantee. But look at the base rates. The U.S. Bureau of Labor Statistics tracked private-sector establishments born in 2013. Ten years later, only 34.7% were still operating. Manufacturing did better at 43.6%.
A hundred years puts all the enterprise jokes in a different light.
These businesses have lived through market crashes. Technologies changed around them. Management changed, and founding partners are of course long gone too. They have had a hundred years of opportunities to disappear.
They did not.
They already know how to run a manufacturing business. We have spent 15 years building B2B platforms on WordPress and WooCommerce for manufacturers.
Our consulting starts where those two types of expertise meet. We challenge the brief and redesign weak parts of the process before we build on top of what already works for them.
We build the B2B platforms and online product catalogues their distributors and sales teams use every day. WooCommerce gives us the engine. We decide how it should work for the business.
After 100 years, "disruption" is a strange thing to promise them. I can even say that we are keeping their business UNDISRUPTED by applying all of our knowledge and engineering to their online platforms. How that sounds as a promise?
I swear, starting from sept 11th I’ll start reporting everyone who is selling WordPress plugins to the CRA Article 14 violation when they don’t have VDPs or proper security reporting info for their plugins and at the same time shit on researchers for reporting issues to a “wrong place” or in general be rude or ungreatful for basically the free work they receive…
Cloudflare does not cache your HTML by default. You can be paying for Cloudflare and still have none of your HTML cached there. Would you be able to tell this is your case?
By default Cloudflare caches by file extension, not by content type. CSS, JS, images: yes. The HTML document, the thing PHP and the database assemble: not by default. Without an explicit rule, every page view still goes to the origin. I have seen more than one setup where "we have a CDN" meant the logo loads from the edge of the network and everything else works exactly as before. Check the cf-cache-status header. If an HTML document comes back as DYNAMIC, the edge is not caching it at all.
What I also hear on calls sometimes are two words traded as if they were two products: CDN and edge. They are not two products. It is the same few hundred machines Cloudflare keeps around the world, and the two names come from different decades. CDN says what the thing does, deliver content. Edge says where it is located, at the far end of the network. For twenty years those were words for the same servers, and then the industry started using edge for the code you can run on them. Nobody planned that, so it is not on you if the two names blur together.
What is different is not the name but the job you give those machines. Handing out copies of files, your images, CSS and JS, comes with the account. No decisions, it's just simple distribution, working from the moment you point your domain at them. The other job is running your logic up there: assembling and caching whole HTML pages, telling a logged-in visitor from an anonymous one, holding the product page while letting the basket through untouched. It's not on by default because it needs rules: what may be stored, for whom, for how long, and what wipes it. And you are the one who knows those rules.
And the principle that ties this whole series together: cache layers are a cascade, not a set of independent switches. The lower layers have to hit before anything propagates up and reaches the edge. Getting that order right is messy work, and it is the reason a site feels smooth when someone has done it.
So before you add another layer, you should be able to say three things about it in plain words. What it stores: files, whole pages, or answers from the database. Who is allowed to get the same copy: everyone, or does a logged-in visitor need their own. And what removes that copy when the content changes, plus who is able to trigger it. If any of the three is a shrug, you probably have something messed up and don't know it yet.
We created a great guide to understanding caching specifically for WordPress. Link in comment
There are two hard problems in computer science: naming things, and cache invalidation.
It's an old Phil Karlton joke, and every two weeks I talk to someone who has the cache done wrong on their website.
The difficulty isn't anecdotal, it comes straight out of how the layers work. The moment you cache a response, you hand over control of the URL. Literally: HTTP has no mechanism to delete a response already stored at an intermediary with a long max-age. Requests hit that copy and stop reaching you, and the protocol gives you no channel to reach into it and remove it. There is nothing you can do but wait for its TTL to expire. The "purge" button in your CDN panel exists precisely because the protocol cannot do this. It is a side channel bolted on next to the spec.
So when and how stale content disappears gets settled while you design the platform, not on the first phone call asking why the old version is still up. In practice you pick one of two: a rigid TTL and waiting for it to expire, or purge on publish. Either way you have to know what a visitor sees in the window between.
Second limit: fresh does not mean fresh. Freshness counts from the moment the response was generated on the origin, not from when it reached you. Whatever cache held the copy on the way stamps an Age header on it. If it arrives with Age: 30, thirty seconds of its life are already gone upstream. You set max-age=50, you really have less than half.
Third, and the most misunderstood: cache does not fix a slow backend. It only makes you rarely see it. Every MISS, every logged-in user, every expiry hits the real, slow backend. And the worst moment is predictable: a popular post expires and a thousand requests see a MISS in the same second. If the layer cannot collapse them into a single request to the origin, the backend that "was not there" comes back all at once.
On the manufacturer platforms we run, this third limit has a concrete face: the logged-in distributor. Anonymous traffic gets the cached page and moves on; the dealer who logs in to place an order bypasses the full-page cache on every click and sees the backend's true speed. The most valuable user of the platform always sees it without the makeup.
Fourth, the one that's on nobody's radar: cache does not know what it is caching. Forget the private directive on a personalized response and a shared cache will hand one user's data to another. We fixed exactly this during an audit for one recent client. MDN calls it by its name: a personal data leak.
Cache buys you time and scale. It does not buy order or stability: a slow origin stays slow, and a wrong response served from cache is simply wrong faster and for more people.
Next episode: how to stack the layers, and why some of these solutions only start to make sense in large systems.
Cache does not make your code faster. It makes your code not run.
From the Batcache readme (Automattic): pages generated in 200 ms are served from cache in 5 ms, and the same WordPress install takes up to twenty times more traffic. Same fact from two sides: optimization makes work faster, cache makes work not happen. The interpreter never starts, the template never assembles, the database gets zero queries. A thousand hits on one post, the work done once.
The underrated part: spike survival. A thousand identical requests at the same moment, and a mature shared cache lets ONE through, then serves the rest its result. Request collapse, cache lock in the docs. Batcache says it straight: "preventing a flood of traffic from breaking your site", even if new users get a page a few minutes old. Freshness traded for staying up.
And the WordPress paradox of the week: plugins use transients "for performance", and by default transients land in wp_options. The exact database they were supposed to relieve. Only Redis takes them out.
Part 2/4 on how caching really works on a WordPress platform. Next: the things cache will not fix for you.
"We have caching enabled" carries zero information.
Between a user's click and your database there is a whole stack of caches. Browser cache keeps full responses on the user's disk. CDN and edge keep copies close to the user. A reverse proxy (Varnish, nginx) keeps whole HTTP objects. Full-page cache keeps rendered HTML. Object cache keeps query results. OPcache keeps compiled bytecode. At the bottom, the database.
Each layer remembers something different and expires by different rules. None of them knows the others exist. How would they? You can have a hit in one and a miss in every other at the same time, so "I enabled caching" can mean seven different things, and some of them do not do what you think.
A performance diagnosis that starts with installing another cache plugin (usually with "cache" in the name) is garbage. Start with the question "which layer", then read response headers. Every layer leaves a trace there.
Part 1/4 on how caching really works on a WordPress platform. Next: what this stack takes off your server. The numbers are better than you suspect.
This sounds as if you were describing a picture. A still frame. Snapshot of the current situation. Which is not true because the business is, as the great ai sage says, „everchanging”. Growth of output enabled by proper use of novel tools is as low as meeting expectations.
If an employee doesn’t know how the employer makes profit (not revenue) from their work, this is only harder for them to get a raise. If on the other hand someone fails to raise the output using methods available to them, by definition their work situation will be reviewed in the least favorable of ways.
Its like being in denial of capitalism.
Anita, all businesses are sweatshops! Thinking any different is being naive. Output improvement is expected and forced. you don't make any economical progress without this paradigm. And a pay rise is a triangulation of many different factors, all optimised by a given business to themselves in a way the collective output of the whole company can be maximised. It's this simple.
I literally can’t believe this was a real, honest question asked by someone genuinely lost at answering it on their own. Even more so, if this came from an employee.
Are we in a day care? It sounds like coming straight from the post covid IT boom era of WFH 4 hr workweek. We are in a „ramp up” phase where people figure it all out, but it is just a little more time before the finish line moves for good and forms the new baseline expectation around toolset and productivity.
Entitlement and tools adoption pushback always trigger me.
Your mapping is correct btw
My train headed for Krakow broke in the middle of nowhere (where else lol) and we are waiting for another train to take us there.
So 3 fire brigades arrives „to assist the passengers with changing the train”
Stay tuned
New agency milestone unlocked: someone is now impersonating us in phishing emails to our clients. Fake addresses: [email protected] and [email protected]. We only write from OSOMSTUDIO .com. Got something suspicious? Reach out directly.