Month: July 2016

Once, I delved into how search works in hybris and sketched out a diagram. It turned out to be seven times longer than the preview you see here. But it should be quite useful for those who develop on Hybris.

It shows the order in which classes and operations are executed from the moment a search request is made until the result is passed back to the controller.

@[100003472810093:2048:Vyacheslav Khudyakov] @[100004031421822:2048:Alexey Lyubimov] @[100000077047562:2048:Marina Zhigalova] @[100001894770015:2048:Viktoriya Shaimardanova]

https://hybrismart.com/2016/07/21/how-facet-search-works-internally-in-hybris/

This is how we spend our evenings 🙂 Lisa was filming

The well-known “Business Youth” is coming to Philly and San Francisco soon to teach Americans how to do business the Russian way, for a fee. Could someone buy me a ticket there – I’ll make an interesting report 😉

People occasionally ask me, was XXX difficult? To pass an exam, deliver a presentation, do something or fix something.

And what does “difficult” mean? I have only one answer for this: something is “difficult” when it creates an unmitigated discomfort (physical or emotional) and there’s nothing you can do about it but endure it or somehow reduce it.

For example, carrying bags from the supermarket home on foot every day is difficult. Dealing with idiots is difficult. Running a marathon without running experience is difficult.

Anything not associated with discomfort cannot be difficult by definition. If public speaking does not cause discomfort, then it cannot be difficult. It can be interesting, or as Americans say, challenging. If a task occupied your brain all your working and non-working hours, and you eventually solved it – that task was not difficult, but interesting. If someone runs the same marathon without feeling significant physical discomfort (to some extent) or when the discomfort is compensated by euphoria, then it is interesting, not difficult.

Although, in this scheme, you can change difficult work into interesting work, but it’s not always possible to change a difficult life into an interesting one.

P.S. My computer crashed almost simultaneously with me hitting the “publish” button. But I made it!

My first presentation in English in front of colleagues) talking about hybris

Live a century – learn forever. It turns out, 75% of banks use dye packs in bundles of bills that spray a chemical substance and heat up to 400 degrees when removed from the bank. As a result, both the money and the thieves are spoiled, and if not spoiled, then marked. Yesterday I saw this in the second episode of the second season of “White Collar,” where these gadgets are part of the plot, today I went to Google it.

https://youtu.be/5gzW4aKg2yg

http://science.howstuffworks.com/question671.htm

And indeed, in America, domestic dogs do not bark.

What a cool bike we have in the parking lot!

New post on my blog.

I managed to set up a little shop on my laptop with 2 million real products, divided into 15,000 categories. The products have attributes which can be filtered by – totaling about 6000 attributes. All this runs faster than ever on my laptop. I could have loaded more – but I didn’t find databases with larger volumes of products.

So, it can be said that on Hybris, it’s indeed possible to create a marketplace with an insanely large number of products, categories, and facets. And I understand how.

However, standard Hybris cannot manage such volumes at all. With a doubling of the product database, various processes like indexing start to operate, roughly speaking, three times slower. Therefore, the architecture in my solution uses a somewhat different approach. In a nutshell, all products are stored directly in SOLR, not in an Oracle or MySQL database. Product pages, product lists, searching – all this interacts directly with SOLR. The link explains how all this coexists with the shopping cart and checkout. @[100001168004708:2048:Erik Babadzhanov] recently asked about how to do this – here’s the answer 🙂

The full loading of products “from an empty database” to a fully functional site takes 25 minutes. Much less time is required for updating the data – for example, the prices for all 2 million products.

Unlike out-of-the-box Hybris, the number of active facets (filters) does not cause any “slowdown” in any processes. Currently, 6000 facets are set up in the demo, and the only barrier to their use is user interface restrictions.

On the link – a proof-of-concept video and technical details.

Tagging @[100001168004708:2048:Erik Babadzhanov] @[100001044160267:2048:Aleksey Kryuchkov] @[100004031421822:2048:Alexey Lyubimov] @[1817791335:2048:Victor Romanovsky] @[100000077047562:2048:Marina Zhigalova] @[100001894770015:2048:Viktoriya Shaimardanova] @[100001735299023:2048:Alexey Pronin] @[1698960808:2048:Alexander Zolotilin] @[100000571996239:2048:Maxim Antonov] @[100002859265802:2048:Ilya Timchenko] @[1328575098:2048:Max Shelukhanov] @[617283947:2048:Renata Mussina] @[1509384824:2048:Anatoly Mokhov] – this might interest you.

Bought a knife. The sticker honestly states that it contains chemicals known in the state of California to cause cancer, genetic disorders, and reproductive system diseases. Just like on cigarettes. Somehow, I got used to the fact that everything else usually has opposite disclaimers.

If I move to California, I must remember to leave the knife here.

Here is the list of these chemicals: http://oehha.ca.gov/proposition-65/proposition-65-list