Forem: Sa

Magmi "1364 Field doesn't have a default value" error - Just be careful

Sa — Mon, 29 Apr 2024 21:30:33 +0000

When importing a fairly large number of goods using magma, I suddenly came across a problem:

SQLSTATE[HY000]: General error: 1364 Field 'image' doesn't have a default value - ERROR ON RECORD #1

Of course, all the fields had all the values.
Then what is it? A Magmi glitch? Magento? Or is it Mysql at all?

I will not give the whole way of finding a solution, I will only describe the result. The Problem was... in me!

The next to the image field in the import file was the width field. In the admin panel width field is spelled out with an indication of the dimension - 2 mm, 3 mm, 4 mm.

And in the import file, of course, just 2, 3, 4.

As soon as the dimension was specified in the import file, everything started working again, like a Swiss watch!

The site is broken due to a hosting error... Who is to blame?

Sa — Sun, 14 Jan 2024 21:25:35 +0000

The mystery of the two unknowns: the site was handed over to the client, fully operational, there were no complaints.
After some time, the site stopped displaying correctly, which the client complained to me about. After digging around inside, I realized a simple truth: some security settings have changed on the hosting. And now, when trying to upload a style file, the client sees a 403 error:

It turns out that it's not my fault. Hosting also tried its best. But the site is not working.
Who is to blame?

Ecwid - good for everyone, if not for one "BUT"

Sa — Sat, 18 Nov 2023 14:36:30 +0000

Ecwid is quite interesting solution for creating an online store, if there is no desire to dig into the code, databases and other programming subtleties.
It is enough to create your own store in the ecwid admin panel, and then just insert a piece of code on the site - without correcting anything, without changing, but simply "CTRL+C" - "CTRL+V":

<script data-cfasync="false" type="text/javascript" src="https://app.ecwid.com/script.js?XXXXXXXX&data_platform=code&lang=YY&data_date=ZZ-ZZ-ZZ" charset="utf-8"></script>

And then a whole store is already available on a certain page of the site:

And it seems that everything is fine - the job is done. But this may be true only if the goal is not to make this store indexed by Google!
Pay attention to the internal links in this store:

See the "#!" in the link? This means that ALL THE PAGES OF the STORE WILL BE LOADED using AJAX. But Google has firmly and unequivocally made it clear that it will not index pages generated by this method: https://developers.google.com/search/docs/crawling-indexing/javascript/fix-search-javascript

And now our joy has almost come to an end.
What to do in such a situation?

There is, of course, a way out, but it is very long and not very simple. And, perhaps, it would be easier to initially create an online store on another platform. But since the store has already been created and there was no desire to transfer all the data, the task was set to solve this issue with the already existing situation.

Let's take a look, first of all, at the list of categories of this store:

You can see that absolutely every category is also loaded by ajax. And what if we can replace "#!" with some secret word, and the result will show the same - dynamic? That is, in this case there will be like https://athos.guide/products/monastic-handicrafts

That's not bad, because it's just a URL that Google can index! Of course, then we need to do the same with the products in each category. It was:

And now:

A completely readable URL that Google likes!
Now let's observe how well these links will be indexed - because now nothing restricts their indexing...

Shopify - the peculiarity that I hit my forehead

Sa — Sun, 05 Nov 2023 10:46:10 +0000

Another client has a store created on Shopify. It is important to clarify that the store sells bags, casual clothes and clothes for an active lifestyle and yachting: https://rewind.com.ua.

That is, three main areas of activity.

And then the task comes: to present these main directions on the main page in the form of cubes. Where each cube represents the goods of its category.

The idea of building was simple: to set an array of categories, and even better - a ligament of the product-category, so that you can decide in manual mode which goods to be withdrawn in cubes. And then, sorting out the array, to configure its conclusion. Nothing complicated.

Working mainly on PHP, such a problem is solved in 3 minutes. If not for one "but" - in Shopify you can’t clearly set an array!

Not wanting to spend a lot of time learning Liquid, I turned to GPT for help.

His idea turned out to be simple and brilliant at the same time: yes, it is impossible to set the array directly, but you can set the line and divide it (as "split" in JS) into array!

And to perform all further actions as if we have an ordinary array. In principle, this is an array, only obtained not in a direct way, but by converting the line.

A trifle, but unpleasant! Thanks to GPT for saved time!

Flower store app. Part 2. Strangeness out of nowhere

Sa — Sun, 22 Oct 2023 12:33:27 +0000

While talks continue about the feasibility of an app for a flower store, I decided to just take a look - what kind of site is proposed to be embodied as an app?
Having run it through all known diagnostic tools, I found one very strange feature: the strange piece of code appears ONLY when checking it through Google's rich-results tester.
Example: look at source of this page - https://saholaflowers.com/. Screenshot of last few lines:

And now the same page, viewed through Google tools: https://search.google.com/test/rich-results/result?id=foGWPW0ALJ5aOdsWqBcjPw

Do you see a new piece of code after the closing script tag?

I'll remind you, this is source code - there can't be a dynamically added JS DOM. But this piece of code only appears when viewed through this tool.

This piece of code has the site's old domain, saholany dot com. I have checked the WHOLE site, every file, the entire database - there is not a single mention of this old domain.

I've reset all the cache on the hosting, on the site, on the browser. But this code is still there.

Not only that, if I make changes to the site code (like adding a tag with the class test-tag-for-goole), the tool will see those changes - so it's not a Google cache or some other cached value, no:

So what is that?

And even more important question, how do you clean this up? This is the first time I've encountered a situation where I can't find the reason for the code appearing on my own site....

To be continued.

Flower store app. Does it make sense?

Sa — Fri, 13 Oct 2023 17:40:49 +0000

I'll try to be brief.

There's a store in New York that sells and delivers flowers: https://saholaflowers.com. The owner asked an off-topic question: would an app, in addition to a website, make sense?

And I'm a bit stumped: usually the client doesn't ask "does it make sense", usually the client says "do this or this for me".

I decided to experiment with chatGPT again: what could it tell me about it?

I won't give you the whole correspondence, I will only summarize my conclusions: the chat offered to analyze whether this application would be relevant? That's a good point, but that's exactly the question I asked the chat!

Thus, for now, my question remains unanswered.
To be continued...

ChatGPT - is it strong and should you be wary of it?

Sa — Thu, 12 Oct 2023 10:21:09 +0000

I won't list the concerns that the chatGPT has raised among the average Internet user - there are tons of them, you can read for yourself.
But here I gave him an assignment: "Can you suggest a regular expression in PHP that would remove all attributes of all tags except src in the text?"...
After each answer he gave, the standard task clarification began:

only the SRC attribute should be left, the rest should be removed
the content of the SRC attribute should remain as well
this code doesn't work, check yourself
in this version the code does not work, because all attributes except style|class|alt remain

And so on.

After playing with it for about 10 minutes, I quickly found a solution here https://stackoverflow.com/questions/33093509/php-regex-remove-all-attibute-except-href-in-html and it is very correct at least from the position that it allows you to easily replace href with src, add a couple of attributes - that is, it gives a flexible, customizable solution. Chat could not give me such a result.

Hence two conclusions:

At this stage, it is not worth being afraid of it - the revolution of cars is still postponed :)
To use it, you need to be able to do what you are asking it to help you with!

And the second conclusion makes you wonder: what is the purpose of a chatGPT? The answer is simple - as a generator of ideas and directions for finding an answer! This is where it is great, as it gives us a squeeze of the entire Internet, saving us hours of searching and processing garbage!