A
A
asnarskiy2020-03-19 14:52:40
Law in IT
asnarskiy, 2020-03-19 14:52:40

How to correctly write down the criteria for downloading a web application in a contract?

Hello everyone,

we ordered the development of a web application from the studio. The application has a relatively large logic (10 thousand editable rows in the database, bound by conditions), the first time the main page was loaded for about 40 seconds, after which the application worked quickly.

Then, the studio "cached" something there and the download was reduced to 3-6 seconds on a desktop and 12 seconds on a smartphone at 4g (which is critical)

Our audience mainly comes from mobile devices. Now, we are ready to make improvements to optimize the speed, but we don’t know how exactly to write this down in the contract.
The studio's wording "we optimize speed" is, in our opinion, blurry. Several questions arose:

- How exactly to register "fast loading" in add. agreement? Do I need to be tied to some specific numbers, for example: "the application should load at slow 3g in so many seconds", or is this nonsense?

- How realistic is it to make a 1-2 second download on smartphones of applications on vue + yii2?

Perhaps the questions are not entirely correct, but we will be grateful for any useful comments. (-:
We don't want to pay and get a slow product.

Answer the question

In order to leave comments, you need to log in

1 answer(s)
T
ThunderCat, 2020-03-19
@asnarskiy

firstly, the download speed depends on 3 parameters:
1) The speed of code execution on the server, which in turn is divided into the actual code, and database queries (which most often create the main load and brakes).
2) Data transfer rate and data volume. The first depends on the location of the server, the distance from the request point, the server channel and the client channel. The second, as you understand, usually depends on the quality and amount of code (html / js / json / pictures) transferred from the server to the client.
3) Page rendering. Depends on the included scripts / styles, blocking of the JS code and the volume and complexity of the markup.
All three parameters affect the download speed, but the server-side code can be checked purely programmatically, from there it must be measured on the server by inserting execution timers into the code. The second and third parameters can be viewed in the browser, in chrome there is a download speed analyzer, such a built-in speed test.
For an approximate assessment of the work of your service, I would compare your performance with approximately similar large analogues of your site.

the first time the main page was loaded for about 40 seconds
This is hell, it shouldn't be like this anyway. It clearly indicates that the architecture is either not thought out, or clearly sloppy on the knee, in the hope that "the cache will pull everything out."

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question