VI |
Привет, Гость! Войдите или зарегистрируйтесь.
in 2026, Excel can't distinguish long numbers ?
Even worse, it deletes significant figures without warning.
The cell format looks especially mocking as a "Numeric" one, when an accuracy of 2 decimal places is immediately set, but the signs before the comma are destroyed.
The number in the picture is obtained by manual input. Excel has the right to crop the calculation results, not the manual input.
The simplest solution is to have double precision for entering manual numbers. We don't know what goals a person is pursuing.
Maybe I'm looking for sequences with calculations, and every sign is important to me.It's like if Word deleted every 16th word in a sentence, because everything is clear anyway.
The simplest thing is the format of the cells of variable bitness. If a person entered 20 characters manually, then the cell must be 10^20 bits. Not necessarily 2^128 .
But even I understand it's easier to have 2 cell formats: 2^64 and 2^128.
Remember IPv4 and IPv6. They didn't bother with 2^32 and 2^33 (although that would have been enough for the coming years.)
Repeating analogies: the transition of the entire world's software from AMD86(х32) to AMD86(х64) is about 2^128 times more difficult in complexity than for Excel to introduce a new data storage format with with an accuracy of 2^128 . But the whole world accomplished this task almost easily and bloodlessly.
I suspected that MS would do this chore for about 2^128 years.
The transition from x32 to x64 required expanding their knowledge to create software for millions of programmers. The new processors required reconfiguration (just throwing it into the trash of History) of the lithographic equipment.
The transition from IPv4 to IPv6 required the replacement of all network equipment worldwide.
Don't compare this to moving to high-precision cells within a single software product.
it beautiful But an Excel sheet can't look like dozens of bigadd()'s in each cell.
This code should have been in Excel itself many years ago. The VBA speed is 0.1% of the C speed.
This is a house on crutches, not on a foundation.
This was justified in the 1980s. When you were saving memory.
Let's take an analogy with UTF-8, the base characters are encoded with 1 byte, the national characters are 2 bytes. Special characters are 4 bytes.
The encoding itself chooses how many bytes to give for which character. User should not think about how much space his number takes up. User should get his number accurately, and not "save on matches" at the cost of an error.
You are defending a memory-saving approach when all the developers of these standards have already died.
Bill Gates said "640K ought to be enough for anybody."
MS’s entire history is built on these "documented limitations" that eventually become anchors dragging down progress.
Adding 2 numbers of 4th precision is difficult for you. ?
Tell this to your phone, which in the game calculates a 3D scene with lighting calculations, with 10 reflections from surfaces, with 6 layers of textures on each surface, with 10,000 rays from each light source, and ... It can repeat all this 140 times per second . And it does it on a processor weighing 10g.
I just call MS Marketing amateurs looking for excuses to do nothing. EXCEL is cutting down on precision for integers because it doesn't have enough space in its brain for cells.
I think that in the 80s, the transition to 64-bit processors began with just such posts on the BW Internet. And the authors of such posts heard the same thing there: don't talk about it, buy paper and write letters to intel. Do not interfere with working on the command line!
You are now playing the role of that retrograde from the past.
The demand to delete my posts is an admission of their low reasoning. It's just like the good old 80s.
I will not rewrite the known facts:
The transition to a 64-bit architecture required a physical change in the crystal topology, as general-purpose registers doubled (from 32 to 64 bits), and their number increased from 8 to 16.
To produce such sophisticated chips with an increased number of transistors and new data buses, the industry was forced to switch to new, more subtle lithography processes.
The old lithographic equipment, sharpened to the old standards and architectures, was literally sent to the landfill, since it physically could not print circuits of the required density and complexity.
The development and implementation of x64 cost billions of dollars due to the need to refurbish factories and change photo masks.
I'm talking about different scales of similar tasks. In both tasks, it was necessary to double the number of digits.
And for software EXCEL, this task does not require retraining of either users or programmers. They shouldn't notice the transition at all. This is an internal task within MS.
The above-mentioned tasks are millions of times more complex, including financially. Increasing the bit depth in EXCEL is the easiest task listed. And she's the latest. Because of someone's laziness. Or from a desire to blurt out the problem. That's what you're doing now, hiding behind your "20 years."
In fact, MS uses the achievements of the global industry and uses 64-bit processors, but limits the cells to only 2^49. And where are the other 32 options ? Where you can write all the service bits. For example, how many characters are lost 
.
You are defending a 40-year-old stagnation while sitting on a 64-bit throne that engineers and lithographers built for you.
You forbid me to talk about the difference in scale, because this is an "EXCEL software problem", and routers and processors have nothing to do with it. Of course, it has nothing to do with it. Both of them have already solved their problem, which is millions of times more complicated.
Your 20 years of lying experience doesn't excuse you. And don't look for your stupid supporters here. Почему так грубо ? Потому что я пропустил эту глупость в 1 раз, но 2 раз -- нет. Wiki quote:
Internet Protocol version 4 (IPv4) was the first standalone specification for the IP address, and has been in use since 1983.[1] IPv4 addresses are defined as a 32-bit number, which later then became too small to provide enough addresses as the internet grew, leading to IPv4 address exhaustion over the 2010s. Its designated successor, IPv6, uses 128 bits for the IP address,
This is literally known even to children. And there's not your meaningful conjunction "AND" to hide behind. Only IPv6, uses 128 bits , not IPv4 .