|
Toe-MAE-Toe, Toe-MAH-Toe.
He waited way to late to make that claim.
I’ve given up trying to be calm. However, I am open to feeling slightly less agitated.
|
|
|
|
|
Meet VS Code Docstring Skeleton Generator and the AI Doc Writer extensions to simplify your documentation process //this is a comment
|
|
|
|
|
this is just silly. I click on these articles and it's a re-hash of the last 40 years. In all the time I have been writing code, and dealing with other peoples' code, I have never, ever seen self-documenting code. Maybe the title is just wrong?
Why do we need to generate documentation for code these days? To be honest, I thought it was the old "should I comment my code" debate.
The problem is that code documentation describes the "what" when the most important part is the "why". Help me as the developer behind you (since you left the company and got a big raise) get into your head with what you are thinking.
That's what I need to know.
Charlie Gilley
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
Has never been more appropriate.
|
|
|
|
|
37GB of Microsoft's internal source code has been leaked online by the Lapsus$ group Don't tell me they got the source for calc.exe too?!
|
|
|
|
|
So Lapsus has released how not to do a search engine. 
|
|
|
|
|
Looking at a typical Windows system shows thousands of threads, with process numbers in the hundreds, even though the total CPU consumption is low, meaning most of these threads are doing nothing most of the time. This had me in stitches
|
|
|
|
|
I rebooted my Win10 PC to kill stale processes/threads, to get a feeling for the 'minimum' process/thread count. It takes surprisingly long time (3-4 minutes) for all sorts of initializations to complete; then it settles at 120 processes, 1180 threads, with a few more of both popping up regularly.
I am not at all worried! Why should I? I have had to explain to dozens of (non-professional) PC users that several of the tips and tricks to speed up your PC essentially is BS. One really stubborn 'placebo tip' says 'Delete those programs you no longer use to make those you use, run faster'. I know lots of people that insist on a significant speedup after a round of file deletions. But the disk page is still there - why would the CPU speed depend on whether the disk page contains the initial formatting bit pattern, program code pointed to buy a directory entry not being referenced, or program code no longer referenced from a directory entry? Even if you run a wipe program to zero the contents of freed disk pages, so the program code goes away: Why would a non-referenced page affect CPU speed?
We may of course construct extreme situations, such as if the allocation strategy uses a bit map, and only a handful pages are free, and they are at the very end of the bitmap: Search for a free page may be a long one. So you delete a file, and a free page pops up early in the bit map, and the next allocation may be a microsecond faster. But no one will notice a microsecond speedup, and the saving comes in for disk page allocation only. Yet users 'notice' a significant speedup ...
Half-studied guys may construct the most fanciful 'rational' explanations for the speed of their PCs: A friend of mine regularly vacuum cleans his (tower type) PC, to keep it from slowing down. He claims that dust will hinder the ventilation airflow, causing the OS to reduce the clock speed to avoid CPU overheating, which would slow down program execution. So he vacuums it every few weeks to keep speed up.
Like unreferenced disk pages: Why should we worry about a few thousand idle threads, and a few hundred idle processes? Sure, if you have used your last RAM page, and even the bottom entry in the MRU page list is in active use, then a few 'superfluous' entries in the process and tasks lists may, at least in theory, increase paging activity by 0.0001%, or even (if paging is disabled) may cause your system to crash due to an 'out of memory' condition. The answer to such issues is not worrying about idle threads or processes, but to buy more RAM!
Some people would claim that the OS will spend more time searching these terribly long lists of entries. So write yourself a tiny test program that runs through a list of ten thousand objects, scan the list ten thousand times, and measure the time used per iteration. Program in C, and turn down all non-essential activities, to get as stable timings as possible. How often do you think the OS has to traverse a list of this size? Reduce the list size to half, five thousand objects, to see how much time that saves you.
The article writer seems to be extremely worried about e.g. thread pool processes, suggesting to reduce their number. Apparently, he thinks it cheaper to create new thread objects at high frequency than skipping over an idle thread in a list ... Thread pools are used to reduce overhead, not to increase it! Having a sufficiently large number of already-initialized worker threads ready to jump in to do some work at a snap is certainly not slowing down application execution!
I am really happy about the number of threads that are sitting there, ready to jump in at any time. The cost of managing a thread that makes no system calls, does not generate a CPU load, does not occupy I/O channels, does not cause any interrupt, is very close to epsilon squared.
|
|
|
|
|
trønderen wrote: A friend of mine regularly vacuum cleans his (tower type) PC, to keep it from slowing down. He claims that dust will hinder the ventilation airflow, causing the OS to reduce the clock speed to avoid CPU overheating, which would slow down program execution. So he vacuums it every few weeks to keep speed up.
Badly ventilated PCs (to include desktops from a number of major OEMs  ) will benefit from regular de-dusting because their baseline thermals are already awful enough to throttle regularly so even a marginal decline will start throttling sooner. OTOH the proper solution is better cooling so that +5-10C of dust doesn't throttle you more and you can be lazy about cleaning your system out.
Did you ever see history portrayed as an old man with a wise brow and pulseless heart, weighing all things in the balance of reason?
Is not rather the genius of history like an eternal, imploring maiden, full of fire, with a burning heart and flaming soul, humanly warm and humanly beautiful?
--Zachris Topelius
|
|
|
|
|
Yep,
Pavel Yosifovich[^] has been awarded MVP status many times over the last decade. He made some great contributions to the Windows Internals books. Good to see his blog here in the news. 
Best Wishes,
-David Delaune
|
|
|
|
|
What he didn't do is take all but 4GB out of his system and then relook. The number of "idle/waiting" threads will drop dramatically as there is memory overhead associated with them. Many system services are written to scale to memory availability and logical processor cores (includes Hyper-Thread cores in the count).
For instance, right now I have 230 processes and fewer than 2800 threads on the system I'm using.
|
|
|
|
|
Based on its Project Z-Code, which uses a “spare Mixture of Experts” approach, these new models now often score between 3% and 15% better than the company’s previous models during blind evaluations. Zat iz where you zwitch all the z'z for z'z
|
|
|
|
|
Map of the Internet 2021 visualizes the most popular websites in the style of an old historical map "Catch the trade winds in your sails. Explore. Dream. Discover."
Beware the Straits of Stonks. Fair winds and following seas.
|
|
|
|
|
I believe the map has been bowdlerized.
Your map upholds good morals by omitting naughty websites.
--Oh, so you're familiar with them?
|
|
|
|
|
Unless you are very careful from day one, any shell script above a certain complexity level is almost guaranteed to be buggy… and retrofitting the correctness features is quite difficult. If you use them wrong, you can have problems!
|
|
|
|
|
So the more complex your script, the more chance it has of having bugs and issues? Lucky for me that is strictly a shell scripting thing. Heaven help us all if that were to trickle down to other coding languages.
|
|
|
|
|
Quote: any shell script above a certain complexity level is almost guaranteed to be buggy
Applies to more than just scripts.
|
|
|
|
|
Skilled coders are receiving almost double the number of interview requests as demand for software soars, with specialist programming languages commanding top salaries. You only need one
Unless you're me, in which case it seems to take 30 or more.
|
|
|
|
|
Go
Ruby on Rails
React
I guess I'm not interview material. I refuse to use those three (and there are many more languages/frameworks/libraries as well!)
|
|
|
|
|
I was going to say, this is just another BS article. RonR is ancient, and to push it as a tech skill is just weird. It's like old school HR people are writing these articles.
Give me a developer full of piss and vinegar any day. I don't care about your frameworks, can you learn?
Charlie Gilley
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
Has never been more appropriate.
|
|
|
|
|
In a new update to Windows 11, a watermark has appeared on the desktop wallpaper for unsupported systems, alongside a similar warning in the landing page of the settings app. Did the old warning have a bad icon?
|
|
|
|
|
this is so much MS bs, it's time to go to the fight club.
Charlie Gilley
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
Has never been more appropriate.
|
|
|
|
|
The result demonstrates one way to avoid the massive amount of computation typically required to tune an AI system, an issue that could become more of a roadblock as such programs grow increasingly complex. Sorry, no DOOM
|
|
|
|
|
if petal length between a and b
and petal width between c and d
and sepal length between e and f
and sepal width between g and h
then iris species 1.
repeat for the other two iris species.
vs. gigabytes of training data and a complicated neural net and hours of CPU training time.
Yup, AI is definitely NOT the solution for certain types of problems. But, lacking any I in the general population, programmers included, let's all code everything in AI! 
|
|
|
|
|
Marc Clifton wrote: But, lacking any I in the general population,
Found the I in "general population".
|
|
|
|
|
Kent Sharkey wrote: Sorry, no DOOM
... yet!
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
