|
Is using a C++-based IDE to write Java code OK? If so, why not the reverse?
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
Writing Java code would be frustrating per se.
Java-based IDE are slooooooooooow.
"In testa che avete, Signor di Ceprano?"
-- Rigoletto
|
|
|
|
|
Eclipse will compile 25,000 java files in minutes and build a comprehensive database of the code during the compilation.
Maybe it is just slow dealing with C++? You might check for linting/ validation options that can be disabled.
|
|
|
|
|
Eclipse was fairly new when I tried to use it, maybe 15 years ago. Maybe the build speed was high even then, but the interactive response was extremely slow. Unless you were running the fastest PC on the market, you were typing blindfolded; you saw no echo from your keystrokes, and for all supplementary functions you might as well take a walk to the coffee machine to refill your cup while waiting.
We suspected - and I still do - that the extremely sluggish interactive response was a result of a bad porting to the MS Windows environment, done by developers who had not learned how to build Windows applications (if this is correct, Eclipse is certainly not alone with such problems!), because Eclipse had a good reputation in *nix environments - that's why we thought it would be the best tool for us. But none of us had sufficiently powerful machines to run it; it was thrown out.
If the problem was a poor port to MS Windows, the problems may have been cured within a release or two. I don't know. To us, Eclipse was so extremely bad, performance wise, that we said: Never more Eclipse. Today, that may be wrong, yet I will be very reluctant to give it a new try, even after that many years. It was that bad. (Besides, I haven't been coding Java for years, and have no plans for picking it up.)
|
|
|
|
|
trønderen wrote: because Eclipse had a good reputation in *nix environments
Sadly it is the best IDE on *nix environments, but not due to its merits.
GCS/GE d--(d) s-/+ a C+++ U+++ P-- L+@ E-- W+++ N+ o+ K- w+++ O? M-- V? PS+ PE Y+ PGP t+ 5? X R+++ tv-- b+(+++) DI+++ D++ G e++ h--- r+++ y+++* Weapons extension: ma- k++ F+2 X
|
|
|
|
|
den2k88 wrote: it is the best IDE on *nix environments
Nah, Emacs was always the best. 
|
|
|
|
|
Emacs is a good OS, but it could use a decent editor.
|
|
|
|
|
Quote: We suspected - and I still do - that the extremely sluggish interactive response was a result of a bad porting to the MS Windows environment, done by developers who had not learned how to build Windows applications
Some things were (maybe still are) difficult to port to Windows.
I had a bad experience and dug into it with a benchmark that: Opened multiple small files, read them, wrote them, then closed them.
The time do do this on a Windows machine on NTFS with the set of test files I had was around 1m30s. On the same Windows machine, running a virtualbox Linux, with the same files on an ext4 filesystem the benchmark took around 12s.
The reason? On Windows, various applications can (and do) hook into the "open", "write" and "close" filesystem calls. If you're running just 1 AV and a filesystem indexer like Everything, each time an application opens a file, it's three times the overhead of doing the same on some other OS.
For applications that may open lots of small files, I don't know how you'd "port" it to Windows other than by rewriting it to not use multiple small files. For something eclipse, I'd imagine that simply referencing other java projects via maven or similar is going to result in many thousands of files being downloaded.
|
|
|
|
|
On the other hand  : I'm using a very old java based tool to search for text in files in windows directory structure. And it is the only one I trust on the results   old furt me 
|
|
|
|
|
I just use VS Code. It's got a few wrinkles, like intellisense just randomly dying on certain files, but honestly I'm surprised it can autocomplete my layered C++ template constructions at all.
To err is human. Fortune favors the monsters.
|
|
|
|
|
Message Closed
modified 15-May-23 19:06pm.
|
|
|
|
|
Considering that the dodo may be on the way back Gene editing company hopes to bring dodo ‘back to life’ | Extinct wildlife | The Guardian that may not be a good example, anymore.
But no, bug reports will still need to be made. Bugs will still occur, some of them will be potentially catastrophic (what do you mean, the plane tried to land 2 meters below the runway???). So bug reporting will still be needed. AI might be able help narrow down the nature and/or location of the bug, maybe even one day correct the bug, but for now, as others here and in other locations have noted AI is very poor AI at writing code. I can't imagine it would do any better at fixing bugs. And companies producing software will still be liable.
Keep Calm and Carry On
|
|
|
|
|
A "meter" is a device that measures something. A "metre" is a distance of about three feet, or 1000 millimetres.
|
|
|
|
|
Actually, that is not true in the USA or Canada.
|
|
|
|
|
I doubt it - from what I see of AI, it can't create, just rehash whatever it was given as input: look at MS recent fun-and-games with Bing; look at ChatGPT output when posted in QA as a "Solution". Most of the time it won't even compile, let alone do what is needed, so what makes you think it could fix a bug? At least, without introducing a dozen more?
And there are areas where bugs and the reporting of them is really important: 99% of Tesla recalls since Jan 22 have been software problems*
If those bugs weren't reported, do you think they would be fixed? Would you be happy with a bank that didn;t fix bugs? A plane? A mobile company?
I don't think you would ... and I don't think anyone would be happy with a car, bank, phione, plane that had it's software fixed by the current generation of "Artificial Intelligence" which does seem to exhibit "Artificial Stupidity" instead.
* a quick google:
99% of Tesla Vehicle “recalls” Since January 2022 Were Simple Bug Fixe[^]
Tesla is forced to 'recall' all Full Self-Driving Beta with update, NHTSA says may cause crash | Electrek[^]
Tesla Recalls 130,000 Vehicles in U.S. Over Software Bug | Hypebeast[^]
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony
"Common sense is so rare these days, it should be classified as a super power" - Random T-shirt
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
Try solving your next question with ChatGPT and you will get a reasonable idea of the answer.
|
|
|
|
|
Some humans are hopeless at writing code, most are barely adequate and some are brilliant at it. The same will apply to AI systems. The first few have been rushed into service by suits who want to test the water (and maybe make a few quick bucks), but I have no doubt that in time there will be specialist code-writing AI systems, trained to do that job, that will work very well. Equally AI debugging systems trained to do that will also emerge.
But I question whether the present crop can be called "intelligence". They do not think, they repeat what they have been "taught". They are just moderately sophisticated computer programs. When asking the same question or setting the same task repeatedly returns better and better answers (confirming that the system is actually considering previous answers and thinking a bit more about it), and showing signs of imagination in their thinking, only then will I consider the label "AI" appropriate.
|
|
|
|
|
haughtonomous wrote: but I have no doubt that in time there will be specialist code-writing AI systems
Why?
Why would they attempt to specialize to a market that has 27 million (in the world) when they could specialize to the average consumer - facebook with 3 billion users, google with 8.5 billion searches a day?
Then consider that World of Warcraft has about 8 million monthly players. Not as many programmers but if you integrate into the game then it exposes all of them. But with programmers you would need to sell to each. And much easier to provide value for WoW.
Or the number of people that watch soccer is about half of the population of the world. Even just in the US 20 million probably bet on soccer matches. So do something that helps them to decide on what bets to place. Again much easier than attempting to support programming.
|
|
|
|
|
You'll first have to convince the AI it's a bug and not a feature. Not possible if marketing trains it. Everyone will customize their AI to comply with their business model.
"Before entering on an understanding, I have meditated for a long time, and have foreseen what might happen. It is not genius which reveals to me suddenly, secretly, what I have to say or to do in a circumstance unexpected by other people; it is reflection, it is meditation." - Napoleon I
|
|
|
|
|
|
As long as there are customers, there will bug reports. Don't see AI being involved in this at all except to possibly weed out actual bugs from the growing geriatric population of users that make mistakes.
|
|
|
|
|
Do we really need AI to code something like:
void insert_bug_report(bug const& bug_report)
{
if (user_age > GERIATRIC)
{
send_reply("ID10T error!");
}
else
{
add_bug_to_database(bug_report);
}
}
(With apologies to all the oldsters who haven't reached their dotage)
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
Replace send_reply("ID10T error!"); with
Replace send_reply("OLD FART FAULT");
"A little time, a little trouble, your better day"
Badfinger
|
|
|
|
|
|
Said who?
"A little time, a little trouble, your better day"
Badfinger
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
