The Stack Overflow 2018 Developer Survey results came out today with the raw data to follow in a few weeks. Here's what I could gather related to Delphi from the analysis:
The Stack Overflow 2018 Developer Survey results came out today with the raw data to follow in a few weeks. Here's what I could gather related to Delphi from the analysis:
Delphi made two prominent appearances. The first, unfortunately, was on the list of "most dreaded" software. :-( Basically, 65% of those who said they used Delphi did not list it in the list of software they wanted to be using in the future. It didn't show up on the list of most wanted software either (people who weren't using a technology now but hoped to using it in the future) even though that list went down to results with only 2.6% of responses. When the majority of your customers wish they weren't and essentially no one who isn't a customer wants to be one, you're really in trouble. :-(
https://insights.stackoverflow.com/survey/2018/#most-loved-dreaded-and-wanted
Delphi had some good news in a surprising place: It showed up prominently in median salaries - U.S. only. I think there's something off about this figure though.
https://insights.stackoverflow.com/survey/2018/#top-paying-technologies
First, 75% of Stack Overflow survey respondents are under age 35. I don't think it's controversial to say that Delphi age demographics are quite different. The average Delphi user probably has ~20 years experience, which means salary differences could be attributed much more due to experience differences than technology differences.
Later though on another language forum someone pointed me to a chart I'd missed:
https://insights.stackoverflow.com/survey/2018#work-salary-and-experience-by-language
This is fascinating to me. It's for the global data. Look at where Delphi sits on this graph: Delphi users have the highest average years of experience of ANY language listed on the chart, EVEN COBOL. This confirms I was right about age/experience differences coming into play. However, what's just as interesting is where it sits on the median salary axis globally: almost the LOWEST median salary!
Some Delphi folks like to claim that fewer Delphi jobs mean higher salaries, while some of us claim to have observed the opposite. This shows that we're right. Years of experience and scarcity of jobs have not led to high-paying Delphi jobs (globally). As Stack Overflow notes:
"Developers using languages below the line, like PHP and Visual Basic 6, however, are paid less even given years of experience. "
Someone in another forum noted that the languages being underpaid for experience tend to be, in their words, "legacy" languages like VB6 and Delphi.
But this brings us back to the U.S. salary results - at $100,000, the median U.S. Delphi salary is more than DOUBLE the median salary globally! That's a HUGE difference in results. We'll have to wait until the raw data is released, but my guess is that there must be very, very few U.S. Delphi users in the survey to be so skewed compared to the global average. The only other explanation would be an enormous outlier. Unless one of you is being paid 2 million dollars a year though, my guess is that U.S. Delphi users have almost gone extinct. Maybe Nick Hodges is the last of his kind. :-) I don't know, but I can't wait to get my hands on the data to find out.
In 2017 "Desktop applications developer" scored 28.9%. This year, "Desktop or enterprise applications developer" scored just 17.2%! This whole field is going extinct!
My personal favorite revelation was just 49.9% of developers used Windows for their development work, marking the first time Windows has fallen to less than half of developers! Sadly, this also means that Embarcadero eliminates half of their potential customers by offering a Windows-only IDE.
There were many good questions asked about gender and sexuality and software ethics. Sadly, it seems most males rank issues regarding a diverse workplace at the very bottom of their list of priorities in judging a prospective employer. :-( For women and transgender people this was the second-highest priority. We've still got a long way to go to eliminate the pervasive biases in the tech field.
These survey results are well worth a read!
https://insights.stackoverflow.com/survey/2018/#most-loved-dreaded-and-wanted
Delphi made two prominent appearances. The first, unfortunately, was on the list of "most dreaded" software. :-( Basically, 65% of those who said they used Delphi did not list it in the list of software they wanted to be using in the future. It didn't show up on the list of most wanted software either (people who weren't using a technology now but hoped to using it in the future) even though that list went down to results with only 2.6% of responses. When the majority of your customers wish they weren't and essentially no one who isn't a customer wants to be one, you're really in trouble. :-(
https://insights.stackoverflow.com/survey/2018/#most-loved-dreaded-and-wanted
Delphi had some good news in a surprising place: It showed up prominently in median salaries - U.S. only. I think there's something off about this figure though.
https://insights.stackoverflow.com/survey/2018/#top-paying-technologies
First, 75% of Stack Overflow survey respondents are under age 35. I don't think it's controversial to say that Delphi age demographics are quite different. The average Delphi user probably has ~20 years experience, which means salary differences could be attributed much more due to experience differences than technology differences.
Later though on another language forum someone pointed me to a chart I'd missed:
https://insights.stackoverflow.com/survey/2018#work-salary-and-experience-by-language
This is fascinating to me. It's for the global data. Look at where Delphi sits on this graph: Delphi users have the highest average years of experience of ANY language listed on the chart, EVEN COBOL. This confirms I was right about age/experience differences coming into play. However, what's just as interesting is where it sits on the median salary axis globally: almost the LOWEST median salary!
Some Delphi folks like to claim that fewer Delphi jobs mean higher salaries, while some of us claim to have observed the opposite. This shows that we're right. Years of experience and scarcity of jobs have not led to high-paying Delphi jobs (globally). As Stack Overflow notes:
"Developers using languages below the line, like PHP and Visual Basic 6, however, are paid less even given years of experience. "
Someone in another forum noted that the languages being underpaid for experience tend to be, in their words, "legacy" languages like VB6 and Delphi.
But this brings us back to the U.S. salary results - at $100,000, the median U.S. Delphi salary is more than DOUBLE the median salary globally! That's a HUGE difference in results. We'll have to wait until the raw data is released, but my guess is that there must be very, very few U.S. Delphi users in the survey to be so skewed compared to the global average. The only other explanation would be an enormous outlier. Unless one of you is being paid 2 million dollars a year though, my guess is that U.S. Delphi users have almost gone extinct. Maybe Nick Hodges is the last of his kind. :-) I don't know, but I can't wait to get my hands on the data to find out.
In 2017 "Desktop applications developer" scored 28.9%. This year, "Desktop or enterprise applications developer" scored just 17.2%! This whole field is going extinct!
My personal favorite revelation was just 49.9% of developers used Windows for their development work, marking the first time Windows has fallen to less than half of developers! Sadly, this also means that Embarcadero eliminates half of their potential customers by offering a Windows-only IDE.
There were many good questions asked about gender and sexuality and software ethics. Sadly, it seems most males rank issues regarding a diverse workplace at the very bottom of their list of priorities in judging a prospective employer. :-( For women and transgender people this was the second-highest priority. We've still got a long way to go to eliminate the pervasive biases in the tech field.
These survey results are well worth a read!
https://insights.stackoverflow.com/survey/2018/#most-loved-dreaded-and-wanted
Nick Hodges
ReplyDeleteI almost got tricked into writing an answer to your question, but then I remembered I already did, in 2015, on the Embarcadero forum. It was long, it was impassioned, it quoted Larry Wall at length for some reason. What was the result of this effort? Here it is:
Bruce McGee (MVP): "Holy sweet mother of blather! It's like two dogmatic, but uninformed hamsters were wrestling on someone's keyboard for 40 minutes."
Rudy Velthius (MVP): "TL;DR. I guess it is what comes out if you set monkeys to produce the works of Shakespeare. "
Nick Hodges (MVP, to whom the reply was addressed): " I'm sorry, did you say something?"
Marcus Humm: "Hm? Quite wrong reaction in my opinion as he's right in some points.And did you notice? He even mentioned your name! ;-) Did you really read it? (yes it's a bit long)"
...
Nick Hodges (MVP): "Are you kidding?"
Markus Humm: "Yes presentation could be a bit better, but your answer sounds like you didn't learn how to gallop through a text. Why not just search for your name in the text and read the passages around it?"
Nick Hodges (MVP): "Because I fell asleep."
Earlier in the same thread, which led to my reply to you, Bruce McGee accused me of lazy reasoning. I cited 14 citations, from Jeff Atwood to Infoworld, Dice, Wired, and O'Reilly Media to prove my point. I quoted two or three sentences from most of them so one did not have to click the links to find the relevant parts. The entire response?
Bruce McGee (MVP) quotes
Joseph Mitzen: "...I also defend my arguments with actual facts and research."
Bruce McGee (MVP): "And a great deal of misinformation, distraction and verbal diarrhea. Who has time to wade through all of that crap to look for the occasional nugget of truth?"
Somehow Bruce is still an MVP and Jeroen Wiert Pluimers isn't.
Like Charlie Brown, Lucy and the football, I've been tricked too many times into believing that certain MVPs actually want a sincere discussion. The last time we tried having a discussion you objected to my detailed analysis regarding number of Delphi programmers in the world solely on epistemological grounds, upset that I couldn't "truly KNOW". Then for fun I quoted from your own blog where you declare that you believed Delphi use was growing - you didn't have any proof, you just "felt it in your heart". :-) :-) :-) Bruce McGee also openly declared that he believed that he, himself, knew the size of the Delphi market better than the publishers at O'Reilly Media, which I recall Dalija Prasnikar found quite a telling remark.
In essence, to have a discussion one needs to be open to the thoughts of others. What this community often considers "discussion" is more like a pro wrestling match. In fact, those last two sentences contain the germ of the answer to your question!
Maybe things have changed, maybe you and I have changed. But if you think I believe that "no one should use Delphi", maybe they haven't.
If you're really interested in anything I have to say, hunt around this group a bit and find the part where I basically tell Lars that if Embarcadero had followed two key ideas you had under CodeGear the future of Delphi would have been rainbows and sunshine! :-) You might not take the time to read what I write, but I dug up some of your old writings in the meantime and saw a potentially different future if you'd been allowed to bring your ideas to fruition.
Nick Hodges Largely bitterness, I think.
ReplyDeleteJoseph Mitzen Please accept my apology for being rude. There's no excuse for that.
ReplyDeleteHowever, I confess I find it hard to read such long posts. That's my problem, however.
How about this -- could you sum up in, say, 100 words, what you reasons are for being here?