Hacker News Reader

Comment by piokoch | Hacker News Reader

"Don't try to make a career out of optimizing the SQL queries to display a preference page on a line of business app at a company that no one has ever heard of. That is not the straightforward path to having other people learn you are capable of doing meaningful work."

This is so sad. Optimizing that SQL might require being really clever and can bring enormous value for the company that maybe nobody heard about but might be providing some not so sexy but useful service for half of the country.

I've never heard about company that is building software that runs nearby power plant, but certainly this company is much more useful and have more positive feedback for the society than something "sexy" like Facebook.

View on HN

What's a funny coincidence is that I've been thinking about almost exactly this for a little while.

Just in the last two weeks I decided to create a podcast where I interview the exact types of developers who optimize SQL inside business apps.

In fact, literally just last night I recorded and uploaded (to YouTube and the various podcast platforms) episode zero. Not an actual interview yet, but a statement of purpose and a placeholder so I can hopefully have a small audience on release day. (I'm hoping for about 50 subs by the time I launch the first)

I was 100% not going to promote on HN until after I had at least 10 episodes published... but fundamentally it was born out of a desire to attempt to solve for the "unsexy" programmer.

Anyway: https://youtu.be/9mmgfxSeqPA

Or you can search for "Ship It" on your favorite podcast platform.

In terms of feedback, any and all is welcome, though I've got a personal preference for constructive feedback as opposed to just telling me that my stupid face looks really stupid.

Please don't start another podcast. Or if you absolutely must, please provide a text version of the content.

Podcasts aren't searchable, they aren't skimmable and their content isn't indexed in search engines. Text is.

Sorry you're getting downvoted on this post. I found your data point useful. Even though I disagree that there are enough podcasts serving the senior developer / corporate programmer niche, being made aware of the existence of "podcast exhaustion" is valuable, and the discussion convinced me that a transcription service is worth it.

I'd hate for negative attention to result in a chilling effect on your decision to voice your opinion in future.

I can't remove your downvotes, but I can say thank you a second time.

I didn't downvote but I think "please don't do X" is really a poor way to frame this opinion. I also don't care much for podcasts and prefer a textual format, but if you think you'd prefer making it a podcast then just make it and theshrike79 and I will find other content to consume. Or maybe it'll be so good and interesting that we'll just listen to it anyway!

Deciding to start on such a project is really cool IMO, and I think if I ever attempted to do something like that I'd really find it very demotivating if the feedback I got was "no wait, not like that". But that's just me.

I agree with this 100%, and just want the parent commenter to see that at least one other person cares enough to say so to offset the toxic dismissal.

It’s easy to shit on work, but hard to create it. You created something, that’s awesome and I wish you luck! Don’t try to make all the naysayers happy, especially in this community, or you’ll go insane lol.

Given a HN post if just some persons opinion, and a podcast is just some persons opinion, aren't you failing your own logic by calling that post "toxic"?

i.e. If you don't like opinion framed this way, find other posts to consume.. Instead you're naysaying a HN comment..

I stand by what I said. The difference in my mind is that someone invested real creative energy into something, and someone else came along and said “please stop making this”.

Yes, I went out of my way to say something here. Put yourself in the shoes of the person making the podcast here. They excitedly shared their work early because it seemed like a great time and place to do it. And one of the first comments is “you should not have made this”. What people say online actually matters, I don’t speak for this person of course, but I know I would be disheartened by such dismissal of my work. I just hate seeing this maybe happening to others and sometimes try to offset the negativity I think I see.

I actually laughed a little at this one

> “you should not have made this”

... because that was 100% my initial reaction. I was like "oh shit, I just got internetted"

I bounced back pretty quickly... for a few reasons. I've been a consumer of podcasts for a couple years now. I know what I like and I know what is missing for the demographic like me.

Further - I went and looked up the "podcast listener" demographic.

Turns out I'm pretty much the prototype. Older professionals in high income families, commuters.

The size of the market is currently around $10b and expected to grow at a 27.5% CAGR until 2027. https://menafn.com/1101605880/81-Billion-Worldwide-Network-F...

I'm confident that both podcasting has a market, that there is a niche I can serve appropriately. Phase 1 is the part I will do for free because it seems like so much fun... but I'm also going to try to use it as an opportunity to prove that I can accumulate enough resources to support the development effort on phase 2 (as Elon Musk would say)

In phase 2, I have further ideas about how to improve the monetization model for the back catalog. I'm not 100% sure if I am going to say it out loud yet, but since I'm going to spend to night compiling the feedback I've received from various sources into a github repo, I may go ahead and in the spirit of openness just say it out loud.

If spotify steals the idea then maybe the ask to hire me so I can help. (Or maybe one of the other any other hosting platforms decides they want to eat spotify's lunch)

While those theories will need testing, I figure that IF I write the scripts to automate the process of increasing the CPM of the back catalog, I can probably move to the next logical step, which is providing them in a SaaS model to other podcasters who make their living from it as Phase 3.

IF 1,2,& 3 clear ... then phase 4 is creating a new sort of marketplace model where advertisers are directly connected with the aforementioned mid to top tier podcasters.

Phase 5 is to combine it all into a single platform, expand into video and, to crush both spotify and youtube before me , and hear the lamentations of their ad managers.

I might not make it to phase 5 - but - you know, it's nice to plan out a few steps ahead. It isn't like I'm going to try and beat Musk to driving the conversion of humanity into the role of a space-faring species, but each phase is roughly 10x on the previous.

As much as some dude on the internet "yucking my yum" was a little jarring as the first comment, the bull case on this party is ~$100B.

In another fun story... I came pretty close to closing my first $500 in sponsorships today. "I know I'll ask my boss and get back to you" is often a "no" ... the dude I was talking to sounded so enthusiastic about the idea it really lifted my spirits.

I'm not at odds with their opinion but rather with their tone and delivery. I'm not a fan of podcasts at all, and I get where they are coming from, but someone's sharing an idea they're clearly very passionate about, and I'm all for giving constructive feedback (that's probably why they picked this particular forum to share their idea, after all), even if such feedback hurts, but I also expect a modicum of positivity to go along with that, and their comment doesn't clear that bar for me.

Let me add another data point: I never listen to podcasts. Ever. I've tried it, and I hate the format. Even when I am forwarded something highly relevant to my interests, I still check for a transcript, and if there is none I move along.

You can't search audio. You can't skim audio. You can't easily hyperlink to audio, copy-and-paste audio, or auto-summarize audio. Audio is not indexed or cached by the big search engines. It's a terrible, terrible, TERRIBLE and inefficient medium to convey information.

Beyond all those very pragmatic reasons, there are also people like me whose personality type (introversion) makes us find listening to audio an emotionally draining task as it simulates interacting with people. I don't get energized by it--it is work. And why should I give you my working energy when I could just as easily have skim-read a transcript or blog post with the same information and gotten energized by the self-study approach?

THAT SAID, you said that you want to do interviews... and that is something that really calls out for an audio medium. Sure you can transcribe it, and people like me appreciate the transcription, but I recognize that the audio is the original source material and transcription is an additional expense. I'd rather have the full interview available than some highly edited transcript... but I would hope that there is at least a short summary of the content in the description of each segment.

I only listen to podcasts that are well-produced and edited. Very few actually are. There are way too many people who think podcasting is just turning on a microphone and babbling away with a friend or guest for however long they can do it.

I only have a few hours a day I can spend on podcasts and I'm not going to waste it listening to some inane meandering chitchat.

"Put it on the background then" ...yea, that's when FOMO hits. What if I miss the good part? You know, the one thing that they said in the description they talk about, but can't keep on subject and keep getting sidetracked.

This is why most podcasts I listen to are from the BBC or NPR, they've got one subject for the episode, it has a clear narrative through-line and the actually get to a point. Exceptions to the rule being 20kHz (audio production professionals) and 99 Percent Invisible.

At the risk of pedantically zooming in on your choice of words:

> I only listen to podcasts that are well-produced and edited. Very few actually are.

I disagree. Good production is a nice extra, but it's secondary. I'd rather listen to Miles Davis on vinyl, than a mediocre musician on high-bitrate FLAC.

Some examples: I recently discovered the Happy Path Programming podcast. It's a good listen as the hosts are knowledgeable and insightful. The production quality is fine, but isn't at the professional standard. The hosts acknowledge this, but make no apology for it, on the grounds that they don't want their podcast to turn into a chore.

In contrast, the Security Now podcast has professional production. Again that's a plus, but that's all. If the podcast were no good, their high production budget wouldn't be enough to save it.

Good production quality doesn't turn a boring podcast into a worthwhile one, and mediocre production rarely spoils a worthwhile one.

> There are way too many people who think podcasting is just turning on a microphone and babbling away with a friend or guest for however long they can do it. I only have a few hours a day I can spend on podcasts and I'm not going to waste it listening to some inane meandering chitchat.

Sure, it makes sense to insist on good quality podcasts. The hosts need certain soft skills to deliver a good podcast. Not everyone has them, but it must be easy to convince yourself you do. People may be tempted to think I could easily do a podcast, or comedy, or acting. The skills aren't as immediately obvious as in, say, gymnastics or theoretical physics.

Podcasting may be the victim of Sturgeon's Law, 90% of everything is crud, but fortunately it's pretty easy to filter out the ocean of poor-quality podcasts and find a few you like.

I think high production value is probably decently correlated to content; after all, people usually won't invest in something they believe to be low-quality. So in that sense it may be good enough a filter for a world awash in crud.

True, there's most likely a correlation. A bit like 'honest signalling' in evolutionary theory.

What I was getting at was more the fact that even if you record 3 hours of stuff, you don't need to put the unedited 3 hours out. It can be recorded on a potato microphone on a clay disk for all I care, just spend some time in the edit and cut it down to a manageable size.

There are very few podcasts that can stay on topic and interesting past the 1 - 1.5 hour mark.

The only exception to this rule are the Christopher McQuarrie Spoiler Special episodes of the Empire Podcast (now behind a firewall). The dude can keep spewing fascinating stuff about his movies longer than the movies' combined runtime =)

One more data point: Personally I only listen to podcasts on long commutes. This only occasionally happens when regularly needing to actually go to a physical office location in a nearby city. Even then podcasts are competing with audiobooks or long phonecalls to people I don't often get enough free time to talk to.

It's not that I don't enjoy the format, I do, but I find it harder to fit into my life.

I'll take that on board! Thanks.

For Episode 0 I did post the full text of my script in the description, but I do think that for the actual interviews I won't be able to afford a transcription service for a while.

For the last few years though, I've been getting a fairly significant proportion of my ongoing professional development via the medium.

I think it scales well in terms of being able to get access to some of the greatest minds I know without creating an undue burden on my guests OR myself as a father to a 2-year old with a mortgage and a day job.

I am 100% willing to commit to writing rich descriptions that address the topics discussed in advance though. As part of the editing process I'll need to rewatch my own interviews a few times anyways, so it isn't a significant amount of extra work to provide a perfectly reasonable value add that I like to see on the pods I listen to.

Thanks!

I use Descript for my podcast, £22 a month gets me ~30 hours of transcription. This probably sounds like an ad but really, it’s one of the best software products I’ve used in a very long time.

This is the most recent episode I’ve published, giant wall of text took a bit of editing but wouldn’t have been possible without Descript. https://www.mql.fm/005-working-in-martech-dan-graap

Thank you! This is exactly the type of recommendation that I was hoping for.

I'd have been happy with the mention even if it was an ad.

If you give it a try for audio editing make sure you record different podcast guests on separate audio tracks, makes the whole thing significantly better.

I wrote a few thoughts on my blog about how I went about starting with the podcast including tools used, might be useful? https://www.jacquescorbytuech.com/writing/starting-a-podcast

You don't even know how valuable this is to me at this exact moment in time.

My first interview is tomorrow night! Fortunately, it is with a friend and co-worker who has the hacker mindset and will be willing to work through hiccups in technology.

I'm trying to pick a video conference platform to try. Do I discord? Zoom? Try to find a way to set up audio line-in from my android phone.

I need a solution for tomorrow night so the blog post has found me at exactly the right time.

Thank you.

Edit: having read the blog, I thought that free zoom ended after 45 minutes! I didn't realize that one on one would allow longer sessions. Great!

The best part of zoom from my point of view is that people don't have to log in to anything to join me on a call.

If I wanted to use discord for example if have to make my guests install it or make an account there. I'm pretty sure that some of the big-time grizzled greybeards I know would probably sooner pull out of the interview than sign up for a new service.

And they're some of the ones that I'm most interested in hearing from.

100% zoom, if only because it lets you record a separate audio track per speaker. Make sure you’re both off mute when the recording starts otherwise the audio tracks won’t be the same length

You might want to try UberConference, which is free and also automatically transcribes conferences with high accuracy (note this may be a paid feature I'm not sure). It also does some extra neat NLP stuff to detect sentiment, topics of conversation, etc.

I’ve had decent luck using Squadcast: https://squadcast.fm/

Yes, a 100% this.

Even better if you can get the interviewee to record audio on their computer too, this way you can get proper audio without compression issues for editing.

Doesn't work for "video podcasts" because sound sync is a bitch, but for audio only it can enhance the quality immensely.

This is brilliant!

Especially for those times when you have internet issues on one end or another.

At the very least you don't lose your content.

On the other hand, podcasts can be listened to in the background, while text must be actively read. There are pros and cons to each format as a consumer, not to mention differences in production workflow.

To add, I think we should leave it to the machines to start indexing podcasts automatically. They're some of the easiest mediums for speech recognition and Google has been indexing YouTube transcripts for a few years now.

They have been doing it badly for a few years know yes, it’s not solved by any stretch of the imagination.

Sure, they keep getting some words wrong all the time, but it is likely good enough to get the general subject of what a podcast is about.

This is why most of the stuff I listen to as a podcast are just random factoids and trivia. (No Such Thing as A Fish, Cautionary Tales, 99pi, etc)

If it's something I actually want to learn and retain, I prefer text so that I can go back, pause, skip etc as I please with an intuitive interface - my eyes =)

The same thing with interviews of interesting people I may want to store quotes from. With text it's just C-c C-v to copy stuff to my notes. With audio I need to transcribe it by hand, which is tedious work.

Exactly

Naively I would think that there must be a service out there that auto-generates transcriptions from podcasts. Youtube already do this for (some) videos. The output isn't perfect, but it gets about 99% of the words correct.

So my suggestion to OP wouldn't be "don't do the podcast". Instead I would say "do the podcast, but consider that it might not be searchable without a bit of extra effort".

100% there are services! I believe that Azure ML/AI has a natural text transcription service.

In theory I could try and tweet at Seth Juarez@microsoft to see if I can get a sweetheart deal.

But the more I think about the previous poster's position on the text, the more I convince myself that it might be worth a few bucks an episode for me to pay out-of-pocket if it means improved accessibility for users with disabilities.

Not really because it's good business, but more because I've really come to believe in the value and power of implementing accessibility features.

I'm actually starting to get a little teary eyed thinking about that blindness activist's viral video where he broke down physically sobbing the words "thank you" when he saw all of the accessibility features that Naughty Dog added to The Last of Us 2. (I'll hunt the video for people who haven't seen it, but be warned that you will probably bawl your eyes out like a school-child).

From a cost perspective, I'll be putting in hundreds of fake dollars worth of time. The people I interview will be effectively donating hundreds of dollars worth of their own time.

The least I can do is put some serious thought into plunking down dozens of dollars of pocket money for a transcription service.

I'm more sold on the idea than I was in my previous post (5 minutes ago)

Edit: the tweet in question. https://twitter.com/stevesaylor/status/1271404306697158659

I haven’t personally used it but heard good things about Descript (also their demo vids are hilarious)

IBM Watson, Google, Azure, and AWS all have speech-to-text APIs. IBM's claims to distinguish between different voices, although when I used it a couple of years ago for Japanese it was a little lackluster. It's pretty inexpensive: IBM gives you 500 minutes free per month and is a few cents / minute thereafter; Google you gives 60 minutes free, and it's 4c/min thereafter. It's an API rather than a service, so you'd have to write a client to use it. IBM's API (and maybe the others) allows you to request time stamps on the output, so you could let people click on your transcription and seek directly to the part of the video they are interested in.

I suppose you could always do it on your own machine with Sphinx, too, although I don't know how they compare to the others in accuracy.

The best podcasts in my experience are:

1. Those that are tightly written and edited. That is, there is a script (which can then, of course, be posted).

2. Are discussions (lectures, interviews, conversations) about a text (book, article, concept, etc.) which itself can then be referenced for more detail.

3. Are narrations of short stories.

Examples of these would include:

- Numerous history podcasts: The History of Philosophy (Without Any Gaps) by Peter Adamson, History of Rome, History of Byzantium, History of China.

- The New Books Network, a series of author interviews on recently-published academic books. Here you're listening to the author and an (almost always academic) interviewer discussing the topic, effectively as an advertisement for the book. It's a great way to be introduced to deeper material.

- A few deeper / prepared podcasts. WNYC has numerous professionally-scripted podcasts, as do other NPR / public radio / government media affiliates in the US, Canada (Ideas), BBC (Tim Harford, various), and elsewhere. These blur the line between traditional podcast and commercial broadcast, but are quite intelligent and informative.

- Stories or nonfiction narrative. "Selected Shorts", "Soundprints", "This American Life", "Reveal", and others come to mind. Here the idea is to simply be entertained or caught up in the narrative.

- Various interview programmes. I'm fond of Terry Gross's "Fresh Air", the Santa Fe Institute's "Complexity" podcast, and others.

None of these is the proverbial dude/chick (and optionally friends) blabbing incoherently into a mic. All are scripted, most have either transcripts, extensive show notes and references, or both (the fiction / narratives being the exception). For the record, the unorganised / low-content flavour of podcast is one that drives me nuts as well. Sturgeon's Law applies. Seek the non-cruddy end of the spectrum.

For times when I'm read out, or otherwise engaged (kitchen, chores, driving), or want to zone out, these are worth listening to.

For conveying infotech content, I'd generally agree that podcasts aren't the best medium. I also generally stray away from anything that's strongly focussed on news, current events, or current topics, and the list of academic, historical, and philosophical programmes above have been quite the useful informative distraction from the moment for me, for the past several years.

Better yet, feed your podcast to an AI for transcription. And host the transcriptions in a wordpress blog. The cost is minimal.

Not every format is for everyone.

I personally listen to podcasts or audio books every day when driving and in my lunch break.

People are different, and I like to learn about other people's methods and the potential benefits thereof.

So let me ask ... when you're listening to podcasts, do you ever take anything away from them that is actionable? Do you take notes? Do you summarise?

How do you ensure that the lessons in the podcast don't get lost?

Do you "Listen with attention" or are they mostly just background?

I find that podcasts are there, then gone, and nothing remains. When I read I take notes, summarise, and use the results to enhance my personal version of a Zettelkasten. I can't imagine doing that efficiently with podcasts, and I'd be interested to know how I can change that.

Kinda depends upon if it is educational, informational or entertainment. I never use podcasts to just fill out the background, I generally use heavy metal music for that.

I tend to either remember how/where to find the information again or the actual thing. So with podcasts I will either remind myself that I can't easily reference it and therefore remember the lesson, or sometimes I will write it down later in the day.

What do you use for your Zettelkasten? I still haven't found a notebook software that I actually like, so my notes are kinda spread between a few different solutions.

I've been using a ZK-like thing since the mid-90s, so it's very much a "roll your own" solution. Over the past 12 months or so I've started to try to de-couple it from the specifics of my setup so I can share it, but it's proving difficult.

It's essentially a wiki, but it's how I use it that makes the difference. A true ZK requires a conversation. You take notes, but then you spend time inter-weaving them into the ZK. Most people use note-taking systems as, well, note-taking systems, and never return to make the connections between the notes. Hence they get lost.

Here are my thoughts about that "conversation":

https://www.solipsys.co.uk/cgi-bin/DiscDAG.py?DiscussionID=Z...

There's a link in there to take you out of "Neighbourhood Mode" if you want to see the whole thing at once.

Thanks, that is quite interesting.

Not all podcasts are for work / learning.

May be just listening for entertainment

> So let me ask ... when you're listening to podcasts, do you ever take anything away from them that is actionable? Do you take notes? Do you summarise?

I can answer this a little from my perspective.

There are three major inspirations for tech podcasts that I use personally.

The first two: https://www.interactive.org/Interviews/the_game_makers_noteb...

and https://www.idlethumbs.net/designernotes

are both in the format of "veterans talk to each other about their careers and the things they've learned". They always roll in the format of two old friends catching up, but in a guided way in terms of their life, experiences and learnings from along the way.

These are the ones that I want to emulate, but with a focus on programming for enterprise, business and startups.

I'm personally most excited in talking to what I think of very respectfully as "cogwheel" or "workhorse" developers. I think developers are vastly under-celebrated when compared to other creatives ... like some person who can cry on demand gets a tiny golden statue and applause ... but the person who re-writes 1500 lines of SQL into a one liner and reduces a week long job down to seconds DOESN'T?! You're telling me I have to go out and buy my OWN tiny little golden statue and play an applause loop on my cellphone?

That's bullshit!

At least that's where I want to start, I've seen some interviews in the last few days with people like

Ton Roosendaal - the creator of blender - https://www.youtube.com/watch?v=qJEWOTZnFeg

After I've developed the interview skills to talk to the people who genuinely make me feel a little like a programming fanboy. Uncle Bob, Douglas Crockford, Carmack, Hanselman, Jeffrey Snover, David Fowler? I've got way too much respect for them to even ask until I look at my own stuff and say "I'm good enough at this to do them justice" (edit, not that I don't respect my friends, but I think I have enough capital with them that they'd do it as a favor and be willing to re-take any technical mistakes)

But seriously ... go up to idle thumbs and listen to the Sid Meier podcast https://www.idlethumbs.net/designernotes/episodes/sid-meier-... (four parts!).

Or listen to Ted Price (founder of insomniac games) talking to Amy Hennig https://interactive.libsyn.com/027-amy-hennig-tgmn-v2

Back to your question - I don't actually learn much that I would use in my day to day. Sometimes there are nuggets, but mostly it just gives me a "warm feeling" to listen to masters talking to old friends about their craft.

NEXT there's the other type where I hop on to the GDC vault and download some of their panel discussion videos as audio files: https://www.gdcvault.com/play/1024856/VR-Best-Practices-Maxi... so I have 48 minutes of industry experts talking about the things they think about when they're designing for VR.

That might be directly applicable in the work I'm doing, or I might have additional questions to ask the next time we start with a VR project. I might have additional possibilities to offer the next time the creative, UX and, strategy teams come to me and say "we want to build for VR, what can we do with a budget of XYZ?"

I'll have a bit of general knowledge on hand for the next time I have to make a decision... or not ... I actually don't do that much VR right now. For some reason the idea of hundreds of people sharing a facemask in a crowded conference hall isn't selling very well right now.

At the very very least, listening to professionals talk about these sorts of things counts to me as a deliberate act of professional self reflection. Asking myself if I'm doing things in the best way I can? Are there better ways to do things? Are there things that I'd like to start working with, but can't fit into the budget of the current projects?

And so on...

Which again, is all stuff I'd like to try to help solve for with this coming podcast.

One other format that a friend and I started talking about being interesting might be to run "architectural differential diagnosis" where we formulate a problem statement and ask 2 or 3 experienced developers to come to a consensus on what technologies they'd use and why.

Though that's unlikely to happen within the first 10 episodes.

agreed

Have you actually tried searching "Ship It" on a podcast platform?

There is absolutely zero chance that I can find the correct one without knowing what the thumbnail looks like.

Search results in order:

  - I Ship It
  - Ship It Show
  - We Ship It Podcast
  - The Ship-it Show
  - Print It and Ship It
  - Ship It or Skip It
  - I Can Ship It!
  - ...

edit: I realized the above might sound a bit harsh. Sorry. Please do some branding to make the podcast easier to find. I was trying to find it in Overcast as I am actually interested in listening to it.

Not at all harsh! And super super valuable. So genuinely, thank you for this.

SEO and discoverability are critical.

It is abundantly clear to me that I've got to workshop the name.

The categories are

- "say-able" can a listener hear the name and find the right one without a spelling mistake?

- is it unique enough to have a top slot search result?

- does it explain what it's about in the title?

- Would someone in the target audience instantly know what it was about if they saw it in a list of 800,000 podcasts? (The number of different pods out there according to a research report I read before deciding I still wanted to do it)

"Ship it" is a reference to the thing they (sometimes) say on GitHub when your pull request is accepted.

The icon has the "ship-it squirrel"

I thought that was cut and referencial so people in the know would recognize it instantly.

It also has a special place in my own heart because I first learned about it when a pull request I sent to one of the Microsoft as-net docker repos. In the version of the story I remember Scott Hanselman himself commented on the PR and said "lgtm :shipit:"

Which on GitHub translates into a picture of a squirrel dressed up as a gangster.

Based on the rules above I rename to "LGTM: the podcast about delivering code for professional programmers"

Maybe just "LGTM - the programmer podcast"

Thanks again. I'm going to work on it.

Edit: or maybe I rename to "LGTM shipit"?

Note: I wanted to thank you again for these data points. I appreciate that you didn't have to put that effort in and I have now officially decided to go with "lgtm: shipit"

And I even made an issue in the github repo to thank you for this effort.

https://github.com/AlexChesser/lgtm-shipit/issues/15

I expect this repo to be a really good place to write readmes, articles in the wiki, put the transcripts, if at some point I get enough money to pay github for LFS ... maybe I even upload the raw content? I have to think about that last bit because I'm considering just going with the MIT license so maybe I don't want my raw content released under MIT.

Who knows? It seems a lot more appropriate to put a programmer-podcast into github than a wordpress blog.

Cool, I have subscribed.

You should probably also add it to Apple podcasts. (edit: I added it as an issue on github)

About using the Github wiki for articles, I think you would be better off having the actual content on your own domain. It will give you more flexibility down the line for monetization or just moving things around.

Sounds interesting. However, I already found two other podcasts with the same name.

I have searched on Spotify and Castbox, but didn't find yours there.

Thanks! I've been able to find it on a few platforms so far, but I'm distributing via anchor.fm

https://anchor.fm/ship-it

I'll have to find a way to differentiate my ship-it from others for people searching.

I'll probably have to rename it, but in the short term I'm going to let it ride.

Better to focus on quality interviews for now. (I think)

Edit: I just looked over the emails that anchor sent me and some of them say it could take as much as a week for the first listing to appear.

It's probably not showing up because it was like 1am last night when I finally hit the publish button on a non-test episode for the first time.

I'll keep an eye on this though. Thanks.

It's the exact type of hiccup that had me not wanting to self-promote before I had 10 episodes online.

edit - per the other discussion in the thread I've gone with renaming the podcast to "LGTM: Shipit!" which will hopefully make it a little more discoverable.

I don't expect as many people have the acronym for "Looks Good To Me" in their title, I removed the space in "Ship It" and hope that search engines trigger for both "shipit" and "ship it".

What's more is that without the space it's even more true to the inspiration for the title ... which is the :shipit: emoji on github.

https://anchor.fm/lgtm-shipit

OK, you got me, as I've been in those trenches myself. Subscribed.

Thank you!

You've reached that intermediate goal: I'm your fiftieth subscriber.

I think you misunderstand the thrust of that argument. It’s not the nameless company that’s the problem, it’s the business result of your work.

What looks better on a resume?

1. I optimized SQL queries for Nameless Corp to improve admin interface load times by 20%

2. I optimized SQL queries for Nameless Corp to increase profits 20%

If possible, always work on some version of option 2.

Counterpoint: when every developer is urged to build something with "measurable impact", you end up like Google, where people launch something visible, get promoted, and move on, leaving the service to wither away.

Arguably that means you have bad easy-to-game metrics. I wouldn't call what many (most?) Googlers do "impactful". More like "Let's keep a bunch of smart people busy so they don't get hired by our competition"

If the service you built withers away when you move to a different team, it was a pet project that The Business doesn't care about (because it has zero impact). Literally the definition of a bullshit busywork job :)

How is this a counterpoint? The above is advice to developers seeking to improve their career, not executives at large public companies making hiring policy decisions.

The fact that goog is well known for starting projects then killing them, suggests to me they don't really appreciate the measurable impact of their work.

How do you even know the actual impact you had on revenue/profits? Engineering is often many layers removed from customer interactions, sales and accounting. Some companies have complex and opaque revenue channels. Frankly I take any claims of direct revenue/profit growth/generation with a grain of salt. Typically at best, you're wildly guessing, at worst, you're outright lying.

The PM you worked with should be able to answer this question. If they can’t, your project is likely a waste of time.

The signal from a hiring standpoint is that you asked and cared about this. Engineers who will just do whatever are less useful/impactful than engineers who push back and care about the value they’re delivering. Even if it’s sometimes fuzzy to estimate.

Your original example is very specific, but your follow up clarifies that the reality is that it is vague guesswork. I think you could claim contribution, but very hard to define that line x of code written last Friday resulted directly in y% of profits.

To me it signals you're probably exaggerating. You better be able to back up your reasoning behind your claim. Bold claims require bold explanations.

It’s not vague guesswork at all. If you can’t measure the success of your project, why are you doing it?

It’s okay, even expected, if the answer is probabilistic. What’s not okay is doing stuff for no reason.

Bonus points if you’re able to tell a story like: We did X to achieve Y, but it didn’t work because Z. So we tried ...

This goes back to the tying of a SQL query directly to profits. That would be an interesting story if it were true, but such a story is probably an extreme rarity.

We're talking about calculating the impact of your weekly or daily work directly to profit generation. There is a level of granularity and detail to your claim that is likely impossible to realistically calculate, so it would be very hard to accept such claims at face value.

"I made the query for X, an operation that is performed Y times per day, Z% faster resulting in N% increase in Xs performed daily. This meant we could (defer hiring additional headcount|do more <profit-generating/cost-limiting process dependent on X> per day)."

Many teams don't invest in the tooling required to measure this kind of impact, but it's possible to ballpark it.

That’s a good example. The math doesn’t need to be precise, and you don’t need to track it down to lines of code. “Part of team of X developers who delivered product Y that produced $Z in revenue.” is enough for anyone to very roughly ballpark-estimate impact.

This is counter to GPs initial point, where there is a specific claim of direct profit generation. The advice is now just getting watered down into vague ballpark estimates and tossing in a revenue number onto your resume. The advice might as well just be "work on successful projects", which is "thanks captain obvious" advice.

The success of the project as a whole is measured, of course.

But it's 20 people working on it, from design to sales to ops to customers relations to devs. How could you possibly pin 20% of the profit (not even the turnover, but the profit) on one guy?

Even if he improved the SQL queries and that immediately led to more sales, making the choice to work on that over other things was probably not his.

Nobody is asking to pin all that on exactly one person. That’s a silly overly engineery way of looking at it.

It’s a team effort. Being part of the team who did X is enough. You can then talk about your specific individual contribution when appropriate.

Just like everyone in an NBA team “wins the championship” and gets to put “winner” on their resume. Or how hockey players say “I won the Stanley cup”, even though they did it as a team.

This seems to be completely counter to your initial post tying weekly/daily work directly to profit to determine what is worth your effort and what to put on your resume. You even give SQL queries as an example, so I don't see how that's not trying to tie it down to the lines of code.

There's a big difference between:

1. I optimized SQL queries for Nameless Corp to increase profits 20%

2. I optimized SQL queries for Nameless Corp on a multi-million dollar project.

> The PM you worked with should be able to answer this question

Most software devs work on tiny fragments of large projects. Let's say I work in Microsoft on Bluetooth stack in Windows 10 - how do I translate that into increased Microsoft profits?

How does the PM know? Are they told or is it something they are required to ascertain?

This should be part of the product spec or even tech spec. If you cannot measure the impact, why're you working on it?

You are talking about the project/product itself. Does your company sit down each quarter and crunch who specifically was responsible for the revenue and profits that quarter, down to the employee or lines of code? Doubtful.

Some things can't be measured.

Challenge accepted. Do you have an example?

Who exactly was responsible in a company for the company's profits going up by 20% in a quarter.

Huge grain of salt. More like a red flag. I don't want BS artists like that anywhere near me.

it’s _very_ easy to calculate your impact, or the one of your team, when you work on meaningful projects. if that’s not the case, you are probably working on non-essential things.

I think that's rather naive. A lot of work engineers do involves either keeping systems afloat or minor alterations that in cohort lead to a new feature but each minor alterations itself not doing anything meaningful. Evaluating individual or even team impact in those cases is _hard_. Yes, product people or in some cases managers do it but most of their numbers come with five asterisks to cover their asses as they are largely bs.

There are definitely people who are impactful enough and deserve to make bold "I" statements but they are in the vast minority.

Down to the lines of code written? Your product may get millions in revenue, but you wrote a single module out of hundreds, how much % do you take credit for?

Also it sounds like a toxic culture if you have to constantly try to tie your day-to-day work directly to how it impacts new revenue and profits or else your work is considered worthless.

If the takeaway is, "never work on internal tools," I've worked at companies where the internal tools were more valuable than anything customer-facing.

If you're great at optimizing SQL, probably just doing that in any place you can get your hands on will make a pretty big impact.

Overall impact is more helpful than resume points, IME, especially when references are involved.

If optimizing an internal tool can raise profits by 20%, I don't understand your disagreement. The comment you're replying to mentioned profits vs. unnoticed efficiency, not internal tools vs. consumer-facing software.

If it makes a big impact, it makes a big impact.

> If you're great at optimizing SQL, probably just doing that in any place you can get your hands on will make a pretty big impact.

This seems like an argument that there are no priorities and no things that are more important than other things. I don't think you believe that.

> This seems like an argument that there are no priorities and no things that are more important than other things. I don't think you believe that.

On the contrary, I'm arguing that making an impact wherever you can is better than avoiding things that don't meet some criteria of "things that are likely to help my resume." The priority I'm advocating is, making an impact by applying your skills (or even better, learning new skills) to solve problems that you can find (or are brought to you).

This is how you'll build a breadth of experience (which will be valuable for employers) and identify a niche that you enjoy (which will be valuable for you).

> If the takeaway is, "never work on internal tools," I've worked at companies where the internal tools were more valuable than anything customer-facing.

Yes, but they're exceptions.

> Overall impact is more helpful than resume points, IME, especially when references are involved.

The vast majority of the best paying jobs out there (FAANG), don't care at all about references.

It's a big world, it depends a lot in which circles you move around, I guess.

Disclaimer: I do work at a FAANG. I'm not sure what circles you're talking about.

"References" in the traditional sense may not be an important thing. But certainly being able to back your resume claims – whether with skills from that experience, or with people who will vouch for you – is important. If it wasn't, we could all just put "Optimized SQL and improved profit 20%" (or anything else) on our resumes, no?

I participate in interviews from time to time. I don't ask a lot of trivia-type algorithm questions, but I do try to size up where a candidate has invested their effort so far. Have they invested it in trying (and learning) to solve problems? Or have they invested in in padding their resume? I'll take the former over the latter every time, even if they don't have a bunch of brag-points on their resume.

That's just me, though shrug

If you “improve admin interface load times by 20%” do the calculation on what that is in saved time and possible money over 5 years. Then you have both.

You should of course have someone in management agree to your numbers and approve that you can use it on your CV.

Unfortunately saving time/money is not as sexy as increasing revenue.

I learned this once by creating a tool at a telecommunications company I was working at. It made it significantly faster for Technical Support to initiate a Remote Desktop session with a customer. We ran reports on how many RDP sessions we started per month; and recorded how much time the tool saved. At the time it was estimated to save about $2.1M in time savings every year, at a time when the Tech Support queues had long waits.

Couldn’t get any traction with management. Had some grassroots growth internally until we were asked to shut it down since it was an unofficial tool.

Meanwhile they would chase sales campaigns or marketing stunts that might result in $50-100k in new sign ups.

Nobody cares how much you saved in time because it doesn’t get reported anywhere. It shows up on no important metrics. Sales going up by $100K is sexy. Admins handling an extra ticket per day because the interface is quicker isn’t.

This is the kind of inefficiency that startups can use to get ahead of established competitors. So don't be sad, this is just a reminder to be an entrepreneur.

If just was as easy. Although I agree with your sentiment.

Absolutely!

> Nobody cares how much you saved in time because it doesn’t get reported anywhere. It shows up on no important metrics. Sales going up by $100K is sexy. Admins handling an extra ticket per day because the interface is quicker isn’t.

I don’t get that. Solving an extra ticket per day should somehow correlate with either revenue increase(indirect) or cost saving(direct), right?

Revenue increase; unlikely. Cost savings? Yes. But does it?

Managers want big teams. It makes them look good. Having 15 direct reports looks better than having 10. That means the manager needs his people to be efficient enough to look good on paper, but inefficient enough that he gets to hire more people. Having a bigger budget and higher headcount makes him look good.

yeah, I've read a lot about automation here where this was a common argument, also I noticed this in the company I work for as well. If you are in IT or the user yourself and automate something it doesn't really get recognized by management. I think the only time automation is recognized by management is when it's on the agenda of management itself, e.g. if it's said that 20% of the people have to go and for that to happen there has to happen automation (and outsourcing). However, in this case the idea comes from management so there would not be much recognition for the ones doing the automation (also with the reduction in workforce it wouldn't be appropriate for mere workers to ask for a bonus/raise! Of course that doesn't apply to management..). Automation really is a good example of how dysfunctional (or averse to change) big companies are.

That’s why I said profits not revenue. Specifically to capture that 1 & 2 might be the same work :)

Just talking in terms of business impact puts you miles ahead of other candidates in many eng jobs. Especially for levels above junior. Very important to have engineers who (want to) understand why they do what they do.

Meanwhile the most wildly successful companies in history did the opposite of that and hired techies to make amazing stuff that printed money.

> hired techies to make amazing stuff that printed money.

you are agreeing with me

Even better: I optimized SQL queries for Nameless Corp to improve admin interface load times by 20%, which boosted profits by 20%.

It's not that maintenance has no marketplace value, but that it's only really valuable when the system is on fire.

I've optimized that kind of SQL statement, and it paid really well, but only because it was at a point where it was causing a whole lot of pain, so the fix provided CEO-visible relief. What is not a great use of your time is to preemptively fix issues like this, unless it's somewhere so system critical that engineering effort is going to be rewarded regardless.

But even then, you might want to do something in a way that is easy to sell later in your career. You want to at least be able to say: Remember how <famous architect> sold his innovative system at <famous conference>, and HN wonder why this talk existed, as there's no way this solution could scale? Well, they were right, it didn't scale, and I was the one that replaced it with something good, piece by piece, without anything falling down.

And that's the reason you don't get a lot of high quality maintenance work happening at most places. The vast majority of systems in the world are just not worth the engineering effort to build very well. And even in companies that can afford to do this kind of work, you will see a lot of effort spent shuffling engineers across projects, as a team that has high visibility tasks in a quarter might not have any the next.

This is exactly what makes the difference between a seasoned and experienced engineer vs "follow the latest nodejs lib kind of guy" ...

I learned a lot exactly when doing non sexy stuff that nobody else wanted to do.

For instance, when it comes to optimizing an obscure query and make the whole system run twice as fast ... you will learn DB internals, and how the DB engine heuristic works when determining whether to use or not and index etc .. All things that are now part of my senior experience and knowledge and for which I am paid much more than a junior :)

My best advice would be : do the hard stuff, the stuff that people don't want to do. Make yourself valuable to the company. People will come and ask for you. That's the best way to learn and also have more leverage to ask for increase+stocks. Knowledge is power.

Everyone can pickup any new shiny framework in just a couple of days. That's not hard.

Some of the most satisfying work I've done has been optimizing very difficult SQL queries for clients nobody has heard of.

I guess I don't really care about optimizing my career in the way Patrick is suggesting.

Patrick uses his tech skills in service of marketing guru with a Personal Brand. What works for him isn't for everyone.

It's a false dichotomy. You can work on deep technical problems and provide lots of value.

Those examples I was giving did provide lots of value, to the client. The section from the article was talking about optimizing for visibility. Deep technical problems with lots of value are not always visible, and I think it's fine to build a career around those types of things.

I wasn't disputing what you said, kind of agreeing with it actually. Although as a consultant for over a decade myself I have to make an effort to make sure my work is visible. I'm not sure I could lurk in the shadows with no one seeing the value I create and still get paid :)

Do you actually get most of your work from promoting yourself to strangers via Github or similar?

In my experience, if you do good work, it's usually not that hard to get more work even without a public showcase of your work. Your clients will come back the next time they need to get a job done. Your clients also tell others about you, and word-of-mouth is extremely valuable advertisement.

I guess promoting yourself on Github and on personal blogs is one way to go about it, but real life personal relationships shouldn't be undervalued in my opinion. Most consultants I know have a surprisingly bare website.

Still don’t really understand what your objection here is. As the parent to your comment said: the OP never said don’t do it, they just said don’t stop once the code is checked in and follow up by pointing out the value to others incase it’s not obvious.

I think the question here is whether this value needs to be publicly visible.

a blog post about it might make a great complement

If you do unsexy work you will get paid unsexy money. If you have a passion for something thats great but if you want to earn a lot of money you will need to play the game.

This is probably the main struggle of my professional life: to see the difference and find the balance between doing a Good Job and doing a Sellable Job.

Creativity and perfection of craft versus efficiency and economy.

This balance is one of the most important things to get right when evaluating prospective job environments, especially when you're more on the thorough end of the spectrum.

I worked on highly visible internal work. Improving your colleagues' quality of life can be highly visible.

I think his point is that working on deep internals (rather than client facing products) is difficult to show off.

I think this is true, because you would have to do extra work beyond what your being paid to do, to make it show-off-able.

However, look at Brendan Gregg for example. He works on deep internals that would be very hard to show off without putting extra effort into sharing his knowledge. But since he does make that effort, he’s seen as an industry leader for his skill set.

I would suspect the company that designed the software for your local power plant is either a huge mega corporation or their name is on the front of the building or both. It’s not like their work is niche to the point of being esoteric. It’s just not FAANG. Also not targeted to consumers.

You’d be surprised how much industrial automation is not built by relatively small contractors. To your point, they tend to use “mega-corp” systems (e.g., Rockwell Automation hardware) but the actual logic is often written by small companies that bid out the project (or are subcontracted to for various reasons not necessarily related to technical acumen).

I haven’t personally worked o power plants but have been involved in various safety critical automation builds related to factories, hospitals, high energetic test stands etc. often, the code is maintained by a technician who isn’t either an engineer or classically trained computer scientist. Not to say those credentials are a must, but it makes more sense why the code I witnessed had so many bad practices in it

Reply on news.ycombinator.com