May 2021 Speculations

General discussion about calculators, SwissMicros or otherwise
ljubo
Posts: 70
Joined: Sat Jun 20, 2020 1:46 pm
Location: Austria

Re: May 2021 Speculations

Post by ljubo »

Would it make sense to go open source? - That could accelerate development, improve quality and also increase sales.
HP-15C, DM15L, HP-35S, DM42
User avatar
Walter
Posts: 3070
Joined: Tue May 02, 2017 11:13 am
Location: On a mission close to DRS, Germany

Re: May 2021 Speculations

Post by Walter »

ljubo wrote:
Thu Sep 22, 2022 8:57 am
Would it make sense to go open source? - That could accelerate development, improve quality and also increase sales.
... but it would reduce SM's profit so I'm uncertain if they will happily support this way. And I doubt the acceleration as well - just my personal opinion, of course, based on a little experience... 8-)
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
gmac42
Posts: 103
Joined: Fri Jun 01, 2018 11:30 am

Re: May 2021 Speculations

Post by gmac42 »

Walter wrote:
Thu Sep 22, 2022 10:30 am
ljubo wrote:
Thu Sep 22, 2022 8:57 am
Would it make sense to go open source? - That could accelerate development, improve quality and also increase sales.
... but it would reduce SM's profit [...]
How so?
DM41X #542, DM42 #650, DM41L #801, HP 41CX, HP 41CV, HP 50G, HP11C, TI 89
redglyph
Posts: 177
Joined: Sat Dec 22, 2018 11:45 am

Re: May 2021 Speculations

Post by redglyph »

I think it depends on the market size, and it doesn't look very large these days.

Even if it were larger, it would likely make more sense to increase the team size rather than relying on open-source where there's little control, predictability and way to direct the development. I'm also very sceptical about any acceleration and quality improvement from a move to OS at this scale. It's just an intuition, I may be wrong.
vaklaff
Posts: 28
Joined: Thu Nov 12, 2020 9:23 am
Location: Czechia

Re: May 2021 Speculations

Post by vaklaff »

redglyph wrote:
Thu Sep 22, 2022 12:49 pm

open-source where there's little control, predictability and way to direct the development
Opensourcing may or may not be beneficial to SM and I’m not in a position to advice them. I’m jumping here because of the argument above, it doesn’t make sense to me. Can you elaborate, possibly give an example?
I worked with and on opensource, even professionally, and I don’t see how the above applies. However it’s not the first time I hear that so maybe I may be missing something obvious…
redglyph
Posts: 177
Joined: Sat Dec 22, 2018 11:45 am

Re: May 2021 Speculations

Post by redglyph »

vaklaff wrote:
Thu Sep 22, 2022 2:52 pm
Opensourcing may or may not be beneficial to SM and I’m not in a position to advice them. I’m jumping here because of the argument above, it doesn’t make sense to me. Can you elaborate, possibly give an example?
I worked with and on opensource, even professionally, and I don’t see how the above applies. However it’s not the first time I hear that so maybe I may be missing something obvious…
As I mentioned it's only an intuition, because I have managed several big projects professionally but no OS projects.

With a hired team, you know the resources at your disposal. You reliably know who they are, what they can do, when they will work on the project. If you rely on OS, you have no idea who will participate, what their skills are, and when they'll be able and willing to work on the project.

I may be wrong, but if I had to manage such a project and stick to a quality/deadline target, I would be very uncomfortable. And I'm not even talking about team dynamics and how to make it evolve for the next project.
User avatar
OlidaBel
Posts: 58
Joined: Thu Mar 11, 2021 8:52 am
Location: Belgium

Re: May 2021 Speculations

Post by OlidaBel »

Boub65 wrote:
Sun Sep 18, 2022 6:03 pm
IMG-20220918-WA0003.jpg

(Screnshot taken from Logan West youtube video)
We can see the DMCP SETUP called SET on shift ON... plus of course the extra upper FN row of 6 keys.
I don't need it, but it has a beautiful look, 2 shift colors. I like it. :lol:
---
Olivier
48GX, Prime G2, 50G, DM15L, DM42, 28S, HP 15c CE
vaklaff
Posts: 28
Joined: Thu Nov 12, 2020 9:23 am
Location: Czechia

Re: May 2021 Speculations

Post by vaklaff »

redglyph wrote:
Thu Sep 22, 2022 3:06 pm

With a hired team, you know the resources at your disposal. You reliably know who they are, what they can do, when they will work on the project. If you rely on OS, you have no idea who will participate, what their skills are, and when they'll be able and willing to work on the project.
Thank you for taking the time, appreciated! I see your point. I’m so used to how OS works that seeing outside contributors as a resource (of course, an unmanageable one) didn’t come to my mind at all :-) Yes, if you need something done in a given time, you are on your own.

Opensourcing can be beneficial, you can get useful contributions - but in essence “asynchronously”. And what’s the price? I typically make public code “prettier” than closed source (i.e. more time needed) and one also must have time and communication skills to talk to the contributors (if there are any, that is :-). I think this is the core dilemma SM would face in the hypothetical case they start to think about opensourcing the DM32 SW - time dedicated to cultivate a community versus potential async contributions.
ljubo
Posts: 70
Joined: Sat Jun 20, 2020 1:46 pm
Location: Austria

Re: May 2021 Speculations

Post by ljubo »

redglyph wrote:
Thu Sep 22, 2022 3:06 pm

As I mentioned it's only an intuition, because I have managed several big projects professionally but no OS projects.

With a hired team, you know the resources at your disposal. You reliably know who they are, what they can do, when they will work on the project. If you rely on OS, you have no idea who will participate, what their skills are, and when they'll be able and willing to work on the project.

I may be wrong, but if I had to manage such a project and stick to a quality/deadline target, I would be very uncomfortable. And I'm not even talking about team dynamics and how to make it evolve for the next project.
Well, it hasn’t to be “either-or”. The internal professional team could host software on e.g. github and develop there. Two things could happen:
  • There is no interest - well, then you have at least a free development infrastructure :D
  • People are forking project and submiting patches - the professional team could cherry pick and merge good contributions into the official project
HP-15C, DM15L, HP-35S, DM42
redglyph
Posts: 177
Joined: Sat Dec 22, 2018 11:45 am

Re: May 2021 Speculations

Post by redglyph »

ljubo wrote:
Thu Sep 22, 2022 7:55 pm
Well, it hasn’t to be “either-or”. The internal professional team could host software on e.g. github and develop there. Two things could happen:
  • There is no interest - well, then you have at least a free development infrastructure :D
  • People are forking project and submiting patches - the professional team could cherry pick and merge good contributions into the official project
I don't see the interest, except if there's a will to make it an open-source project.

As a company, I would never host a project outside, even if the chance that the external service disappears is minimal. There's the privacy and security issues, then I think that GitHub has limitation regarding the capabilities unless you upgrade your account. What we did is install our own GitLab server internally, so we could do anything we wanted with it. :)

If it had to be an open-source, I would release that once the development is finished.

It would probably take as much time for a team to start reviewing what has been submitted by unknown people as doing the tasks themselves. And again, by default there is no control on what is done or when.

When you develop a project in a team, you often discuss with the others to see what next task should be selected because it makes the work easier for the others, or maybe there's a specific knowledge that makes it more efficient. If instead you have to synchronize with outsiders, it's opening a whole can of worms. You need to use a public tool to manage the stories and tasks, hope that when a task is selected by an outsider, it won't take 5 times the expected time or won't just be abandoned, and so on. And they'll find it strange not to be paid to do the same work. What happens 6 months later when a bug is found in someone's code (which is likely less in-line with internal policies and probably hard to read, with no notes whatsoever) but that someone is not there anymore? Without the contractual ingredient, I think it's just not worth the trouble.

EDIT: Obviously I'm only giving my own opinion, besides I haven't tried in that context - I did try with "external resources" who were only loosely bound to the project, so it was similar, but it never turned out well. Now, I have no idea what SM would think about such an arrangement; I don't know them and I can't speak for them.
Post Reply