> Keivan should probably be a (multi) millionaire now, sitting comfortably with his family doing whatever it is they enjoy doing. Instead, someone at Microsoft used this as a way to get themselves a promotion.
To be fair, AppGet was never really that popular, and Microsoft's replacement is even less so. I'd be surprised if anyone got a promotion based on their WinGet work. There has been no significant adoption of WinGet in the market. They're lucky Microsoft doesn't dissolve the team. We use Chocolatey instead, which I think is more popular (but still not very popular).
I have no horse in the race, nor do I know anything special about this situation, but it's fairly clear to me that the author failed the interview. I don't know if he understands that and is trying to downplay it, or if he doesn't realize that's what happened. That's why everything dried up after the interview, and why the followup says the position "didn't work out."
> I'd be surprised if anyone got a promotion based on their WinGet work. There has been no significant adoption of WinGet in the market.
Because it's an terrible implementation of a package management. Don't get me wrong, it's better than nothing but still...
For instance:
- we can't pin packages (there's an open issue for God knows how long) Problem with that is that if you do "winget upgrade --all" it will upgrade all your programs even those that you don't want to upgrade. The alternative is simply upgrading one by one.
- Some programs have a problem with their version (GOG Galaxy for instance). Winget simply doesn’t know that there’s no newer version available and keep trying to upgrade everytime you do “winget upgrade –all”. This, in turn, make the previous point more annoying as I can’t say to winget ignore “GOG Galaxy” and upgrade all the other packages.
- Some programs are downloaded from websites that are probably heavily throttling downloads from winget, this make downloading qbittorrent (20mb) takes a LONG time and usually failing.
I have no idea of how to fix these issues as I'm just a user but there's a stark difference between using windows winget and apt or pacman.
> Some programs have a problem with their version (GOG Galaxy for instance). Winget simply doesn’t know that there’s no newer version available and keep trying to upgrade everytime you do “winget upgrade –all”. This, in turn, make the previous point more annoying as I can’t say to winget ignore “GOG Galaxy” and upgrade all the other packages.
This was resolved in the latest preview release (well, the PR merged a couple months ago, but they seem to release once every million years). If you want to upgrade to preview temporarily it will still automatically update to stable releases.
> Problem with that is that if you do "winget upgrade --all" it will upgrade all your programs even those that you don't want to upgrade. The alternative is simply upgrading one by one.
Well, TBF, that's exactly what "winget upgrade --all" sounds like it should do.
Every Linux package manager that I use have an option to hold one or more packages from upgrading. It’s common sense that one or more piece of software might break on update and the user should be able to hold it until a fix arrive. Even arch where it’s heavily discouraged to do partial upgrades have an option to pin packages.
What winget upgrade —-all *should* do is to follow this common sense and do a upgrade all /minus package that the user choose not to upgrade.
Admittedly I don't have any experience with this sort of thing, but I have always read it as; Microsoft want his work, they should have just paid him X million for it, or paid him to keep working on it.
Instead, they made it into some sort of interview, got all of his input then went off and copy-paste'd the code (an exaggeration) into WinGet and ghosted him.
To add insult to injury he got a nice little "fuck you" the day before they released it.
winget is protected as it's providing the API for the Microsoft Store to manage Win32 apps. I think there's been a decent amount of adoption, there's lots of install pages (Git is one I noticed recently) that have winget install right next to brew install.
They also don't really promote it as far as I can tell. I didn't find out about it until I was looking up some stuff about choco. It's terrible what they did to this guy. I know he shared his stuff as open source but still to string him along for over a year whilst also having a team working on what he had built and barely acknowledging it or helping him out in some way.
> it's fairly clear to me that the author failed the interview.
What's important here is how he "failed the interview": He was probably never going to pass it. They'd already "failed" him beforehand for being an Open Source freak; the whole charade was just to string him along.
Microsoft employs many people that are open-source advocates. Someone's perspective on that would not impact their status in an interview.
If anything, the reason feedback took forever to him was because someone may have asked for an exception to the interview failure so they could hire him. Large companies are behemoths and things do go wrong, mistakes are made, and Microsoft did own up to it in this case (https://devblogs.microsoft.com/commandline/winget-install-le...) but open source did not play a part in it.
Source: myself, I worked at Microsoft in a past life and did many interviews. No one cares about someone's stance on open source anymore.
> it's fairly clear to me that the author failed the interview
I’m not so sure. The interview could have been a pretense to get more information, e.g. strategy for future development.
Also, in a sufficiently machiavellian environment, candidates can easily “fail” interviews by just seeming smarter than the hiring manager, scaring that manager into thinking they’ll be leapfrogged.
To be fair, AppGet was never really that popular, and Microsoft's replacement is even less so. I'd be surprised if anyone got a promotion based on their WinGet work. There has been no significant adoption of WinGet in the market. They're lucky Microsoft doesn't dissolve the team. We use Chocolatey instead, which I think is more popular (but still not very popular).
I have no horse in the race, nor do I know anything special about this situation, but it's fairly clear to me that the author failed the interview. I don't know if he understands that and is trying to downplay it, or if he doesn't realize that's what happened. That's why everything dried up after the interview, and why the followup says the position "didn't work out."