Selective Install-Options

[warm monolith]

Assuming, I want a lightweight Medicat Stick without some PortableApp or images, I think it would be cool to select the Apps/images/funktions you need before installing Medicat.

[ornate escarp]

kinda impossible coz the installation zip is a highly compressed archive, so we could make different zips instead for different user needs

[tame kindle]

It seems quite possible from my testing. You can provide a list file to 7z exe during extraction to specify the files / folders you want to extract

I wouldn't touch this until medicat installer is moved to a more useful language. Batch is some hot garbage to work with

[ornate escarp]

making a gui c# installer bruh

[tame kindle]

What UI framework you going with?

[ornate escarp]

just basic win forms

[tame kindle]

Actual L

[ornate escarp]

idfc

it's to install a toolkit

not to sell gucci things

[tame kindle]

Is winforms cross platform? I haven't used it for like 8 years

[ornate escarp]

idk

[tame kindle]

Anyway, point is you could totally add in the selection based install. Also batch sucks ||I'm sure we can agree on that much||

[ornate escarp]

alr

[tame kindle]

If you ever open source the installer project, I could look into it for you. Feature creep is a bitch

[ornate escarp]

i have to find time to finish it before

[honest bison]

Not particularly. This is something I had advised jayro years back when they transitioned from mini windows to usb but it was totally shut down, probably either due to arrogance or lack of understanding.

Instead of hosting a single compressed archive, you could have multiple archives for different things.

Separate shorter archives for common things. Linux distros could go in one. Tools in another. This or that in another. Mini win and portable apps in another. Extra portable apps separately. People could pick which one to download and use instead of downloading full heavy files.

And also makes it easier on the developer/uploader. Upload files that were changed instead of having to repeatedly prepare a full image and upload that from scratch.

Unnecessary implementation of medicat usb so far. Having to download 20gigs just because someone wants a particular portion of it is kinda cumbersome. Maybe they can afford all the upload and disk space. Not everyone can. Gladly it's 20 gigs now. When it first started you hadto download a full 20+ gb, extract that which would end up as a 64 gb file or something. Early iterations of usb were just that bad.

[honest bison]

You could also just host smaller archives and have the installer download necessary ones. Something to consider. I haven't touched medicat since the whole transition to usb. Thought to chec it out but still, nope.

If you're looking for examples, then the snappy driver installer is already a good example.

It has options for a full download as well as smaller downloads too. That's what you call being efficient

[honest bison]

Even with batch they only needed to download the ventoy portion and image separately. Imho, instead just have created a script that'd allow users to create a ventoy based usb, ask users how big they'd want the image size to be. And after creating that then download the necessary files that the user could select.

[tame kindle]

I honesty think using powershell bundled with spectre console would be easy to maintain and setup. Alternatively, you could roll your own TUI library. Keeps the initial download to a few Mbs at most. This doesn't need to be some super crazy thing for what it is. I like the split downloads idea, but the maintenance could be easier / harder depending on how it is being automated (if it is being automated?) at the moment.

at the end of the day, I'm not going to jump in anytime soon. I have enough projects I need to not develop for (because lazy). I don't need to add this one to my list