Compare versions properly

[Trenly]

@OfficialEsco

• Resolves [New Feature]: Neutral/Better if newer version #117

[Trenly]

@OfficialEsco If you want me to rebase these on the latest to make testing easier LMK

[OfficialEsco]

Naah i cherry picked every branch into 1 on my repo ^^ Just doing testing under normal usage which is why some PR's gets approved slowly as they require a bit more stable testing

Edit: not sure how much you test your code, but if you do you'll probably pick a few that is mentioned, which doesn't always cover every corner (:

Edit2: I daily drive every patch, however i'm not sure what the state of #112 is, also its probably a bit too early to use since we don't have all the areas covered (MS Side)

[Trenly]

Naah i cherry picked every branch into 1 on my repo ^^ Just doing testing under normal usage which is why some PR's gets approved slowly as they require a bit more stable testing

Fair point; Just didn't know if it was easy for you

Edit: not sure how much you test your code, but if you do you'll probably pick a few that is mentioned, which doesn't always cover every corner (:

Depends on the case. If the issue is clear and easy to solve like the architecture bug, I do a little less testing. For more complex things, I do a little more testing.

Edit2: I daily drive every patch, however i'm not sure what the state of #112 is, also its probably a bit too early to use since we don't have all the areas covered (MS Side)

I'm thinking of waiting on #112 for a bit. I know there is a bug open for release date validation. I'm going to mark #112 as draft probably, just since it isn't critical to be merged in

[OfficialEsco]

Well, here is one issue microsoft#36724
Also looks like i missed this one microsoft#36633
Pretty sure this one also crashed microsoft#36642

Wait, are 2 of those intentional?

if ($AllVersions.Count -eq '1') { $CommitType = 'New package' }

[Trenly]

2 of those are intentional; Yes

Can you describe the issuse with each of those three packages in a bit more detail?

[OfficialEsco]

Nevermind, all of those are actually a trigger of 1 version in the repo

• Guilded: Dynamic url, AutoUpgrade
• SubSync: there was 1 version in there, Simple Update
• Sandboxie.Plus: I submitted 0.9.8d first, dragged those files into the folder and SimpleUpgraded to 1.0.0.0.0 so i expected New version: instead of New package:

The .Count line should probably be last in the elseif's or rewritten so it dosn't trigger when there already is a version in there. (Because i regularly update Dynamic urls and it looks weird when it always says New package:)

[Trenly]

Nevermind, all of those are actually a trigger of 1 version in the repo

• Guilded: Dynamic url, AutoUpgrade
• SubSync: there was 1 version in there, Simple Update
• Sandboxie.Plus: I submitted 0.9.8d first, dragged those files into the folder and SimpleUpgraded to 1.0.0.0.0 so i expected New version: instead of New package:

The .Count line should probably be last in the elseif's or rewritten so it dosn't trigger when there already is a version in there. (Because i regularly update Dynamic urls and it looks weird when it always says New package:)

Ahh, fair point. I’ll take a look at those cases and see where the logic needs to change

[Trenly]

• Guilded: Dynamic url, AutoUpgrade

I believe this is fixed

• SubSync: there was 1 version in there, Simple Update

I believe this is fixed

• Sandboxie.Plus: I submitted 0.9.8d first, dragged those files into the folder and SimpleUpgraded to 1.0.0.0.0 so i expected New version: instead of New package:

Not entirely sure what happened here. I wasn't able to replicate. If it is still happening, maybe we can chat on Gitter to figure it out?

The .Count line should probably be last in the elseif's or rewritten so it dosn't trigger when there already is a version in there. (Because i regularly update Dynamic urls and it looks weird when it always says New package:)

This actually was the root cause of both issues that were resolved. If there was only one version of the package, it was automatically casting the variables as strings. I explicitly cast them as arrays, which should have fixed the issue

[OfficialEsco]

I've not had too many weird packages, but as far as i can tell this is working as expected