4.5.1.3538

[SilverAzide]

P.S.: ModernGadgets Disks Meter works with 4.5.0, crashes on all 4.5.1 releases (including prereleases).

[jsmorley]

P.S.: ModernGadgets Disks Meter works with 4.5.0, crashes on all 4.5.1 releases (including prereleases).

Yes, I have the same results. We are looking into it.

[SilverAzide]

Yes, I have the same results. We are looking into it.

You guys are probably way ahead of me, but I think I found the bug. This skin has a single SysInfo measure:

Code: Select all

[MeasureUserName]
Measure=Plugin
Plugin=SysInfo
SysInfoType=USER_NAME
IfCondition=1
IfTrueAction=[!CommandMeasure MeasureGetLocalGroup "Run"]

This causes a crash for some bizarre reason. Changing this to an identically functioning alternative works fine:

Code: Select all

[MeasureUserName]
Measure=String
String=%USERNAME%
IfCondition=1
IfTrueAction=[!CommandMeasure MeasureGetLocalGroup "Run"]

[jsmorley]

You guys are probably way ahead of me, but I think I found the bug. This skin has a single SysInfo measure:

Code: Select all

[MeasureUserName]
Measure=Plugin
Plugin=SysInfo
SysInfoType=USER_NAME
IfCondition=1
IfTrueAction=[!CommandMeasure MeasureGetLocalGroup "Run"]

This causes a crash for some bizarre reason. Changing this to an identically functioning alternative works fine:

Code: Select all

[MeasureUserName]
Measure=String
String=%USERNAME%
IfCondition=1
IfTrueAction=[!CommandMeasure MeasureGetLocalGroup "Run"]

That's super puzzling. I can put that exact measure into a new skin, and it works fine. And yet in the context of that DisksMeter skin it doesn't. The only thing that jumps out at me is that DisksMeter skin is encoded as ANSI instead of UTF-16, but I don't see how that should matter. Brian will need to weigh in.

[Brian]

Although it may seem weird, this is actually caused by a bug in section variables. Specifically the EscapeRegExp and EncodeUrl functions.

Here is another example of the crash:

Code: Select all

[FreeDiskSpace]
Measure=FreeDiskSpace
Type=0
Label=0

[OOPS]
Measure=String
String=[FreeDiskSpace:EscapeRegExp]
DynamicVariables=1

Note: The OOPS measure doesn't really make sense, but it does crash prior to Rainmeter 4.5.2

We will get a fix out ASAP.

Thanks to Jeff and SilverAzide for all your help!

-Brian

[Brian]

Fixed for 4.5.2

https://forum.rainmeter.net/viewtopic.php?f=13&t=38392#p197456

-Brian

[Jeff]

What version of Windows are you using?

-Brian

Win10 Home, v21H1, build 19043.1165
Windows Feature Experience Pack 120.2212.3530.0

EDIT: I just woke up 2 mins after the fix was posted, I'll still leave the diagnostics here, thank you!

[Brian]

Win10 Home, v21H1, build 19043.1165
Windows Feature Experience Pack 120.2212.3530.0

EDIT: I just woke up 2 mins after the fix was posted, I'll still leave the diagnostics here, thank you!

Thanks!

-Brian

[death.crafter]

Fixed for 4.5.2

https://forum.rainmeter.net/viewtopic.php?f=13&t=38392#p197456

-Brian

Hmm...

I have a suggestion about pre-releases.

I know it's irrelevant but still:

• We should get update notifications for pre-releases too.
• Pre-releases should be ahead of final releases, so that if we are on a pre-release, we wouldn't miss the updates. (I know if we are pre-release testers, we are supposed to keep coming to the forums for testing and stuff)
• Keep the previous system of major.submajor.minor.revision (ofc, the terms are not correct) when releasing final versions. I know it sounds a bit of a cliche, but it would help keeping the pre-releases ahead or equal of finals.
  That way, betas, or pre-releases would work like they are supposed to. When you feel like it's a big enough change to bump the minor version, bump first in pre-release and then finalize it, unlike you bumped to final 4.5.2 from pre-release 4.5.1.3538.

[jsmorley]

Hmm...

I have a suggestion about pre-releases.

I know it's irrelevant but still:

• We should get update notifications for pre-releases too.
• Pre-releases should be ahead of final releases, so that if we are on a pre-release, we wouldn't miss the updates. (I know if we are pre-release testers, we are supposed to keep coming to the forums for testing and stuff)
• Keep the previous system of major.submajor.minor.revision (ofc, the terms are not correct) when releasing final versions. I know it sounds a bit of a cliche, but it would help keeping the pre-releases ahead or equal of finals.
  That way, betas, or pre-releases would work like they are supposed to. When you feel like it's a big enough change to bump the minor version, bump first in pre-release and then finalize it, unlike you bumped to final 4.5.2 from pre-release 4.5.1.3538.

Agreed. This was my bad. What we will go going forward is have say release 4.5.2.3600, then prerelease 4.5.2.3700, which will end up as release 4.5.3.3700. In any case, yes, prerelease versions should prompt for upgrade when the release version is out.