what is the best practice to upgrade the dependency v-calendar from beta to the latest version v2.3.0?
I know the question is opinion based, but I am new to the field, what are the pros and cons of the way to do the upgrade.
We use "v-calendar": "~1.0.0-beta.14" in production. The latest version is currently v.2.3.0
I can think of two ways:
update to the latest version with yarn add v-calendar#latest?
upgrade incrementally until the latest version is reached with yarn upgrade "v-calendar"?
What should I do to reach the latest version? thank you
ps. we are using vue: ~2.6.10
Unless the package you are using has an upgrade guide (99.99% of them don't), there is absolutely no point in doing an incremental upgrade. If it has an upgrade guide, read and follow its instructions.
The easiest way to upgrade is to change the package version to latest (in most IDE's if you press Ctrl/Cmd and hover the package number shown in package.json a tooltip will show you currently installed version, latest wanted version and latest available version). After you changed it (by typing the latest version in), run yarn install.
In the vast majority of cases, that's all you need to do, because most packages are built with backwards compatibility (existing features remain and new features are being added). If that's not true in your case, you'll have to reimplement it following their documentation. Typically it's no big deal.
Also, note there is no risk in attempting to change to a newer version. If things break, you just go back to package.json, change version back to lower, run yarn install again and everything is back to square one.
Related
In the given scenario, new Dev A has a new version of npm (8.3). He does a clone of the repo and npm says the package.lock file needs to be upgraded from lockFile version format 1 to version 2. Then he checks that in. So now the repo has a lock file with format version 2 and other devs pull that down.
Time goes by with no problems. Then Dev B, who's on version 6.13, installs a package. Is it normal that the lock file will change back to version one FROM version 2? In other words, each time a dev with a different version of npm that employs a different format, upgrades or installs a package, is it normal that it changes the lockfileVersion format over and over depending on their npm version and its format? Or should it stay with lockFileVersion 2?
Trying to identify what happened recently with our packages, and I'm looking to exclude this as a possible issue.
From what I read fileLockVersion 2 is supposed to be backwards compatible. But is it supposed to change back and forth like that? I wouldn't think so because how can someone with version 2 (Dev A) use version 1 if it was required to upgrade to version 2 in the first place?
So am I correct that once it's on version 2 it should stay that way? And if so, what would cause it to go back to version 1.
Thanks
No, it's not normal for developers to keep flipping versions like that, but it's fairly common for it to occur. We also had this happen specifically with package-lock.json so we told everyone to upgrade, and we made sure to upgrade all of our build agents. Similarly, in Visual Studio solutions some developers occasionally changed the version of VS back and forth from 2017 to 2019 a few times before we told everyone to just upgrade to 2019.
Generally it's best to avoid it if possible.
How does different mikro-orm packages versions correspond to each other? Could I possibly use #mikro-orm/core 4.0.0 and #mikro-orm/nestjs 3.1.0 or other variations of different #mikro-orm packages and their versions? Should all #mikro-orm packages be installed with same version?
It's 3 questions but I feel they are pretty much the same since they try to clarify concept of version matching in mikro-orm for me.
All packages that come from the main repository are released together, you should always align their versions.
https://github.com/mikro-orm/mikro-orm/tree/master/packages
The rest is not dependent on exact version, but in general, use the most up to date one. There are currently 3 packages that live outside of the main repo, the nest adapter you mentioned and 2 highlighters.
Btw there is no such release as #mikro-orm/nestjs 3.1.0, first release was 4.0.0 and current is 4.2.0.
edit: apparently lerna is aligning only minor/feature releases, so it can happen that some packages are not on the very latest patch version, example of that are mongo and reflection packages, that are still on 4.1.0, while the rest is on 4.1.1. but again, use the latest everywhere and you will be fine. Your IDE should guide you in knowing the latest versions.
I have a product and the version is 1.3.9. It has a severe bug when uninstall it. I fix the bug and make the product 1.4.0. Then I want to overwrite the 1.3.9 version whitout unintall it. Is there anyway to do that?
Minor Upgrade: Yes, you need a minor upgrade. It upgrades the product in-place and does not trigger uninstall. After that is done you can run the uninstall - now with a corrected uninstall sequence - and you can use major upgrades again the normal way - but you need to migrate yourself out of trouble with a minor upgrade. Minor upgrades have many limitations.
Note: it is also a problem that when your next version comes along you need to provide a minor upgrade again to migrate users who are still at the problem version. There are various approaches here - such as using the same minor upgrade wrapped in a setup.exe launcher.
Clunky Version: You can also make the new version ask for the user to manually uninstall the previous version. Not a good solution, but possible. Then you avoid the need to have to deliver a migratory minor upgrade as part of future launchers?
Maintanence of minor upgrades for broken major upgrade fixing (Michael Urman of Installshield)
Various Fixes: There are many hacks and tools you can use to clean this problem out, none are ideal (do visit that link).
We're using NodeJS for some projects and are faced with an issue that must have a simple solution (seeing as nobody else seems to have the problem).
In the packages.json there are a bunch of dependencies mentioned with a minimum version, each of which may have overlapping dependencies of their own. The default way a dependency is added is using the ^ operator which seems to mean 'compatible with' or 'same major version, but minor versions may differ'.
The way I understand npm to work is on npm install to take the highest minor version available that matches. Unfortunately 'compatible with' is not quite as enforced as you'd hope.
The situation this puts us in is that for instance on a developer machine version 1.1.0 is installed, but between development and publishing a new version 1.2.0, that has a bug, is introduced. On our build machine a fresh build is made which ends up using 1.2.0 and we've introduced a bug that wasn't there in development.
We tried changing the ^ operator to = for instance, but this gives us trouble when dependencies have subdependencies that aren't compatible with the requested version.
All in all I'm a bit confused, but this thing keeps biting us anytime something changes since the development machines don't do anything on npm install if the package is already there, but the build machine always gets fresh copies.
I know from NuGet that it always takes the lowest version that matches all combined requirements. Since this is always the same for a given set of dependencies, I much prefer this approach. Is there a way to make npm work like this too?
To answer my own question:
npm has introduced a new command npm ci which does something similar to npm install but enforces that the specific versions are used that were also used when a package was initially added by using the package-lock file.
See https://docs.npmjs.com/cli/ci for more information
I'm trying to create patches using the method from this tutorial. An issue I'm running into is that I can't install a new patch on top of a previous patch.
I can full install Version A,then patch to Version B. After that I can't patch to Version C.
I can full install Version B, then patch to Version C.
Currently we just do full installs with major updates each time which is working fine, but because of the frequency of our (internal) updates the file size and update time is becoming a burden so we're looking to reduce the update time (both downloading and installing) especially when most of the files don't change.
Edit: Another requirement is that at any given time a full install can be done instead of a patch. The solution I came up with setting a static product code made full installs on top (without manually uninstalling) doesn't work.
If you're not doing a major upgrade, but you are changing versions, you're doing a minor upgrade. To be able to install the next version .msi file over an existing installed previous version, you're going to have to set REINSTALL to a list of modified features somewhere (or to ALL if you're lazy and willing to put up with Windows Installer doing extra work). Often setting REINSTALL handled by the bootstrap, but it is possible to set it in the .msi and reset it to empty ({})when the previous versions are not installed (condition Not Installed).
Looks like the issue was that I was previously making all upgrades major upgrades, but that's not supported with patching. Changing to a static product code rather than auto-generate fixed it.
Edit:
Looks like it solved the first problem of Install A Patch B Patch C not working, but now trying to do a full install of D on top doesn't work.