michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-822196779
>
>
> I spotted this (refCount was not final), but will run your test tomorrow
and try to reproduce.
Can you explain why `final` should fix this?
--
This i
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-822059795
@cstamas Ran the first tests with:
```
for run in {1..5}; do echo "Testing run #$run"; rm -rf repo;
~/apache-maven-4.0.0-alpha-1-SNAPSHOT/bin/mvn clean install -Prun-i
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-822004920
Working on the PR now...
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-818908950
>
>
> K, merged the "mini PR", I hope this addresses all your concerns.
I guess so. I will test with the few projects I know to fail. You can
already squash a
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-818710284
>
>
> > I understand this now, but am not happy with for the following reasons:
We have a hard requirement on reentrancy, it is up to the lock to implement
that pro
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-818679028
>
>
> > There is one more thing I do not understand and think that this
undermines reentrancy: While I understand that you maintain a concurrent map in
`NamedLockFa
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-818655591
>
>
> @michael-o just my 5 cents: "long running PRs" (this one is slowly 1/2
year old 😄 ) suffer from very same issues as "long running feature branches":
is easy t
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-818012784
>
>
> re formatting: the maven site exposed formatting rule for Idea IDE drives
me nuts, it is not aligned with checkstyle, so several "reformat" attempts,
sorry fo
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-817388228
> yes but the problem is that the lock is at the same level as the
collection. It should be at the level of individual file and the JVM together
with OS would lock it by it
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-817384481
We need one lock per artifact to coordinate access to files.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHu
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-817376215
@Tibor17 If I understand you correctly you want to obtain locks
concurrently. If yes, this will be a source of deadlocks because they have to
be acquired in serial from wi
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-817367241
There is one more thing I do not understand and think that this undermines
reentrancy: While I understand that you maintain a concurrent map in
`NamedLockFactorySupport` to
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-813447624
Question: Does it make sense to add an enum to `NameMappers` which denotes
whether they are local (in-VM) or distributed (multi-VM) and appropriate impls
you reject mappes
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-813444925
>
>
> > A left a few comments, please go through.
> >
> > * I have a few more homework to do because I do not fully understand all
involved layers with `Adapt
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-813441893
>
>
> General remark: my personal preference for local variables is to declare
them as they are, so `Type variable = new Type()`, in cases when the variable
is stri
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-811833880
>
>
> @hboutemy
> @cstamas
> @michael-o
> Some long time ago I was talking about the following isea in our Slack.
> It was about atomic move for the artif
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-811828377
Let me go through this today/tomorrow. Thanks for the elaborate response!
--
This is an automated message from the Apache Git Service.
To respond to the message, please lo
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-808807986
I forgot to mention. I do really like the idea of the name mappers. I have
used a similar approach in a completely different area years ago, but that code
serves me daily t
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-802992676
Scheduled for this weekend...
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go t
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-784020330
Thanks for the ping. Will pick up again during this week.
This is an automated message from the Apache Git
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-764929160
I need to go through another iteration because I wasn't really happy with
some things last time, plus I need test myself too. I am currently on a short
leash, but will try
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-764929160
I need to go through another iteration because I wasn't really happy with
some things last time, plus I need test myself too. I am currently on a short
leash, but will try
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-754170465
Will pick up by Wednesday.
This is an automated message from the Apache Git Service.
To respond to the mess
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-751690622
@rmannibucau I would not exclude them from the delivery, but mark as
experimental. I want people to test them. If we don't make it publically
available it won't be usable f
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-751685099
As for testing Redis, I would not rely on tools like Docker because it is
too OS specific. One could probe whether the redis daemon is in `PATH`. If yes,
start it. Run test
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-751682400
@rmannibucau
1) This makes sense, but I don't want to create a mapping from properties to
the object ob Redisson. The YAML configuration is so vast that I want to leave
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-751531066
Very nice, I will happily go through.
This is an automated message from the Apache Git Service.
To respond
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-731603812
So no code changes are required to do a single node setup like with Redis?
This is an automated message fro
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-731601588
When I read your note about HZ, does it make sense at all when it requires
at least three nodes to work? This seems like overkill. I was able to achieve
very decent peforma
michael-o commented on pull request #77:
URL: https://github.com/apache/maven-resolver/pull/77#issuecomment-731601299
@cstamas Do you want to perform a final review yourself before I go over
again?
This is an automated messa
30 matches
Mail list logo
