RELOAD!

[Minesweeper] Games Played: 122, Bombs Used: 63, Moves Performed: 14603, New Users: 15

At work, we just started using SVN.

@M.Doerner my condolences and congratulations at the same time?

Kind of. We had scc at all in the department. I convinced my department boss and some others that it would be a good idea. However, since our sister department in our parent company currently uses SVN and not git, they decided that we would use SVN as well instead of git.

> mumble mumble this wasn’t the SCC mumble I wanted mumble.

My only experience is with git. Is SVN simply older and not as well designed?

IINM, SVN was what motivated Linus to come up with git

Brief read also says SVN is single source of failure. It goes down you be SOL. I it also network access dependent, no local files.

yeah, all SCC back then were server-client

and I think (but am not sure) that SVN is also using a locking model (whereas git uses branching model)

> It has always been done this way. If it’s not broken, don’t try and fix it.

I understand how someone could have gotten frustrated to the point of “This blows donkey balls. We need a better system. This one screws the pooch horrribly.”

yeah In Linus' case the Linux project probably was much more loosely organized and as such they didn't want to have a single point of failure

server-client locking model is probably OK for corporate development but not so OK for a bunch of cowboys rustling some tumbleweeds

TFW you encounter an intermittent AV, develop a semi-reliable way to reproduce the AV, then attempt to fix it, undo the fix to prove that the fix really did prevent the AV from occurring and failing to get the AV again after several attempts.

@this locking vs. branching really isn't the point there

@this the problem for linux was less the single point of failure and more the issue of opening access to patch-based external contributions

@this lol

@Vogel612 ah ok. To me that's the biggest difference to grasp when transitoning

whereas git just allows you to make conflicting changes at will until you merge

the core difference really is that git has no single source of truth

everything else, even branches must follow from that idea, so long as you want a shareable history

but tbh when people talk about git being "a decentralized SCC", I find that more distracting & less helpful

because transitoning from the locking model, I was "where's my checkout?!?"

So understanding how branching & merging works was more important in adopting git.

yea, but it's the actual core of the difference

locking makes no semantic sense if your localhost is the equivalent of the SVN remote server

that's why people talk about git being decentralized when the locking model is brought up

but that's a really interesting point.

my current workplace switched from SVN to git on monday

I'm not directly on the dev-team, but there is a few older folks on the team that possibly have the same problem that you describe

yeah, it didn't really click until I saw a tutorial/movie where the focus was on branching & merging

the "push" and "pull" doesn't really make much sense until that.

probably because the first intuition is to pretend that "pull" is "checkout" and "push" is "check-in" but that's a poor analogy

yea

I wonder why Visual Studio seem to think that RubberduckParserState isn't handling the OperationCanceledException exception.

@this TFW that the solution involves calling a Placeholder3() and you have no idea what it does.

[rubberduck-vba/Rubberduck]: 1269 stars vs. [decalage2/oletools]: 1459 stars