Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ddmd -> dmd transition #7135

Merged
merged 3 commits into from
Dec 15, 2017
Merged

ddmd -> dmd transition #7135

merged 3 commits into from
Dec 15, 2017

Conversation

@dlang-bot
Copy link
Contributor

Thanks for your pull request, @wilzbach!

Bugzilla references

Your PR doesn't reference any Bugzilla issue.

If your PR contains non-trivial changes, please reference a Bugzilla issue or create a manual changelog.

@dlang-bot dlang-bot added the Review:WIP Work In Progress - not ready for review or pulling label Sep 14, 2017
@JinShil
Copy link
Contributor

JinShil commented Sep 14, 2017

If this is to investigate better naming for the compiler-as-a-library API, would it be better to just use the name d in an effort to make it compiler-agnostic?

@wilzbach
Copy link
Member Author

If this is to investigate better naming for the compiler-as-a-library API, would it be better to just use the name d in an effort to make it compiler-agnostic?

Yes and no. tl;dr: it's picking up an effort to clean up the folder hierarchy started by @RazvanN7 and @andralex half a year ago (ddmd comes back from the old days when it meant "D version of dmd").
See also:

#6489
#6562
#6595

@JinShil
Copy link
Contributor

JinShil commented Sep 14, 2017

(ddmd comes back from the old days when it meant "D version of dmd").

Yes, I'm aware of that, but I was questioning the use of dmd as a package name as it is specific to only one brand of D compiler. My suggestion was to use the name d instead.

@wilzbach
Copy link
Member Author

Yes, I'm aware of that, but I was questioning the use of dmd as a package name as it is specific to only one brand of D compiler. My suggestion was to use the name d instead.

While I appreciate your input, there have been many lengthy discussions about this and a decision has been made to rename to "dmd":

#6489 (comment)
#6489 (comment)
#6489 (comment)

Unfortunately the rename was left in stage 2 and I am just trying to take it to the finish line, which sadly is though enough. Also, there are (and hopefully will be) other D compilers which aren't based on the DMD frontend (e.g. SDC), so taking the d namespace would be quite unfriendly (especially if lazy people used d to organize their private packages).

@CyberShadow regarding:

.generated/stable_dmd-2.072.2/dmd2/linux/bin64/rdmd --compiler=.generated/stable_dmd-2.072.2/dmd2/linux/bin64/dmd -conf=.generated/stable_dmd-2.072.2/dmd2/linux/bin64/dmd.conf modlist.d ../druntime ../phobos ../dmd --ex=gc. --ex=rt. --ex=core.internal. --ex=core.stdc.config --ex=core.sys. --ex=std.algorithm.internal --ex=std.c. --ex=std.concurrencybase --ex=std.internal. --ex=std.regex.internal. --ex=std.windows.iunknown --ex=std.windows.registry --ex=etc.linux.memoryerror --ex=std.experimental.ndslice.internal --ex=std.stdiobase --ex=std.typetuple --ex=tk. --ex=msvc_dmc --ex=msvc_lib --ex=ddmd.libmach --ex=ddmd.libmscoff --ex=ddmd.objc_glue --ex=ddmd.scanmach --ex=ddmd.scanmscoff \
	--dump object --dump std --dump etc --dump core --dump ddmd >.generated/modlist-prerelease.ddoc
core.exception.AssertError@modlist.d(75): ddmd can't be found.
----------------

(this is about the menu generation for the dlang.org (Ddoc version))
Q: Would you be okay with making this error soft-failing for the time being?

@CyberShadow
Copy link
Member

Q: Would you be okay with making this error soft-failing for the time being?

You're asking the wrong person, DMD docs on dlang.org are mostly pushed by W/A. But I suggest finding a way to do things that doesn't imply that "just break it for now, we'll fix it later", as "later" may be postponed more than expected.

Why not detect the name of the DMD package/directory?

@ibuclaw
Copy link
Member

ibuclaw commented Sep 15, 2017

I'd still prefer some agnostic name as per comment I made many months ago in linked pr above.

@andralex
Copy link
Member

I'd still prefer some agnostic name as per comment I made many months ago in linked pr above.

It's a nice point but gets us into rebranding the front end. I'd say we move forward with this obvious step and defer that.

@andralex
Copy link
Member

Let's do this already. @MartinNowak ? @WalterBright ?

Copy link
Member

@MartinNowak MartinNowak left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Anytime

@ibuclaw
Copy link
Member

ibuclaw commented Sep 23, 2017

Just fix the build.

@wilzbach
Copy link
Member Author

wilzbach commented Oct 5, 2017

Seems like we need to update the auto-tester again: https://auto-tester.puremagic.com/show-run.ghtml?projectid=1&runid=2816476&dataid=19934733&isPull=true :/

-> braddr/at-client#4

At least CircleCi and Jenkins have been green.

@andralex
Copy link
Member

andralex commented Oct 5, 2017

rebase?

@andralex
Copy link
Member

andralex commented Oct 5, 2017

Let's boldly go see regressions we haven't seen before :).

@wilzbach
Copy link
Member Author

wilzbach commented Oct 5, 2017

rebase?

I really appreciate the enthusiasm, but we have to wait at least for dlang/dlang.org#1903 and braddr/at-client#4 (no point in constantly rebasing until then).
This turns out to be much more pain than initially thought :/

@andralex
Copy link
Member

andralex commented Oct 5, 2017

Let's hope gain will follow the pain. Thanks for the hard work! (Also please hop on slack.)

@wilzbach
Copy link
Member Author

ping @braddr - this is waiting for you please pull braddr/at-client#5 thx!
I've updated a couple machines with these changes. Assuming they survive without problems I'll finish deploying to the fleet tomorrow or thursday.

Thanks a lot for pulling this @braddr 🎉 🎂

@ other: FYI I just rebased this PR and it's looking good so far. There were some changes on dlang.org, so we have to update this again (-> dlang/dlang.org#1971), but apart from that, this PR is ready and waiting for the other auto-tester machines to be updated :)

@wilzbach wilzbach force-pushed the ddmd_dmd_transition branch from 8552ed4 to 8dedc6b Compare December 13, 2017 23:49
@wilzbach wilzbach changed the title [WIP] ddmd -> dmd transition ddmd -> dmd transition Dec 14, 2017
@wilzbach wilzbach removed the Review:WIP Work In Progress - not ready for review or pulling label Dec 14, 2017
@wilzbach wilzbach force-pushed the ddmd_dmd_transition branch 2 times, most recently from 4291752 to 7396f68 Compare December 14, 2017 05:59
@wilzbach
Copy link
Member Author

Alrighty - this is passing on all auto-tester machines now:

image

According to @braddr

Changes are deployed to most machines. The ones left I can't access remotely, so will be at least another day before they're updated.

This means in a very foreseeable future we can finally move on with this.
Any last-minute or final thoughts?

This branch has conflicts that must be resolved

Don't get confused by this. The current stream of merged PRs into dmd is just too high to be able to constantly rebase this. Note that every merge PR will arise a conflict and as soon as we merge this here, all open PRs will have to be rebased once. This is fairly trivial though as git can handle renames quite well.

@wilzbach wilzbach force-pushed the ddmd_dmd_transition branch from 7396f68 to d937ba7 Compare December 14, 2017 10:32
@andralex
Copy link
Member

Can't wait!!

@wilzbach
Copy link
Member Author

Can't wait!!

image

It looks really good. There are just a few hosts which haven't run these changes yet:

  • bellevue.puremagic.com (sets DMD for Druntime properly to /home/braddr/sandbox/at-client/master-132655-FreeBSD_32/dmd/generated/freebsd/release/32/dmd)
  • mac-vm-1.local (Last Heard From: 2017-12-13)
  • work-vm-1 (Last Heard From: 2017-12-13)

And there seem to be two other other hosts in the list which are solely used to the new FreeBSD OS changes:

  • ip-172-31-43-66 (idle, FreeBSD project)
  • ip-172-31-46-60 (idle, FreeBSD project)

But let's better wait for @braddr to give his okay.

@wilzbach wilzbach force-pushed the ddmd_dmd_transition branch 2 times, most recently from 8df0f47 to cd3ca68 Compare December 15, 2017 02:27
sed -i "s/ddmd/dmd/g" -i **/*.[cdh]
sed -i "s/ddmd/dmd/g" -i **/*.sdl
sed -i "s/ddmd/dmd/g" -i **/*.mak
sed -i "s/ddmd/dmd/g" -i **/*.visualdproj
sed -i "s/ddmd/dmd/g" -i **/*.vcxproj
sed -i "s/ddmd/dmd/g" -i **/*.filters
sed -i "s/ddmd/dmd/g" -i CODEOWNERS
@wilzbach wilzbach force-pushed the ddmd_dmd_transition branch from cd3ca68 to cee698b Compare December 15, 2017 02:29
@wilzbach
Copy link
Member Author

But let's better wait for @braddr to give his okay.
All the machines are updated now.

Wohoo 🎉 🎈 🎂

Awesome. Thank you so much!
Everyone ready?

@wilzbach
Copy link
Member Author

I just saw @klickverbot's comment from January this year:

The actual file-level reorganisation is something all of us can do given 20 minutes and a sed implementation of choice (well, okay, maybe an extra 10 to fix up the Windows makefiles).

...

@andralex
Copy link
Member

So should I force merge this?

@wilzbach
Copy link
Member Author

So should I force merge this?

Yes, please. Jenkins is failing due to it running out of space (CC @MartinNowak) and Travis seems to have hit a runnable/test17868b which appears to be a spurious failure (after all, it passed a couple of hours ago).

@andralex
Copy link
Member

alrighty then

@andralex andralex merged commit a2697c4 into dlang:master Dec 15, 2017
@andralex
Copy link
Member

no mercy

@wilzbach wilzbach deleted the ddmd_dmd_transition branch December 15, 2017 05:38
@RazvanN7
Copy link
Contributor

Awesome. Thanks!

@MartinNowak
Copy link
Member

Nice, a bit unfriendly to merge this just on the beta branch day.
https://forum.dlang.org/post/oknjtm$1unb$1@digitalmars.com

@andralex
Copy link
Member

@MartinNowak sorry, lmk how I can help make things better

@wilzbach
Copy link
Member Author

Same here! Did you ran into any issues already?
Sorry Martin. I lost totally track of the branching day.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

8 participants