Mikhail Ivchenko

- [ ] Preallocate reasonable amount of bytes in ```MarshalRequest``` https://github.com/go-hep/hep/pull/183#discussion_r185445598. - [x] Create code generator for marshalling to be used instead of reflection https://github.com/go-hep/hep/pull/183#discussion_r185294377.

- [ ] kXR_admin - [ ] kXR_auth - [ ] kXR_bind - [ ] kXR_chmod - [x] kXR_close - [ ] kXR_decrypt - [x] kXR_dirlist - [ ] kXR_endsess...

> @EgorMatirov want to give this a try? First thing, that I have noticed: `C++` version writes file by buckets of 16 MBytes while `Go` version uses buckets of 16...

> it seems there is now also: > > * https://docs.microsoft.com/en-us/windows/desktop/projfs/projected-file-system > > > Recent versions of Windows 10 ship with a new system facility called “Projected File System” (ProjFS)...

According to [gsi-msg-specs, p.3](https://github.com/go-hep/hep/files/2089932/gsi-msg-specs.pdf): `the context establishment phase is nothing more than normal SSLv3 handshake messages being exchanged`. I was using [RFC6101, p. 25](https://tools.ietf.org/html/rfc6101#page-25) as a specification of SSLv3 handshake...

>I guess at this point it's best to ask the xrootd mailing list :) >(or on xrootd/issues.) done. :) > feel free to push your WIP branch so I can...

Yep, I'll take a look. Thanks for reporting!

As for now, I was able to get such [stacktrace](https://gist.github.com/EgorMatirov/f89110ac00b8ab2e28d97911c91f1284), however, I still don't really understand what happens.

> (but I suspect they'll want a smaller reproducer...) The problem is that I still wasn't able to reproduce it locally. I'll try more, though.

Hi. I have made PKGBUILD for git version. Now Archlinux users can simply install it from AUR: https://aur.archlinux.org/packages/btcd-git/