My 2 cents, as a LWN "professional hacker": Mercurial planning to transition away from SHA-1

 

LWN article:

Mercurial planning to transition away from SHA-1
https://lwn.net/Articles/832111/

My comment:

https://lwn.net/Articles/833425/

Content of my comment:

My 2 cents, as a LWN "professional hacker":

Existing Mercurial users are using Mercurial with existing old-and-aging Mercurial repositories that use SHA-1.
At my job, we use Mercurial, because it is what it is.
Unless an existing repository can have both old SHA-1 hashes and $NEW-HASH hashes at the same time, I don't think that this is going to be a thing. Users just don't care.

Branching

You can think of Mercurial branches as Commit Namespaces rather than true branches as intended by CVS, Subversion, Git and really computer science in general. In Mercurial, each commit belongs to a Mercurial branch explicitly and forever. (The information of branch name is encoded in the data of a commit.) In Git, a commit does not contain in its data the branch name.

So, what did Mercurial do to fix that, in order to add standard branching semantics to replace "Mercurial Branches" ?
Bookmarks !

In Mercurial, bookmarks are basically Git branches. (A temporary label on a commit, that will advance as you add commits on top.) In Mercurial, most commands (push, pull, export, ...) support "bookmarks". You can do "hg push origin -B topic", which is equivalent to "git push origin topic". But, "hg log" does not have the -B switch for bookmarks.

In August 2020, I submitted patches to add bookmark support in the "hg log" command.
The patch https://phab.mercurial-scm.org/D8973 has been approved.
But, I have not seen it land in the maintainer's repo at https://www.mercurial-scm.org/repo/hg or even at https://www.mercurial-scm.org/repo/hg-committed/ ...

I was a bit disappointed. Was my patch series silently put to rest, behind the curtain ?

Deduplication

Mercurial has a fancy format called "Bundle" for the wire protocol. But, the .hg/store does not use "Bundle", and the result is that if you add 10 files with the same content to a Mercurial repository, the Mercurial object store will have 10 copies of it. There is no deduplication. Git has deduplication (since day one) which does the right thing of not storing 10 objects with the same content. Linus Torvalds calls that a content-addressable store.

Conclusion

To me, Mercurial is a dinosaur software. It does not evolve, and is going extinct.

At least, the third DVCS, Fossil, acknowledges what it wants to be in its name !

Comments

Post a Comment

Popular posts from this blog

My 2022 Calisthenics split routine

Hey my fellow blog readers ! At the beginning of year 2022, I watched some YouTube videos from Tom Merrick, The Bodyweight Warrior . The video 3 Things I Wish I Knew When I Started made me think a lot. Since I began my bodyweight journey in July 2020, I used a full-body routine, 3 times per week. As I became stronger, the volume of strength work became too high. Watching videos from Tom Merrick convinced me to split my full body routine into a split routine. My new Calisthenics routine is a split routine. Splitting the routine allowed me to increase rest time between upper body workouts, and also to increase rest time between lower body workouts. It also allowed me to increase from 3 (full-body) workouts per week to 4 workouts (upper body, lower body + core, upper body, lower body + core) per week. Here what my split routine looks like: Wednesday: Upper Body Workout (1h15min) Pair 1 Ring Pull-Up Pseudo Planche Push-Up  Pair 2 Tuck Assisted Victorian Box Handstand Push-Up Pair 3 Ring R
Read more

Adding ZVOL VIRTIO disks to a guest running on a host with the FreeBSD BHYVE hypervisor

On atlantis.domain (my FreeBSD BHYVE hypervisor host), I use these commands: Install FreeBSD on a 10G disk. vm create -s 10G guestname   vm install FreeBSD-12.1-RELEASE-amd64-memstick.img vm console guestname Add 8 30GB zvol-based disks vm add -d disk -t zvol -s 30G guestname vm add -d disk -t zvol -s 30G guestname vm add -d disk -t zvol -s 30G guestname vm add -d disk -t zvol -s 30G guestname vm add -d disk -t zvol -s 30G guestname vm add -d disk -t zvol -s 30G guestname vm add -d disk -t zvol -s 30G guestname vm add -d disk -t zvol -s 30G guestname You will need to do vm poweroff guestname vm start guestname because the guest does not seem to see the 8 30GB without power cycling. (rebooting won't work) *** If you want the boot disk (the one create by "vm create") to be a zvol, you must edit the template: root@atlantis:~ # zfs snapshot tank/vm-bhyve@2020-06-11-2145 root@atlantis:~ # mount -t zfs tank/vm-bhyve@2020-06-11-2145 /tank/vm-bhyve@2020-06-11-2145 r
Read more

Le tissu adipeux brun, la thermogénèse, et les bains froids

Image
Figure 4 de Mapping of human brown adipose tissue in lean and obese young men https://www.pnas.org/content/114/32/8649 BAT = Brown Adipose Tissue    Avec l'hiver à nos portes, je partage ci-dessous mon intérêt pour le tissu adipeux brun. Le tissu adipeux blanc a un rôle d'isolant et de stockage. Le tissu adipeux brun ("gras brun") diffère du traditionnel tissue adipeux blanc. Le tissus adipeux brun a un rôle de plinthe électrique (thermogénèse). Et non, manger de la galvaude ne donne pas plus de gras brun.😄 Selon le livre Human Anatomy & Physiology de Élaine Marieb , 5ème édition, publié en 2001, le tissu adipeux brun est présent seulement chez les bébés, qui ne peuvent pas frissonner. Le tissu adipeux brun fonctionnel a été découvert chez les adultes en 2009 ( New England Journal of Medicine ). Le tissu adipeux brun a été cartographié chez l'humain adulte en 2017 ( Proceedings of the National Academy of Sciences ). J'ai mis la Figure 4 de l'articl
Read more