12,539 reputation
11746
bio website
location
age
visits member for 2 years, 10 months
seen 21 hours ago

Oct
15
revised Strategy for code review before merge to master from feature branches
more about how it works in Linux
Oct
15
revised Strategy for code review before merge to master from feature branches
added 552 characters in body
Oct
15
comment Strategy for code review before merge to master from feature branches
The revisions exist and are pointed to by reflog entries. But as branches there is just one, topic. Because branch in git is just the name.
Oct
15
comment Strategy for code review before merge to master from feature branches
There should be no topic-0, topic-1 and topic-2 branches. The second the rebase is complete, the previous version is irrelevant. So all there would be is topic@{1}, topic@{2}, topic@{yesterday}, topic@{3.days.ago} etc. to save your butt in case you find you screwed conflict resolution in the rebase.
Oct
15
answered Strategy for code review before merge to master from feature branches
Oct
15
comment Is garbage collection necessary?
@RichardJ.RossIII: Reference counting is generally the slowest memory management technique. There is simply so many reference and unreference operations that it gets really noticeable. If done manually, you can often optimize many of those operations away, but the compiler usually can't.
Oct
14
comment Why doesn't Git daemon start in the background?
Running with & won't work, because such process will still receive hangups from the terminal.
Oct
10
comment Etiquette for editing someone's pull request
Remember that git distinguishes "author" and "committer". So when you rewrite their commits to only contain what you actually want, they will still be considered authors. And you the committer.
Oct
10
answered How do the blogging sites or sites that host prose like contents store the data?
Oct
9
comment Python decorators and Lisp macros
@delnan: Technically, lisp is not modifying it either; it is using it as source to generate a new one and so would python, yes. The problem lies in the absence of token list or AST and the fact the compiler already complained about some things that you might otherwise allow in the macro.
Oct
9
comment Python decorators and Lisp macros
@SK-logic: In Lisp the source language also has to be lisp. Just Lisp syntax is very simple but flexible, while python syntax is much more complex and not so flexible.
Oct
9
comment Python decorators and Lisp macros
The thing is that macropy is not implemented using decorators. It uses the decorator syntax (because it must use a valid python syntax), but it is implemented by hijacking the byte compilation process from an import hook.
Oct
9
answered Could a reflog replace tags for bookkeeping?
Oct
9
comment When is a BIG Rewrite the answer?
@Andomar: Because the risk of incremental rewrite is a lot lower and because you have something working throughout the process.
Oct
9
answered Python decorators and Lisp macros
Oct
7
comment c++ web with nginx - Do I have to worry about thread safety?
@user2758004: There is basically no difference between threads and processes. Processes do a copy-on-write on few pages and that's it. Nginx is faster mainly because it is simply much simpler.
Oct
7
comment c++ web with nginx - Do I have to worry about thread safety?
@user2758004: Ok. So it's CGI. Therefore the server starts a new instance for each and every request. Nothing more to concern about.
Oct
7
awarded  Yearling
Oct
7
comment c++ web with nginx - Do I have to worry about thread safety?
Is your application run by nginx as CGI, FCGI, backend HTTP server (nginx is reverse proxy) or module? The answer depends on that (and mostly not on whether you use nginx or other web server in particular).
Oct
7
comment Handling Server Response Timeouts
@eddardstark: Do you mean the authentication one? Transferring money requires serious security, which means a two-factor authentication. There are many methods of it and I don't know which one you choose. It is totally irrelevant to this question except for the fact, that you use the same value when resending request.