The Big Programming Thread - Page 753

Prev 1 751 752 753 754 755 1032 Next

Thread Rules
1. This is not a "do my homework for me" thread. If you have specific questions, ask, but don't post an assignment or homework problem and expect an exact solution.
2. No recruiting for your cockamamie projects (you won't replace facebook with 3 dudes you found on the internet and $20)
3. If you can't articulate why a language is bad, don't start slinging shit about it. Just remember that nothing is worse than making CSS IE6 compatible.
4. Use [code] tags to format code blocks.

amazingxkcd

GRAND OLD AMERICA16375 Posts

August 10 2016 19:57 GMT

#15041

anyone know why getting a Hystrix stream from a hystrix enabled server return a 404? github pages are not helping here -- dashboard works fine and event-stream dependency is injected, but why would curling the hsystrix.stream URI giff 404?

grumbles

berated-

United States1134 Posts

August 11 2016 00:02 GMT

#15042

On August 11 2016 04:57 amazingxkcd wrote:
anyone know why getting a Hystrix stream from a hystrix enabled server return a 404? github pages are not helping here -- dashboard works fine and event-stream dependency is injected, but why would curling the hsystrix.stream URI giff 404?

grumbles

Did you register the servlet?

https://github.com/Netflix/Hystrix/tree/master/hystrix-contrib/hystrix-metrics-event-stream

Hhanh00

34 Posts

August 11 2016 02:51 GMT

#15043

On August 11 2016 03:11 Manit0u wrote:
Is it just for Windows or is Linux also vulnerable? I guess it'd be much harder to pull of on Linux since you need to download the stuff, make it executable and then it needs proper permissions to do any real damage. Quite a few hurdles to overcome.

Not necessarily, in some cases the open source nature of Linux was used against it. A flaw in a common code affects many systems from computers to routers. Take Heartbleed and Shellshock as an example.

Manit0u

Poland17435 Posts

August 11 2016 03:21 GMT

#15044

On August 11 2016 11:51 Hhanh00 wrote:

Show nested quote +

Heartbleed was actually the result of lazy maintainers who didn't give a damn. That's how LibreSSL was born (a fork of OpenSSL where they deleted half a million lines of code during the first 30 days of working on it) and picked up by people who care. They fixed a lot of bugs, improved performance and security greatly.

Ropid

Germany3557 Posts

August 11 2016 04:08 GMT

#15045

All you really care about will be accessible by your user account, so it's not necessary to "infect" the base system. There's sand-boxing schemes that can be used to protect a user from for example Firefox accessing everything. Those are right now not popular and not used by the packages you have on normal Linux distros. I think in theory Linux is just as vulnerable without those, and there's just a difference in practice because of popularity.

JJoNeEightY

United States509 Posts

August 11 2016 06:13 GMT

#15046

Reading these last few pages has made me start implementing a proper backup strategy. Finally. Derp.

What do you guys use for personal remote backups? Backblaze? AWS? Anybody really fond of a particular service/piece of software?

Also, Hi Tofu o7

Hhanh00

34 Posts

August 11 2016 06:33 GMT

#15047

On August 11 2016 12:21 Manit0u wrote:

Show nested quote +

Well - you are entitled to your opinion. I don't see how it is relevant to Linux being less vulnerable than Windows.

mantequilla

Turkey779 Posts

August 11 2016 12:15 GMT

#15048

Whenever I want to import data into mysql with a bunch of insert statements, a random error occurs at the nth insert but up to n-1 gets inserted. I fix the random error and then I have to revert that n-1 changes manually because they are already committed and won't work because of duplicate primary keys etc.

I know I'm supposed to wrap all of my inserts in a transaction to be able to revert in case of an error but it doesn't work like this:

set autocommit = 0;
start transaction;
//bunch of operations
commit;
set autocommit = 1;

why the fuck this does not work, do I have to read and understand a whole chapter about transactions to do this? why these things aren't a little intuitive?

Acrofales

Spain18120 Posts

August 11 2016 12:39 GMT

#15049

On August 11 2016 15:13 JJoNeEightY wrote:
Reading these last few pages has made me start implementing a proper backup strategy. Finally. Derp.

What do you guys use for personal remote backups? Backblaze? AWS? Anybody really fond of a particular service/piece of software?

Also, Hi Tofu o7

You talking about professional, or personal stuff? Why use a professional service for personal backups when Dropbox, Google Drive, OneDrive or some other virtually free service works perfectly well? Between an external hdd and a cloud service you have more than enough redundancy.

tofucake

Hyrule19155 Posts

August 11 2016 12:45 GMT

#15050

hi

our company has a colo with weekly backups, plus the base images of everyone's workstation, and aws backups

personally anything I want stored safely I keep on externals or google drive, but for the most part I don't care

hi aeth o7

Hhanh00

34 Posts

August 11 2016 13:33 GMT

#15051


use testdb;
create table test(a int not null primary key, b int);

start transaction;
insert into test values(1, 1);
insert into test values(2, 1);
-- crap, this doesn't work
insert into test values(2, 1);
-- so rollback
rollback;

start transaction;
insert into test values(1, 1);
insert into test values(2, 1);
-- ok now
insert into test values(3, 1);
-- so commit
commit;

select * from test;

What doesn't work?

Manit0u

Poland17435 Posts

August 11 2016 13:56 GMT

#15052

On August 11 2016 22:33 Hhanh00 wrote:


use testdb;
create table test(a int not null primary key, b int);

start transaction;
insert into test values(1, 1);
insert into test values(2, 1);
-- crap, this doesn't work
insert into test values(2, 1);
-- so rollback
rollback;

start transaction;
insert into test values(1, 1);
insert into test values(2, 1);
-- ok now
insert into test values(3, 1);
-- so commit
commit;

select * from test;

What doesn't work?

Were you trying to insert duplicate primary keys the first time?

On August 11 2016 15:33 Hhanh00 wrote:

Show nested quote +

Well - you are entitled to your opinion. I don't see how it is relevant to Linux being less vulnerable than Windows.

The sole fact that nothing you download has execution privilages by default is a great help.

Linux can be much less vulnerable if you're using chroots, user groups and permissions properly. If you're scared of losing some sensitive data you can always change folder permissions for it so that your every-day personal account doesn't have access to it. Then, nothing that you'll get infected with on this account can affect anything stored in a secured folder. You can then set up your tools (IDEs and such) to be launched with proper access rights to secure folders.

You can also set up secure folders on Windows if I remember correctly (at least you used to be able to), but it's way easier to bypass its security. The fact that you can launch javascript from command line in Windows is pretty scary, you can use it to create system messages that execute JS for example.

mantequilla

Turkey779 Posts

August 11 2016 14:27 GMT

#15053

On August 11 2016 22:33 Hhanh00 wrote:


use testdb;
create table test(a int not null primary key, b int);

start transaction;
insert into test values(1, 1);
insert into test values(2, 1);
-- crap, this doesn't work
insert into test values(2, 1);
-- so rollback
rollback;

start transaction;
insert into test values(1, 1);
insert into test values(2, 1);
-- ok now
insert into test values(3, 1);
-- so commit
commit;

select * from test;

What doesn't work?

Thanks for answering but I didn't get your example. The behaviour I was expecting is "if an error occurs rollback everything, else commit everything". In the example, you know where the error will occur and purposely rollback so, what's the point? Like:

start transaaction;
insert ...
insert ...
insert ...
commit;

(if an error occurs at 3rd insert statement, rollback 1 and 2)

Acrofales

Spain18120 Posts

August 11 2016 14:34 GMT

#15054

Well, if all this shit starts as JS exploits, then presumably they can create binaries with execution permissions (which you don't need root for in most distros). From there, the world is their oyster and as long as they have some exploit for getting root, or can operate without root (like ransomware just encrypting your whole home directory), then Linux is not inherently safer than Windows. Linux is safer because most end users, including those in corporate environments, run Windows, and it is therefore simply a numbers game: make a Linux (or Mac) virus, you can expect to hit X users. Make a Windows virus, and you hit an order of magnitude (or two) more users.

And yes, you can set up a Linux box to be very secure. However, you can also set up a Windows box to be very secure. It's just a pain in the ass to do so for either system. I hate typing in passwords, as do most users. Having to change users (and thus type in a password) to access my photos, documents, etc. would be fucking irritating, and therefore I don't do that. On the other hand, backups I run as a separate super user, so the only one with write access to the backups (of my photos, docs, etc) is that superuser, so while a virus/ransomware might destroy/encrypt/corrupt the originals, it does not have write access to the backups. This is also further backed up on Google Drive. The only way of corrupting the backups is by actually backing up the corrupted data, which would overwrite my clean backup. Even then, Google Drive should be able to restore earlier versions, but it'd (probably) be a major pain in the ass. Luckily, I have never needed to go that far. I have also never gotten ransomware or a serious virus, and my backup regime has suffered at times, but luckily threads like this exist to keep me paranoid

Hhanh00

34 Posts

August 11 2016 15:23 GMT

#15055

On August 11 2016 23:27 mantequilla wrote:

Thanks for answering but I didn't get your example. The behaviour I was expecting is "if an error occurs rollback everything, else commit everything". In the example, you know where the error will occur and purposely rollback so, what's the point? Like:

start transaaction;
insert ...
insert ...
insert ...
commit;

(if an error occurs at 3rd insert statement, rollback 1 and 2)

Yeah, this doesn't work that way. If a exception occurs, the client gets notified and the behavior depends on what it wants to do. The transaction isn't always rolled back. In workbench, it displays an error and you can continue. mysql client will quit in which case the transaction rolls back automatically. If you want control, wrap everything in a procedure with an exception handler like this:


delimiter //
drop procedure if exists p_test //
create procedure p_test(in v int)
begin
	declare exit handler for sqlexception rollback;
	start transaction;
		insert into test values(1, 1);
		insert into test values(2, 1);
		insert into test values(v, 1);
	commit;
END //
delimiter ;

-- will rollback
call p_test(2);
select * from test;

-- will commit
call p_test(3);
select * from test;

Edit: It's usually preferable to catch the error in the client because then the application would be aware of the failure.

mantequilla

Turkey779 Posts

August 11 2016 17:39 GMT

#15056

Hmm. I think I get it now. To roll back or not is client's choice.

Is the sample code standard sql or mysql specific?

amazingxkcd

GRAND OLD AMERICA16375 Posts

August 12 2016 13:35 GMT

#15057

On August 11 2016 09:02 berated- wrote:

Show nested quote +

Did you register the servlet?

https://github.com/Netflix/Hystrix/tree/master/hystrix-contrib/hystrix-metrics-event-stream

yup. its wierd

norlock

Netherlands918 Posts

August 13 2016 10:39 GMT

#15058

http://www.teamliquid.net/forum/tech-support/512785-creating-a-teamliquid-app If anyone is interested in helping me out. First of all I still need to wait for docs of the API (hopefully they can provide me with that).

graNite

Germany4434 Posts

August 14 2016 13:53 GMT

#15059

I am trying to code a builder for table tennis exercise graphics. I want it to show a table for each rally of the exercise, with arrows on it that show the movement path of the ball and rectangles of areas where to play.

Right now it looks like this: https://codepen.io/graNite/pen/grqXOo

You can add new tables, remove the last one, and draw two fixed arrows on the same canvas layer on the first table.

What I want to enable is to draw arrows* on each table via click-and-drag and delete them with right-click on them.

*(like they are drawn in the drawArrows function)

What is the best way to do this?

I already implemented a canvas layer for the arrows and could even do so for every single arrow on each table, but how can I access an arrow that is in a middle layer if I right click on it to delete it?

Djagulingu

Germany3605 Posts

August 14 2016 19:07 GMT

#15060

On August 13 2016 19:39 norlock wrote:
http://www.teamliquid.net/forum/tech-support/512785-creating-a-teamliquid-app If anyone is interested in helping me out. First of all I still need to wait for docs of the API (hopefully they can provide me with that).

I'm down for this shit.

Prev 1 751 752 753 754 755 1032 Next

Please or register to reply.

The Big Programming Thread - Page 753

Completed

Ongoing

Upcoming