Allow connection to spec. host (not to all)

Hi
Is situation, for example: some app want to connect with internet to specific IP address.Shows the message:

http://makeuseof.com/images/comodo-firewall-alert.png

do you want allow this request?->yes. But when I go-> Firewall->Network security policy->This app Rule is (something like that): Allow from any to any but not allow from any to specific IP address?
Edit: this is picture from google from old version.

It is a very old CIS bug, you have no way to allow a single IP but to write it on a piece od paper and modify the new rule for all IP to only concern the one you want to allow.

and modify the new rule for all IP to only concern the one you want to allow.
Yes, just I doing. It is not yet fixed ???This serious error, CIS is not developed?

hey and Welcome!

I recommend you to download the newest CIS. uninstall the one you have but before that download CSC(comodo system cleaner) and I recommend these settings before you put CSC to clean.

run CSC. CSC —> Settings —> General. Here you mark all besides the last option.

CSC —> registry cleaner —> settings. Here you mark the second and the last option.

CSC —> privacy cleaer —> settings. Mark the second and the last option

CSC —> disk cleaner ----> settings. Mark the second and the last option.

Regards,
Valentin

I recommend you to download the newest CIS.
Ok, yes I do but my ver. is 5.0.163652.1142 and when I check program updates (CIS)is no avaiable newest version.

okey… I only said that you should download the newest because I thought you use CIS 3x version

regards,
Valentin

Hmm, and what now is any solution?

I suggest that you add the wanted exe file in firewall. that you do through CIS —> Firewall —> network security policy —> add —> select —> Browse…

Regards,
Valentin

Valentinchen, some people told you several times and today again that you do not have any obligation to post unless you want your number of posts to grow, and that you should not answer when you obviously don’t know what you are talking about.

Unless i was misleaded, i remind i am not more then you a native english speaker, the question is:
why when a Comodo allowing rule is made for a peculiar software, it is generalized not only to the ip raising this alert, but to any ip.

And, as far as i know and as i allready said, no CSC, exe, or i don’t know what “solution” is relevant:
the firewall has the behavior observed by the OP, it has been known for a very long time, and nothing inside the firewall itself but manually amending the said rule as soon as it is written can help it.

Period.

off topic:

I am here to help and not seeing my post nr grow!

I wrote this post because I thought he had an old version of CIS (I looked at the pic). My mistake, next time I will erase the posts that are not related to the topic.

Please note, greg111, that the irritation leaded me to be somewhat expeditive:
there actually is a way to get alerted per ip, setting the firewall to custom with maximal alert level, and the asking programs themselves to custom.

In these conditions (2 good examples are avira if you don’t use cis av or your usual mail client if you don’t trust it as a global mail client, and i am not even speaking of p2p and im clients, as then requested ip are changing every day or so), you shall get per ip alerts…but no way to set from the alert window an allowed or denied range, coming back to the first issue:
unless your software is using a single ip, you have no way but to amend the rule once it is made.

Edit: i forgot, the intervention of the Sandbox is, as usual, very nefarious in this regard: whatever trusted executable is by definition allowed to do a lot of things in your back, at least on the outbound way.
If you want fine control of your connexions, both Sandbox and trusted files/executables/vendors should of course be disabled.

I wrote this post because I thought he had an old version of CIS (I looked at the pic).
Sorry for this picture-I found them on the google pictures, my mistake.
setting the firewall to custom
I have that setting. Nevermaind, I be doing this by hand(Firewall->Network security policy->App Rule).
unless your software is using a single ip, you have no way but to amend the rule once it is made.
I not use p2p.I for this moment i have idea-I paste particular IP to program IPNetInfo http://www.nirsoft.net/utils/ipnetinfo.html and I see range IP adresses and blocks(or allow-depend what I want to do, of course) this IP range. Maybe i new version could receive an option -block/allow "inetum" (see in image (again from google ;) :

http://imagenes.sftcdn.net/es/scrn/36000/36366/3_ipnetinfo.jpg

I know by this option is possible block/allow to many hosts, but this is should be only option.
Sorry for my english.

there is no bug. and since i use comodo, there wasnt a bug which causes this.

if you look at the setting, it explains to you exactly what you get when you choose one of these settings (firewall is in custom mode):

firewall----settings of the behaviour of the firewall-----alarm settings.
if you want to have a rule FOR EACH adress, you have to choose “frequency of alarms: very high”. you can read in the examples that all other settings are way more simple.
NOT because of a bug, BUT because of people who dont want to have too much questions.

the default setting is made for people who want to answer one time per application. thats why i allways say: go through each window and choose what you want to have. its all there, you just have to find it :wink:

if you want to have a rule FOR EACH adress, you have to choose "frequency of alarms: very high". you can read in the examples that all other settings are way more simple.
Works fine, thanks O0
there is no bug. and since i use comodo, there wasnt a bug which causes this.

If wanting to be very strict with syntax, yes, we actually can’t speak of a bug in the formal point of view, as a bug is supposed to be a failure taking advantage of a programming fault.

Here, we have, if we want to respect the meaning of the words, not a bug, since there’s no programming fault, but no programming at all.

You are kidding: explain to others that, on a per IP alert basis, they should write as many rules as existing IP in, e.g. and taking a very simple situation, 80 discrete ip (in the mathematical acceptation of the term, as you like syntax) in the (random but representative of what happens) 191.211.008.001-191.212.064.255 range if your mail provider, software, or whatever you call it works in the said range.
Not serious, i reming e.g. the avira update rule according to this behavior, leading at the day speaking to a set of dozens of non contiguous ip in 3 different ip ranges, and still not enough to do the job.

brucine
you speak to me? i dont understand what you try to say. and english syntax is not my usual one. so dont nail me onto special spellings :smiley:

the topic opener wants to have ip based questions, and ip based made rules then. now he got the abillity through the hint of a setting.

IF you allready know that your program needs many ip adresses to work, you can answer one question, and then you can edit this rule like you wish. as you speak about avira, why not giving it the permission for OUTgoing only? and btw, i went through the process to give avira all ip based permissions… :wink:

its your choice to choose the high setting, or to stay with “one question per application”. so i am not kidding. when you choose lower settings, many of your choices are made by the firewall.

i dont see why you find a problem with that. my suggested setting gives you the chance to make it ip based, or not, BEFORE a general allow rule is created for “not yet answered things”. so it is a plus of specified security, if wished.

it is definitely strange, to get a question which talks about ONE ip, but when you allow that, you find a rule with a general permission. i fully understand the topic openers question. and my suggestion answers his question as you saw. so all is fine.

comodo firewall should ask based on what a rule it is about to create. then the user would exactly know, what is really asked and made! example: “.exe tries to connect ‘the internet’ to port 80. is it allowed?” to show that the application will get all permissions to reach port 80. this wouldnt be clear (as we saw), when the full details are presented in the question, even if they dont be part of the decision. its luck which ip adress appears in the first (and only) question, so it is useless to see it, while my answer would produce a rule that doesnt care about ip adresses at all, but only the port.