Wikipedia:Reference desk/Archives/Computing/2018 August 6
Computing desk | ||
---|---|---|
< August 5 | << Jul | August | Sep >> | August 7 > |
Welcome to the Wikipedia Computing Reference Desk Archives |
---|
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages. |
August 6
[edit]Counting IP addresses
[edit]Imagine two IP addresses, 123.456.78.000 and 123.456.81.255. How many addresses are in the range defined by those two? [I just picked random digits; if there's something special in this particular range, please ignore it.] I'm guessing 1,024, since it's 000-255 for .78, .79, .80, and .81, but I've never understood ranges well; in particular, the bit at mw:Help:Range blocks about only two can be assigned to a computer has always confused me. Nyttend backup (talk) 15:47, 6 August 2018 (UTC)
- That's not a very helpful description for a page about range blocks. If I can generalise a bit, in terms of network design the top and bottom addresses are used for network-related things, hence in a block of four only the middle addresses would be used by clients. The lowest IP ('network address') is used for finding the network, and the highest IP ('broadcast address') is used for broadcasting to everything on the network. But as a Wikipedia sysop you couldn't care less how the network is designed, so you'd include them all. In your example you'd be better off using 123.456.0.0 and 123.456.3.255 as that's a simple /22 range containing 1024 addresses of which at most 1022 are going to be used to edit. But you'd block them all. -- zzuuzz (talk) 18:33, 6 August 2018 (UTC)
- ^ This. It is usually poor form to answer "how to do X" by "do Y instead", but here I really do not see why you would want to count "real" IP addresses (if even you can do this from outside the network). TigraanClick here to contact me 18:44, 6 August 2018 (UTC)
- Re "why" — it's a real-life situation unrelated to a rangeblock; I brought in rangeblocking only because that help page is the only place where I've ever seen an explanation of IP ranges. I'm just trying to figure out the raw number of addresses: that's why zzuuzz's penultimate sentence is helpful. Nyttend backup (talk) 20:07, 6 August 2018 (UTC)
- You might find this useful. I have generalised a bit - you might find the network actually contains a few smaller networks, or is part of a larger network. That's to say, there may be none or many network and broadcast addresses which won't be used, but you'd still probably count them all as part of the network. -- zzuuzz (talk) 20:15, 6 August 2018 (UTC)
- Re "why" — it's a real-life situation unrelated to a rangeblock; I brought in rangeblocking only because that help page is the only place where I've ever seen an explanation of IP ranges. I'm just trying to figure out the raw number of addresses: that's why zzuuzz's penultimate sentence is helpful. Nyttend backup (talk) 20:07, 6 August 2018 (UTC)
- ^ This. It is usually poor form to answer "how to do X" by "do Y instead", but here I really do not see why you would want to count "real" IP addresses (if even you can do this from outside the network). TigraanClick here to contact me 18:44, 6 August 2018 (UTC)
- In this old-style IP number system, each of the four parts can be from 0 to 255. So in your example it is 1 x 1 x 4 x 255 = 1024. Bubba73 You talkin' to me? 17:42, 7 August 2018 (UTC)
- Minor nitpick: the example IPv4 address given in the OP (123.456.78.000) is invalid because the second component is larger than 255. (Also the final component would normally be written as just "0" rather than "000".) But ignoring those quibbles, the OP's original guess and Bubba73's identical response are correct. CodeTalker (talk) 20:20, 7 August 2018 (UTC)
- Not-so-minor nitpick: 1 x 1 x 4 x 255 is 1020, not 1024. Another way of doing this is to utilise the "geolocate" link at the bottom of the IP's contribution page. For 123.45.67.89 the "Decimal" is 2066563929. For 123.45.68.125 the "Decimal" is 2066564221. Subtracting the one from the other, the difference is 292 IP addresses. 95.150.52.197 (talk) 14:25, 8 August 2018 (UTC)
- Presumably Bubba73 meant 1 x 1 x 4 x 256 since there are 256 numbers from 0 to 255. TigraanClick here to contact me 15:46, 8 August 2018 (UTC)
- Not-so-minor nitpick: 1 x 1 x 4 x 255 is 1020, not 1024. Another way of doing this is to utilise the "geolocate" link at the bottom of the IP's contribution page. For 123.45.67.89 the "Decimal" is 2066563929. For 123.45.68.125 the "Decimal" is 2066564221. Subtracting the one from the other, the difference is 292 IP addresses. 95.150.52.197 (talk) 14:25, 8 August 2018 (UTC)
- We can call it a TV IPv4 address. (I noticed a long time ago that IPv4 addresses in TV shows always seemed to be invalid generally because they had one number greater than 255. I thought this was funny but it later occurred to be it was likely generally intentional the same way Fictitious telephone numbers are often 555 (telephone number) in US TV shows.) Nil Einne (talk) 14:10, 9 August 2018 (UTC)
- BTW regarding the point about explanation above, it's covered in Classless Inter-Domain Routing although it may be a little technical if your understanding of the subject is close to zero. Nil Einne (talk) 14:10, 9 August 2018 (UTC)
- This is fascinating. How do IP addresses feature in TV shows? A Sussex lady had the misfortune to be targeted by Donald Trump in a case of mistaken identity [1]. 109.180.237.248 (talk) 18:30, 9 August 2018 (UTC)
- I found we have a Fictitious Internet resource which mentions IP address. Another trick mentioned is evidently to use unassigned or private IP address. I probably should have mentioned I'm not just thinking of TV shows but also movies although I don't really watch that many movies. Nil Einne (talk) 07:22, 10 August 2018 (UTC)
- This is fascinating. How do IP addresses feature in TV shows? A Sussex lady had the misfortune to be targeted by Donald Trump in a case of mistaken identity [1]. 109.180.237.248 (talk) 18:30, 9 August 2018 (UTC)
CD pipeline to use with multiple CI providers
[edit]What continuous delivery pipeline tools are capable of integrating with both Travis-CI and AppVeyor for the same product, either directly or by exposing multiple web hooks -- one for each provider -- to be called after all tests pass? Are any of them available as hosted services, with a free tier for open-source projects? NeonMerlin 18:35, 6 August 2018 (UTC)