|
it's perfectly reasonable in the 21st century to assume that an email matches .+@.+\.+
|
# ? Sep 23, 2017 12:40 |
|
|
# ? May 17, 2024 07:46 |
|
redleader posted:it's perfectly reasonable in the 21st century to assume that an email matches .+@.+\.+ lol i dont think any emails match that
|
# ? Sep 23, 2017 12:46 |
|
Powaqoatse posted:lol i dont think any emails match that maybe? have i got my regex wrong? some letters, then an @, then letters, then . , then more letters?
|
# ? Sep 23, 2017 13:20 |
|
redleader posted:maybe? have i got my regex wrong? some letters, then an @, then letters, then . , then more letters? a string your regex will reject: "dongs@butts.com" a string your regex will accept: " @ ....."
|
# ? Sep 23, 2017 13:24 |
|
You forgot a non-literal . after the literal . Otherwise I feel that's really close enough
|
# ? Sep 23, 2017 13:56 |
|
Soricidus posted:a string your regex will reject: "dongs@butts.com" oooooh whoops, lol. these things suck
|
# ? Sep 23, 2017 14:03 |
|
Vanadium posted:You forgot a non-literal . after the literal . yeah i'm definitely not arguing in favor of doing something more complicated i'd stick with "is there an @ in it? ok validate it by emailing it a confirmation link". now you don't need regular expressions. nobody uses ! any more and local addresses without an @ only apply in very specific cases where you're not validating an general-purpose email address to begin with. but there are going to be cases where there's no . in the domain such as the aforementioned "nerd@google"
|
# ? Sep 23, 2017 14:07 |
|
Everyone saying c# is bad should own up to what terrible language and platform they prefer. The .net core/standard stuff is impenetrable, but the language is great and so is the stdlib, especially considering its age.
|
# ? Sep 23, 2017 14:16 |
|
havelock posted:Everyone saying c# is bad should own up to what terrible language and platform they prefer. JAVA BICTH
|
# ? Sep 23, 2017 14:24 |
|
speaking of which https://www.oracle.com/corporate/pressrelease/java-se-9-and-ee-8-092117.html JAVA 9 MOTHERFUKCERS!!!!!!
|
# ? Sep 23, 2017 14:26 |
|
oh good, time to take another brief look at getting any of our code to work with it before sadly shaking our heads and underlining the bit in the documentation that says only java 8 is supported
|
# ? Sep 23, 2017 14:29 |
|
also reactjs and some other fb js projects just went mit licensed
|
# ? Sep 23, 2017 14:29 |
|
Sapozhnik posted:speaking of which this is as excited i get about anything
|
# ? Sep 23, 2017 14:41 |
|
a witch posted:C# is a great language with a dumpster fire of a standard library. you misspelled build system
|
# ? Sep 23, 2017 18:13 |
|
you misspelled package manager
|
# ? Sep 23, 2017 18:30 |
|
msbuild and nuget are so loving bad.
|
# ? Sep 23, 2017 18:41 |
|
microsoft hosed up .net when they decided to open source it
|
# ? Sep 23, 2017 18:44 |
|
they hosed it up when they decided to do .net core and didn't create a working build + package system. they're the biggest problems with the current state of .net
|
# ? Sep 23, 2017 18:47 |
|
they hosed up by not having maven
|
# ? Sep 23, 2017 18:50 |
|
yeah they should have used the litterrall pom and then just made a native runtime + plugins for it.
|
# ? Sep 23, 2017 18:51 |
|
Soricidus posted:but there are going to be cases where there's no . in the domain such as the aforementioned "nerd@google" good, gently caress people who use turdy tld-only email addresses I'm 100% ok with refusing to allow them to create an account
|
# ? Sep 23, 2017 18:52 |
|
Maven, more like Maybe, as in maybe it will let you build what you need "I'm sorry Dave, I'm afraid I have a different opinion"
|
# ? Sep 23, 2017 18:57 |
|
Shaggar posted:they hosed it up when they decided to do .net core and didn't create a working build + package system. they're the biggest problems with the current state of .net i still don't understand why they didn't just work on fixing mono.
|
# ? Sep 23, 2017 18:58 |
|
Soricidus posted:but there are going to be cases where there's no . in the domain such as the aforementioned "nerd@google" do people really do this gmail won't even let me send an email to a TLD
|
# ? Sep 23, 2017 18:58 |
|
Arcsech posted:good, gently caress people who use turdy tld-only email addresses imagine opening the inbox of test@test
|
# ? Sep 23, 2017 19:00 |
|
Sapozhnik posted:speaking of which
|
# ? Sep 23, 2017 19:02 |
|
Arcsech posted:good, gently caress people who use turdy tld-only email addresses i wanted an @aws but 1) there's a much better plan for it than nerd cred and 2) it would break on half the internet wish we gave people @a2z.com addresses tho FamDav fucked around with this message at 19:16 on Sep 23, 2017 |
# ? Sep 23, 2017 19:11 |
|
it's prohibited apparently? https://www.icann.org/news/announcement-2013-08-30-en
|
# ? Sep 23, 2017 19:19 |
|
JewKiller 3000 posted:it's prohibited apparently? https://www.icann.org/news/announcement-2013-08-30-en oh cool that makes life easier then
|
# ? Sep 23, 2017 19:30 |
|
jony neuemonic posted:i still don't understand why they didn't just work on fixing mono. indirectly, they did. from what I've heard, mono 5.x (which was released after and with a lot of code contributions from lolnetcore) is a lot more stable than 4.x and mono now has a webassembly compiler in alpha, which is way more interesting to me than any mild refactoring of aspnet and entity framework
|
# ? Sep 23, 2017 20:44 |
|
also paket solves practically every problem that nuget has. always use it for anything non-trivial. i don't know of any tool that similarly fixes the horror of msbuild though. there are a few interesting projects but none are close to production ready. best you can do is keep your msbuild project files as small and simple as possible (admittedly here the netcore guys actually did well, by making the new project files actually manageable) and offload anything beyond basic compilation to your language-agnostic tools of choice.
|
# ? Sep 23, 2017 20:54 |
|
Arcsech posted:good, gently caress people who use turdy tld-only email addresses i get my gonadic+company@gmail.com rejected far more than i would expect. let alone gonadic.io+company@gmail.com
|
# ? Sep 24, 2017 00:10 |
|
jony neuemonic posted:i still don't understand why they didn't just work on fixing mono. cause it would be way too hard. there are too many win32 based assumptions in .net framework, and thus mono, that separating out the pure .net stuff makes sense.
|
# ? Sep 24, 2017 00:59 |
|
NihilCredo posted:also paket solves practically every problem that nuget has. always use it for anything non-trivial. what makes paket better than nuget? more generally, what makes one package manager better than another? what makes a good package manager better than a bad one?
|
# ? Sep 24, 2017 01:28 |
|
redleader posted:what makes paket better than nuget? for starters, if you add some packages to your project, then share that project with other people, when those people build the project they will always get the exact same package versions you used. this is not guaranteed with nuget second, if you install a big package that has a lot of other dependencies, paket will keep separate track of the package you actually wanted and the ones that that package merely brought in, for the purpose of updating etc. (nuget actually fixed this with nuget 3, not sure if it survived the death of project.json though) other benefits: the files are human readable, it can reference direct source files, it can temporarily override dependencies with local versions for testing purposes, it can create conditional references based on the framework you're targeting, it can create load scripts which are super useful for quick testing, it can easily manage packages for a whole solution at once while keeping separate references for each project so as for your question "what makes a good package manager", it basically boils down to two things: - don't do unexpected stuff - don't waste the user's time - (optional) let the user reference more things
|
# ? Sep 24, 2017 01:46 |
|
two features make a good package manager: * define a project's build process using inert data instead of providing a turing-complete languge that waltzes with its os environment while making GBS threads a trail of side-effects in its wake * when i check out a snapshot of my project from source control, and it cites a particular set of transitive dependencies at that point in time, i loving expect a bit-identical copy of each of those exact transitive dependencies to be provided. because, you know, reproducible builds are generally considered to be a good idea.
|
# ? Sep 24, 2017 03:01 |
|
lolmfao if you aren't using java/maven
|
# ? Sep 24, 2017 03:03 |
|
NihilCredo posted:indirectly, they did. from what I've heard, mono 5.x (which was released after and with a lot of code contributions from lolnetcore) is a lot more stable than 4.x Shaggar posted:cause it would be way too hard. there are too many win32 based assumptions in .net framework, and thus mono, that separating out the pure .net stuff makes sense. fair enough. it's a shame to see them essentially start over but i probably shouldn't make "fix mono" sound like a trivial thing.
|
# ? Sep 24, 2017 03:10 |
|
Sapozhnik posted:* define a project's build process using inert data instead of providing a turing-complete languge that waltzes with its os environment while making GBS threads a trail of side-effects in its wake
|
# ? Sep 24, 2017 03:56 |
|
|
# ? May 17, 2024 07:46 |
|
jony neuemonic posted:fair enough. it's a shame to see them essentially start over but i probably shouldn't make "fix mono" sound like a trivial thing. well it's Microsoft, there's either five competing internally produced products of varied viability or it's not worth doing
|
# ? Sep 24, 2017 03:56 |