Thomas Delrue on 31 Mar 2016 07:28:42 -0700 |
[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]
Re: [PLUG] Bash on Windows? |
First off: running bash on Windows is not a big thing, it's just conhost.exe that now hosts bash. The rest is relatively straightforward. The big thing here is that unmodified plain-ol' ELF binaries are running on top of Windows. *That* is the big thing in my opinion. On 03/31/2016 08:07 AM, Anthony Martin wrote: > This is meant as a tool for developers more than actual desktop use from my > understanding. I think Microsoft is grasping at straws trying to pull > themselves out of the hole they are in that just keeps getting deeper with > smother walls. I don't think MSFT is 'grasping at straws'. The hole they've dug themselves isn't that bad of a place to be in judging by their latest earnings reports. As much as I like to 'bash' (get it?) MSFT, I don't think this is a particularly good argument. I think what we are seeing is the first E of the tried-and-tested EEE (Embrace-Extend-Extinguish)... and it is not surprising to me that Canonical is a part in this saga. > Bringing SQL server to Linux is proof they realize that > windows is slipping in the enterprise environment and this move with adding > "BASH" to Windows 10 is admitting they are losing a lot of developers to > Linux and OSX. I couldn't disagree more. Bringing SQL to Linux has little to do with those developers. As a developer, you literally don't care what SQL server is running on. As long as your app can talk to it and you can talk T-SQL to it, you don't care one hoot about it. I think bringing SQL to Linux is a combination of the following: 1) bring more linux folks into the microsoft eco-system: oh, you are running linux, not to worry, you can come and join our playground and run SQL Server on it. All in all, even though /I/ like Postgres better than SQL Server, SQL Server is still a pretty good product. Essentially, this ties in with "oh you have Linux folk in your org? Now they can do Windows as well without any re-training needed! No need to spend more monies or hire more folk, everyone can do everything now! Also DEVOPS!!!") 2) It is also an internal technical exercise in trying to reduce the mess that is the SQL codebase; just like how Win8's re-architecture had as a side effect that there was now ONE, count them: ONE, code-base for all the windowsen in the world: one codebase which runs on desktops, on tablets, on phones and on XBoxen - this is great to reduce your support surface. And doesn't that explanation of reducing your support window now suddenly also start to explain the forced upgrade to Win10? ;) SQL's code is a royal mess (I used to work on that code, btw, half a decade ago, but still). It's its own OS which runs inside an OS and does it OS things it shouldn't be doing like SMP, I/O & memory management, etc. If you can make SQL run on non-Windows OS's (and do it properly), you will have simplified your codebase dramatically. 3) This ties in with their long used strategy of "Better Together"/Vertical Integration: Oh, you're running a web app on Linux already? Well, you don't have to acquire a new skillset to run SQL or even a new box running Windows because we know you likely don't have those. (pssst, did we mention that we'll help you migrate to Windows in a couple of years as well). 4) Having SQL run on linux goes directly head-to-head with Oracle. *That* is the real target for this particular exercise. MSFT hates, Hates, *HATES* Oracle and wants to cull it into oblivion. Having SQL run on Linux opens up the market of replacing Oracle running on Linux. There is one thing where I'll agree with you though: With the second advent of cloudness (the first one being Mainframes and time-shares), as much as I dislike it, the market for OS's is drying up. They're all pretty much doing similar things anyway and most of them are being used to run a webserver (I kid, I kid...!). The OS is no longer the differentiating factor: capabilities are the differentiating factor now and having SQL Server as a capability is a big differentiator. I think that this point is evidenced by MSFT offering Linux in Azure. > On Wed, Mar 30, 2016 at 10:21 PM, JP Vossen <jp@jpsdomain.org> wrote: >> On 03/30/2016 09:05 PM, Mike DePaulo wrote: >>> On Wed, Mar 30, 2016 at 8:53 PM, Chris Grabowy <cgrabowy@gmail.com> >>> wrote: >>> >>> ...[snip]... >>> >>> Wow. Remember that Microsoft retired SUA/SFU. >>> https://en.wikipedia.org/wiki/Windows_Services_for_UNIX This is a good point. Many of us don't even remember SFU. >> http://news.slashdot.org/story/15/06/02/1950234/microsoft-to-support-ssh-in-windows-and-contribute-to-openssh Good tools are good tools regardless of the OS. It's been mind-boggling to me that the main way of doing SSH on windows was to use a crappy app like PuTTY. >> On the other hand, IF it actually works, having useful tools (esp. SSH) on >> Windows would be cool for us on this list. But if having useful tools on Windows is what you care about, surely this particular approach is a bit shotgun-wedding-y... no? Just write the tools for the native platform. (Oh wait, that's not cool to do anymore, because it's not HTML5+Javascript+CSS+WebScaleDB - sorry, I'll go back to the point now). Instead of SSH'ing out of a winbox, it'd be cooler if I could SSH *into* a WinBox and manage it like that. Mark my words regarding EEE: Embrace, Extend, Extinguish. Something, something, history repeating, something...mmmummmble mumble >> But for those new to Ubuntu, how does saddling Ubuntu with all the pain, >> instability and insecurity of Windows affect the perception of Ubuntu? Ubuntu receives two things right now: 1) Publicity (real publicity because people fawn over themselves trying to make sense of what is going on - FUD because people think this is the end of Windows/Ubunty) 2) A big fat truckload of cash right from Redmond, I can guarantee you that. Which leads me to a deeper question: what is the financial state of Canonical? Do they need this cash or is it just nice to have? I'm not trying to FUD here, but somehow this was (made?) interesting to do for Canonical, why is that? >> And related, how will this affect Cygwin? Is that thing still being actively developed? The question is a good one and I think that this is bad news for cygwin. If this takes off, I expect cygwin to fade into oblivion because "Bash On Windows(tm)" will be the officially supported thing.
Attachment:
signature.asc
Description: OpenPGP digital signature
___________________________________________________________________________ Philadelphia Linux Users Group -- http://www.phillylinux.org Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce General Discussion -- http://lists.phillylinux.org/mailman/listinfo/plug