Building a Software Solutions Team / Group to start a small business
-
@alexntg said:
@JaredBusch said:
@scottalanmiller said:
@pol.darreljade Why would it be so common if money isn't in abundance? It is a relatively rare development platofrm in the US and in Europe because of the very hogh cost associated with using it - it is expensive to license, expensive to support and the mainline development tools for it are quite costly. For desktops, sure. But for servers running applications, it is definitely a niche choice for companies looking to spend a lot of money or for shops looking to spend money on technology and save money on developers since Windows development is generally cheaper as it is in less demand.
This is completely out of touch. Every single small business I deal with that has not migrated out to hosted solutions use software packages built on and designed to run on windows. Yes new stuff should not be designed that way, but the embedded market has it and has had it for years and will not just change for no reason.
Hell, we just decommissioned a System 36 (not Windows, I know but extremely long time in service is the point).^This! For those developing for a web platform, the end-user doesn't see the backend platform, so the OS doesn't matter. For desktop OSes, Windows still has the market by a staggering majority. http://www.netmarketshare.com/operating-system-market-share.aspx?qprid=10&qpcustomd=0 Why would a company spend time and money developing for a platform that has less than 2 percent of the marketshare?
Linux has the largest server and end user market share. But the web has effectively 100% penetration.
Why would anyone choose to pigeon hole when cross compatible options are ao plentiful and powerful like HTML5 and Python.
You're argument seems to be to limit deployment options for their own sake. What business goal do you perceive that you are meeting by overspending and under delivering?
-
Don't get me wrong, I love .NET as a library and platform. I've run multiple .NET teams at different companies. I'm certified on it and helped write one of the third party .NET certs. I've managed .NET SaaS (as a vendor) for nine years.
But the places where it makes sense today and few and far between and almost solely based on fear of the unknown.
-
@scottalanmiller said:
@alexntg said:
@JaredBusch said:
@scottalanmiller said:
@pol.darreljade Why would it be so common if money isn't in abundance? It is a relatively rare development platofrm in the US and in Europe because of the very hogh cost associated with using it - it is expensive to license, expensive to support and the mainline development tools for it are quite costly. For desktops, sure. But for servers running applications, it is definitely a niche choice for companies looking to spend a lot of money or for shops looking to spend money on technology and save money on developers since Windows development is generally cheaper as it is in less demand.
This is completely out of touch. Every single small business I deal with that has not migrated out to hosted solutions use software packages built on and designed to run on windows. Yes new stuff should not be designed that way, but the embedded market has it and has had it for years and will not just change for no reason.
Hell, we just decommissioned a System 36 (not Windows, I know but extremely long time in service is the point).^This! For those developing for a web platform, the end-user doesn't see the backend platform, so the OS doesn't matter. For desktop OSes, Windows still has the market by a staggering majority. http://www.netmarketshare.com/operating-system-market-share.aspx?qprid=10&qpcustomd=0 Why would a company spend time and money developing for a platform that has less than 2 percent of the marketshare?
Linux has the largest server and end user market share. But the web has effectively 100% penetration.
Why would anyone choose to pigeon hole when cross compatible options are ao plentiful and powerful like HTML5 and Python.
You're argument seems to be to limit deployment options for their own sake. What business goal do you perceive that you are meeting by overspending and under delivering?
Are we even talking about the same thing? In my last post, I split out web applications from local applications. Linux in most cases makes sense for a web platform. For installable and many traditional client-server applications, Windows platforms have a solid majority. Despite what you may feel, Linux does not have the largest end-user market share. The vast majority of business endpoints are still Windows. It's one thing to look to the future, but it's another to ignore the present.
-
@alexntg said:
@scottalanmiller said:
@alexntg said:
@JaredBusch said:
@scottalanmiller said:
@pol.darreljade Why would it be so common if money isn't in abundance? It is a relatively rare development platofrm in the US and in Europe because of the very hogh cost associated with using it - it is expensive to license, expensive to support and the mainline development tools for it are quite costly. For desktops, sure. But for servers running applications, it is definitely a niche choice for companies looking to spend a lot of money or for shops looking to spend money on technology and save money on developers since Windows development is generally cheaper as it is in less demand.
This is completely out of touch. Every single small business I deal with that has not migrated out to hosted solutions use software packages built on and designed to run on windows. Yes new stuff should not be designed that way, but the embedded market has it and has had it for years and will not just change for no reason.
Hell, we just decommissioned a System 36 (not Windows, I know but extremely long time in service is the point).^This! For those developing for a web platform, the end-user doesn't see the backend platform, so the OS doesn't matter. For desktop OSes, Windows still has the market by a staggering majority. http://www.netmarketshare.com/operating-system-market-share.aspx?qprid=10&qpcustomd=0 Why would a company spend time and money developing for a platform that has less than 2 percent of the marketshare?
Linux has the largest server and end user market share. But the web has effectively 100% penetration.
Why would anyone choose to pigeon hole when cross compatible options are ao plentiful and powerful like HTML5 and Python.
You're argument seems to be to limit deployment options for their own sake. What business goal do you perceive that you are meeting by overspending and under delivering?
Are we even talking about the same thing? In my last post, I split out web applications from local applications. Linux in most cases makes sense for a web platform. For installable and many traditional client-server applications, Windows platforms have a solid majority. Despite what you may feel, Linux does not have the largest end-user market share. The vast majority of business endpoints are still Windows. It's one thing to look to the future, but it's another to ignore the present.
Linux moved past Windows on end points a year or two ago. Not on desktop, on end points
But writing to either platform, or any OS, for business apps is crazy and has been for years. HTML5 has been the target of choice for some time now. Windows, Linux, Mac, mobile.... All covered at once.
-
What use case are you seeing for client server apps? I haven't seen a legitimately designed one if those since the 90s.
-
@scottalanmiller said:
What use case are you seeing for client server apps? I haven't seen a legitimately designed one if those since the 90s.
Some of the things that come to mind in regards to client-server:
Rich email clients - While a basic web-based email client's sufficient for some, when it comes to integrating and automating between email and other sources, such as CRM, the rich client's where it's at.Production data acquisition - Most manufacturing equipment runs on PLCs, but those report into Windows servers or client computers that act as servers. Propietary equipment manufacturers, especially custom ones, often don't have the resources to write their software for multiple platforms, and will go with the most requested platform, traditionally Windows client with central Windows software (often MSSQL). If the majority of customers start requesting a different architecture, then they may reconsider. This ranges from $500k+ manufacturing equipment down to simple timeclock interfaces.
Niche software - There's lots of unique use cases, such as ACD software for Lync server. In addition to the ACD server, each user has a client installed on their computer that relays data to other applications for automation purposes, such as MS Dynamics CRM. Most ERP systems favor one particular platform over another, and also have locally installed clients. Most vertical-specific document management systems are client-server.
Speaking of Lync, VOIP clients are a great example of client-server. While there are some web applications that get the job done, the most feature-rich, high quality clients are, well, clients. The same goes for instant messaging clients.
-
@alexntg you are listing where it IS used, not where it made sense to use. I don't dispute that horrible decisions are made or that people don't just keep doing things that they know.
But that's not the discussion. It's about good design and what you should do when you get to do the design and make the decisions.
Doing bad things just because other people also do bad things is not good logic. It might help to justify mediocrity but it doesn't promote excellence. Making bad software is relatively easy, that's why so many companies do that.
