I am wondering if Umbraco 5 is going to require rebuilding extensions and packages like 4.5 did... It was a lot of hassle figuring out what packages were or were not compatible with the latest versions. What about macros, XSLT and user controls?
Version 5 will in most cases require a remake of a package, since the intire underlying codebase wil be rewritten. for one thing, it will be based on MVC instead of webforms. The concepts of Umbraco will remain the same, so you will still be able to use your knowledge of umbraco to build websites using macros, XSLT and so on.
If you are working directly with the .net api, then you will have to rewrite the parts of the code that interact with that. The good thing is that it look to be a much nicer API, and your code will most probably end up being much nicer to work with than today.
I think it will be pretty clear which packages work for V5, and which don't since they simply cannot install on V5, since the packageing engine will also change.
That is REALLY disheartening seeing as how we just went through the whole 4.0 to 4.5 package migrations which was not fun at all... I can't recall too many major open source projects that break existing functionality like Umbraco upgrades have. I'm pretty sure WordPress, Joomla and Drupal, among others, would not be as successful as they are if every few versions a bunch of the functionality broke!
I think this is a case where the Umbraco team is too focused on developers and not caring enough about the average website builder who just wants a solid CMS that works and can be upgraded with having to working about things falling apart.
I dunno... I am sure there are some advantages of MVC, but I am skeptical that they're so great it's worth going through yet another painful upgrade cycle.
Remember that umbraco 4.x is not going anywhere for a long time. Umbraco HQ has anounced full support for the platform for 2 years (i think it was) from the time that the final V5 is released, so if you don't need to upgrade, don't. The 4.x is a great product, and will continue to be so.
The reason for the big rewrite and move to MVC, is that the core that 4.x is based on is more or less pushed to it's limits now. And exactly because of a relatively large focus on not breaking the upgrade path, it just is too difficult to incorporate the things that v5 will bring on the v4 platform.
But again, if you are running sites on V4, and you don't absolutely need to upgrade to v5, then don't. I know that we have quite a few sites that we will not be upgrading, because it would not make sense to do so.
With regards to building new sites, yes, maybe you cannot use the same packages, but I'm sure that the community will be pretty quick at building new versions of the favourite packages.
Morten, with all due respect, how is 4.x pushed to it's limits? It's fast, it's stable... what does a CMS do? Retrieve and manage simple data, pretty much. How much complexity do we need here? I can't imagine what magic MVC will bring to the end user and content manager that 4.x doesn't already provide.
If what I understand about MVC is true, I would not assume that the community would be exactly "quick" to build new versions of packages... I know next to nothing about MVC, but I do understand it's quite different than web forms and conversion of packages will probably not be a trivial task.
And again, I am worried about the trend of breaking changes as the normal course of business for Umbraco... no other CMS out there seems to introduce new versions that don't play well with previous versions as much as Umbraco and that must limit Umbraco's popularity to some degree.
Don't get me wrong... I LOVE Umbraco, but I have found upgrades to be nasty and error prone with too many versions... I really, really hate that but it does not seem to be of much concern to the core team.
It is not MVC in itself that is brining new features to the platform, it is the brand new architecture below the hood. One of the things that wil lcome in V5 is the concept of a "Hive". This means that you would be able to use umbraco to manage content that is not stored by umbraco itself, and integrate that directly into Umbracos own content tree structure. Without the end user ever having to know that there is a difference to where it is stored. This is just one example of something that would not be possible/practical to implement in the current V4 architecture.
With regards to the community upgrading packages, the most active package devs are very talented people, and I trust that many of them will be quick to update the packages that would be requested from the community.
On the topic of other CMS'es, I disagree. I have worked with quite a few, and every single one of them has broken my code at one pint or another, after performing an upgrade. I don't think Umbraco is much different in that sense. Granted, the V5 will be a large shift in tech, but in my view, it will be well worth it as it will be bringing an arcitecture that is built to scale from day 1.
The upgrade difficulties you are experiencing now, is in a sense a symptom of the v4 architecture which is not testable, thus making any changes to the core more likely to introduce new bugs. So continuing to develop on the v4 core would only bring more buggy upgrades in the future, not less. V5 will bring a large degree of test coverage, which in my view provides much greater confidence in the platforms stability going forward from v5 and up. So ironically the change is needed in order to avoid the very thing you are worried about.
As long as we see the death of WebForms which IMHO broke the web for developers, then I'll be happy. These changes are really more to do with Microsoft changing the .NET goalposts every 3 years, for commercial reasons I'm sure. No point in making something that's great and then walking away, they have tens of thousands of wage cheques to pay every month;) For bespoke development MVC is definitely the way to go. MS is actually a bit behind the curve on this so I guess Umbraco is just keeping up.
Hello there, I only really got onto umbraco around V4.5, so I don't know about the headaches of the past.
As far as I am aware Drupal has no backwards compatability between major versions. Joomla V1 extensions only work in V1.5 in legacy mode and V1.5 to 1.6 needed rewrites. I haven't used wordpress but it looks simplistic enough to me that it wouldn't be impacted by change much. I'm not sure how Umbraco is any different to these.
Yeah, understandable to be anxious. Say for example from my own perspective, because I am fairly new to Umbraco and know there is change coming, I am holding off from getting my hands tooooo dirty. I am probably not the only noobie with a similar attitude. I would guess there is some form of momentum loss.
Regardless though from someone who hasn't really dug in deep yet, seems to me the Umbraco guys making the decisions know what they are doing :) They made a great CMS so far anyway.
Umbraco 5 Concerns...
I am wondering if Umbraco 5 is going to require rebuilding extensions and packages like 4.5 did... It was a lot of hassle figuring out what packages were or were not compatible with the latest versions. What about macros, XSLT and user controls?
Version 5 will in most cases require a remake of a package, since the intire underlying codebase wil be rewritten. for one thing, it will be based on MVC instead of webforms. The concepts of Umbraco will remain the same, so you will still be able to use your knowledge of umbraco to build websites using macros, XSLT and so on.
If you are working directly with the .net api, then you will have to rewrite the parts of the code that interact with that. The good thing is that it look to be a much nicer API, and your code will most probably end up being much nicer to work with than today.
I think it will be pretty clear which packages work for V5, and which don't since they simply cannot install on V5, since the packageing engine will also change.
That is REALLY disheartening seeing as how we just went through the whole 4.0 to 4.5 package migrations which was not fun at all... I can't recall too many major open source projects that break existing functionality like Umbraco upgrades have. I'm pretty sure WordPress, Joomla and Drupal, among others, would not be as successful as they are if every few versions a bunch of the functionality broke!
I think this is a case where the Umbraco team is too focused on developers and not caring enough about the average website builder who just wants a solid CMS that works and can be upgraded with having to working about things falling apart.
I dunno... I am sure there are some advantages of MVC, but I am skeptical that they're so great it's worth going through yet another painful upgrade cycle.
Remember that umbraco 4.x is not going anywhere for a long time. Umbraco HQ has anounced full support for the platform for 2 years (i think it was) from the time that the final V5 is released, so if you don't need to upgrade, don't. The 4.x is a great product, and will continue to be so.
The reason for the big rewrite and move to MVC, is that the core that 4.x is based on is more or less pushed to it's limits now. And exactly because of a relatively large focus on not breaking the upgrade path, it just is too difficult to incorporate the things that v5 will bring on the v4 platform.
But again, if you are running sites on V4, and you don't absolutely need to upgrade to v5, then don't. I know that we have quite a few sites that we will not be upgrading, because it would not make sense to do so.
With regards to building new sites, yes, maybe you cannot use the same packages, but I'm sure that the community will be pretty quick at building new versions of the favourite packages.
Morten, with all due respect, how is 4.x pushed to it's limits? It's fast, it's stable... what does a CMS do? Retrieve and manage simple data, pretty much. How much complexity do we need here? I can't imagine what magic MVC will bring to the end user and content manager that 4.x doesn't already provide.
If what I understand about MVC is true, I would not assume that the community would be exactly "quick" to build new versions of packages... I know next to nothing about MVC, but I do understand it's quite different than web forms and conversion of packages will probably not be a trivial task.
And again, I am worried about the trend of breaking changes as the normal course of business for Umbraco... no other CMS out there seems to introduce new versions that don't play well with previous versions as much as Umbraco and that must limit Umbraco's popularity to some degree.
Don't get me wrong... I LOVE Umbraco, but I have found upgrades to be nasty and error prone with too many versions... I really, really hate that but it does not seem to be of much concern to the core team.
Hi Robert
It is not MVC in itself that is brining new features to the platform, it is the brand new architecture below the hood. One of the things that wil lcome in V5 is the concept of a "Hive". This means that you would be able to use umbraco to manage content that is not stored by umbraco itself, and integrate that directly into Umbracos own content tree structure. Without the end user ever having to know that there is a difference to where it is stored. This is just one example of something that would not be possible/practical to implement in the current V4 architecture.
With regards to the community upgrading packages, the most active package devs are very talented people, and I trust that many of them will be quick to update the packages that would be requested from the community.
On the topic of other CMS'es, I disagree. I have worked with quite a few, and every single one of them has broken my code at one pint or another, after performing an upgrade. I don't think Umbraco is much different in that sense. Granted, the V5 will be a large shift in tech, but in my view, it will be well worth it as it will be bringing an arcitecture that is built to scale from day 1.
The upgrade difficulties you are experiencing now, is in a sense a symptom of the v4 architecture which is not testable, thus making any changes to the core more likely to introduce new bugs. So continuing to develop on the v4 core would only bring more buggy upgrades in the future, not less. V5 will bring a large degree of test coverage, which in my view provides much greater confidence in the platforms stability going forward from v5 and up. So ironically the change is needed in order to avoid the very thing you are worried about.
As long as we see the death of WebForms which IMHO broke the web for developers, then I'll be happy. These changes are really more to do with Microsoft changing the .NET goalposts every 3 years, for commercial reasons I'm sure. No point in making something that's great and then walking away, they have tens of thousands of wage cheques to pay every month;) For bespoke development MVC is definitely the way to go. MS is actually a bit behind the curve on this so I guess Umbraco is just keeping up.
Hello there, I only really got onto umbraco around V4.5, so I don't know about the headaches of the past.
As far as I am aware Drupal has no backwards compatability between major versions. Joomla V1 extensions only work in V1.5 in legacy mode and V1.5 to 1.6 needed rewrites. I haven't used wordpress but it looks simplistic enough to me that it wouldn't be impacted by change much. I'm not sure how Umbraco is any different to these.
Ok, I'm convinced... Just a little anxious about such a major rewrite, I guess.
Yeah, understandable to be anxious. Say for example from my own perspective, because I am fairly new to Umbraco and know there is change coming, I am holding off from getting my hands tooooo dirty. I am probably not the only noobie with a similar attitude. I would guess there is some form of momentum loss.
Regardless though from someone who hasn't really dug in deep yet, seems to me the Umbraco guys making the decisions know what they are doing :) They made a great CMS so far anyway.
is working on a reply...