if a new version is going to be released

[flippedcracker]

i think it needs to written from scratch. that's the problem we ran into before. the original gm code is so convoluted and mashed together that it's just not nearly as efficient or proper as it could/should be. this is a huge part of why we didn't do new versions often. i personally feel that a new version should not be release unless it's a complete rewrite.

[petefinnigan]

Hi flippedcracker,

I am not so sure about that. If we did a complete re-write the mountain may be too big to climb in one go. I am more of a person who takes something and moulds it into something else. I think that there is scope to move GM to a more stable base and effectively a complete rewrite but not in one go.

For me, the two key things are to get comments working so that I can avoid spam. I have found most of the old mods via googles cache and was planning to summarise them here tomorrow. I am also working on the comment moderation mod I suggested some time ago. I also want to include having commentors enter a unique number before being allowed to post and also a mechanism where the first post is moderated and after being approved a poster can then post without further moderation. I will post details of how I plan to do all of this in the next day or so.

The second thing I would like in the short term is draft posts. This is easier to do. I start posts on my Oracle security blog and tend to leave windows open in IE on my different computers. I wnat to be able to start a post and then save it and come back to it.

In terms of a complete rewrite, I think Coldstones suggestion to isolate the file write code so that a data access API can be created is a good one. This will allow a move to a database easilly, a move to multiple blogs and also a move to categories. If the functional code is seperated from the data layer it will be easier to do this.

I think if we started on a complete rewrite it would be harder than to morph the code into what we want.

What is everyone elses views?? - can you add a poll coldstone?

cheers

Pete

[flippedcracker]

petefinnigan said:

The second thing I would like in the short term is draft posts. This is easier to do. I start posts on my Oracle security blog and tend to leave windows open in IE on my different computers. I wnat to be able to start a post and then save it and come back to it.

this can already be done. just make the post, and then close it. i guess the only problem with that is that the time and date will be from when you first save the post, not when you actually make it live.

the reason i think it needs a rewrite is because if you keep building on or changing bad code, it's just gonna get worse. once you really get in there and try to start modifying it adding stuff, you'll see just how bad it is. yes, it's a big monster to rewrite. exactly why it's failed twice already. but just covering up old bugs inserting work-arounds for other features really isn't going to help it much.

[petefinnigan]

Hi Flippedcracker,

Thanks for your post. I was aware of using closed posts as drafts but thats not an ideal solution. It would be better to create a status for posts of 'D' and modify the code to deal with the extra status.

I can see your point of view on the code re-write and in an ideal world it would be great to start with a clean slate but as you said two attempted rewrites have failed. I am still for cleaning up the existing code, making it modular and adding the features but I will bow to the group pressure. For now I will continue to look to modify the code.

cheers

Pete

[coldstone]

Yeah, I am a similar person as Pete I think. I think GM is a starting point. I mean it works (with all its issues and what have you, but a refactor is definatly quicker and we can learn from the originals mistakes). As I go through the code, I mark issues I see, and 'use strict' and 'use warnings' where I can.

I have some personal dev guidlines I would like to post when I post the code I have changed. I think we can definatly modify the current code in a positive way, which is why I think 1.6.1 should have few 'new' features.

[coldstone]

Just an update for anyone following. Since this thread was started there has been 6 Greymatter releases over the last year+. Greymatter is being rewritten, just in bits and pieces at a time.

With everyone's help and support, I believe Greymatter is more secure, less buggy, and contains more features than previously. Something I am not sure I could say if we had attempted to do a complete rewrite.

[petefinnigan]

I agree with Coldstone, I think the re-factoring approach is vindicated as shown by the great progress made on GM, the code is miuch more compact, readable and is getting into great shape to add major mods that we talked about originally.

[marlinnut]

Agree _totally_. I'm not smart enough to write code, but smart enough to understand it and tell good from bad. I love the fact that the code is being rewritten a section at a time without orphaning features - or users (are you listening, MicroSoft ...)

I think we should continually benchmark GM against the other options, and steal ... er, borrow ... liberally from those features that make folks want to jump ship to the competition. At the same time, we need to continue the process on rebuilding the basic foundation so that it never gets so stale folks want to give up on it again.

Just one user's $.02 ...