I am back to using SDIdesk wiki after trying Tiddlywiki (not scalable and disorganizing), freemind (too much screen real estate) and wikidPad (no calendar or network diagram). The prime reason is that it was software designed by a programmer, you, for your own use and it shows rather than someone designing an application because he (or she) thinks it can be done.
Even though, in my opinion, the choice of Windows and Visual Basic was abominable, the features were (and remain) very useful (being html-based, various table-building structures, the two kinds of calendar and LBNL the network diagram).
I have abandoned Windows so I am running SDIdesk in an emulator on Linux - Crossover Office in a Windows 98 "bottle" very successfully indeed.
Obviously I'm chuffed to know that you're back with SdiDesk. Maybe we can discuss a bit more which are the really important features so I can prioritise them in future work.
As you know, I pretty much agree that Windows / VB was an unfortunate choice. I dabbled with getting it upgraded to VB.NET in 2008 which I've been intending to release formally. But I've found the VB.NET environment pretty horrible to use and this is another languishing project.
Presumably you have Python installed in your Linux, so if I were to send you a very experimental Python program which loads the PageStore and shows pages in the browser (but doesn't yet do networks and calendars etc.), would you be able to install and run it? (Basically running a couple of setups.)
We could then explore from there what features would be most important to make it a replacement for the VB version for you.
It's doable. I am running a hacked version of Linpus Lite on the Acer Aspire One netbook (can install and remove programs, advanced menu, command line, screen, etc.). It's based on Fedora 8, so I am running python 2.5.1, wxpython 2.8.4, tkinter 2.5.1, pyQt4 4.2, etc. Let me know what other dependencies you are using and I will install them.
Phil, As I said before, what I liked the most about SDIdesk is that you actually use it yourself as opposed to being just the creator of an application. I personally am looking forward to starting with features that you would like to use yourself. Why? So then they will be implemented in a way that you would find convenient to use.
At some point you mentioned that a two-pane outline editor format was attractive to you. I feel that to ally an outline editor with a wiki, you need to abandon the one-to-one relationship between a node (in the tree) and its associated page (or leaf in outline jargon). One of the applications / environments that I tried was wikidPad, which looks like a hierarchical two-pane outline editor but isn't. While it has two panes, i.e., one showing nodes and their children and grandchildren, etc. and the other showing a bi-modal page (one being an editable mode containing the wiki text and the other the html-generated text), the nodes do not form a tree since a child can have multiple parents. So the tree-like pane is really an index.
The way it works is each time you create a wiki word in a page that wiki word generates a potential child node. Once you have filled the new page with wiki text, the node becomes a child node. So if you use the same wiki word in different pages, each of these pages becomes a parent.
I mentioned this as I use the network diagram as a traffic cop to direct me to a particular page in my wiki (as well as indicating the relationship or mapping between the pages). So if you decide to postpone this feature then perhaps giving the wiki the appearance of a two-pane outline editor with an index and associated pages would enable me to easily find the particular page I want. This would also allow you to put off implementing a search tool.
5 comments:
Way off topic but I thought you ought to know.
I am back to using SDIdesk wiki after trying Tiddlywiki (not scalable and disorganizing), freemind (too much screen real estate) and wikidPad (no calendar or network diagram). The prime reason is that it was software designed by a programmer, you, for your own use and it shows rather than someone designing an application because he (or she) thinks it can be done.
Even though, in my opinion, the choice of Windows and Visual Basic was abominable, the features were (and remain) very useful (being html-based, various table-building structures, the two kinds of calendar and LBNL the network diagram).
I have abandoned Windows so I am running SDIdesk in an emulator on Linux - Crossover Office in a Windows 98 "bottle" very successfully indeed.
Albert
Hi Albert
Obviously I'm chuffed to know that you're back with SdiDesk. Maybe we can discuss a bit more which are the really important features so I can prioritise them in future work.
As you know, I pretty much agree that Windows / VB was an unfortunate choice. I dabbled with getting it upgraded to VB.NET in 2008 which I've been intending to release formally. But I've found the VB.NET environment pretty horrible to use and this is another languishing project.
Presumably you have Python installed in your Linux, so if I were to send you a very experimental Python program which loads the PageStore and shows pages in the browser (but doesn't yet do networks and calendars etc.), would you be able to install and run it? (Basically running a couple of setups.)
We could then explore from there what features would be most important to make it a replacement for the VB version for you.
It's doable. I am running a hacked version of Linpus Lite on the Acer Aspire One netbook (can install and remove programs, advanced menu, command line, screen, etc.). It's based on Fedora 8, so I am running python 2.5.1, wxpython 2.8.4, tkinter 2.5.1, pyQt4 4.2, etc. Let me know what other dependencies you are using and I will install them.
Thanks,
Albert
Phil,
As I said before, what I liked the most about SDIdesk is that you actually use it yourself as opposed to being just the creator of an application. I personally am looking forward to starting with features that you would like to use yourself. Why? So then they will be implemented in a way that you would find convenient to use.
At some point you mentioned that a two-pane outline editor format was attractive to you. I feel that to ally an outline editor with a wiki, you need to abandon the one-to-one relationship between a node (in the tree) and its associated page (or leaf in outline jargon). One of the applications / environments that I tried was wikidPad, which looks like a hierarchical two-pane outline editor but isn't. While it has two panes, i.e., one showing nodes and their children and grandchildren, etc. and the other showing a bi-modal page (one being an editable mode containing the wiki text and the other the html-generated text), the nodes do not form a tree since a child can have multiple parents. So the tree-like pane is really an index.
The way it works is each time you create a wiki word in a page that wiki word generates a potential child node. Once you have filled the new page with wiki text, the node becomes a child node. So if you use the same wiki word in different pages, each of these pages becomes a parent.
I mentioned this as I use the network diagram as a traffic cop to direct me to a particular page in my wiki (as well as indicating the relationship or mapping between the pages). So if you decide to postpone this feature then perhaps giving the wiki the appearance of a two-pane outline editor with an index and associated pages would enable me to easily find the particular page I want. This would also allow you to put off implementing a search tool.
Hope this helps,
Albert
Albert.
Thanks again. Useful comments. And would you like to mail me (interstar AT gmail) so we can discuss me sending a prototype to you to try out?
Post a Comment