August 25, 2014 at 20:06Organizer for students: history and plans
Hello!
Tomorrow this blog will cease to exist, because in the end I would like to tell the story of the creation of our organizer for students and share plans for the future.
It all started in the spring of 2013, when Raymondo and I finally decided to automate the accounting of deadlines for tasks, because before the session there was a real problem: how to put all the remaining tasks in a heap and distribute their execution time taking into account priorities.
To begin with, we decided to write Model in C ++ / SQLite, so that in the conditions of limited resources we can cover as many platforms as possible. There were no problems with the implementation, but it turned out that if you still use such a scheme with Objective-C, then fussing with Java wrappers is clearly not a good idea. In addition, the model for freeing memory of a minimized Android application imposes some restrictions on the architecture. In general, it became clear that you won’t cook porridge, and you need to go on the other side. Well, they took up the desktop version on Qt, which was written over the summer. But alas, the release did not work - slurred SQLite crashes began during intensive work with the program. It was not possible to quickly fix them, I had to rewrite Model to QSql, and since all the same one fails at all.
After the new year, the release finally turned out, and by March they even hit Habr:
It became even more obvious that you need a mobile application, on the other hand, by this time Digia released Qt 5.2, which announced support for mobile platforms. Already guess what the rake was next?) That's right, we thought that since this is the case, we need to write a cross-platform application on Qt Quick, focusing on starting with Android. Alas, the technology turned out to be very, very crude, especially for writing traditional applications. Very many widgets have to be implemented independently, and those that are or look not native or do not work as they should. Yes, and I had to tinker with scaling for different screens. As a result, we got the maximum viewer of the schedule and tasks, with which we can’t see high places in the store as our ears. For the prospect of at least some kind of monetization, we need not only full-fledged and native mobile applications, but also access to Western students. Well and also such trifles as motivation and free time.
It will be a pity if, in general, a good and probably necessary product will wither, we decided to open the sources of all the developments, both desktop and mobile. You only need to comb the code a bit, in general coming soon. To ensure synchronization through DropBox, which is implemented, but is not yet available in the current version, the official assembly will be published, because it will not work to lay out the secret key in open access.
Of course, we got an armful of invaluable experience in the full cycle of application development, defended a diploma on this topic, and just went for just fun, so we have no regrets. The experience of creating a desktop application is described in thisa series of articles, and I plan to start sharing my QML development experiences in the coming weeks, including in articles on writing custom components. I am glad that the guys from Digia are not sitting idly by, actively fixing bugs, and this fall there will be a version with support for the native theme of Android design, so the direction is very interesting.
You can download the organizer here , vk group here .
PS Many thanks to TM for their startup support program - this is really great.
Tomorrow this blog will cease to exist, because in the end I would like to tell the story of the creation of our organizer for students and share plans for the future.
It all started in the spring of 2013, when Raymondo and I finally decided to automate the accounting of deadlines for tasks, because before the session there was a real problem: how to put all the remaining tasks in a heap and distribute their execution time taking into account priorities.
To begin with, we decided to write Model in C ++ / SQLite, so that in the conditions of limited resources we can cover as many platforms as possible. There were no problems with the implementation, but it turned out that if you still use such a scheme with Objective-C, then fussing with Java wrappers is clearly not a good idea. In addition, the model for freeing memory of a minimized Android application imposes some restrictions on the architecture. In general, it became clear that you won’t cook porridge, and you need to go on the other side. Well, they took up the desktop version on Qt, which was written over the summer. But alas, the release did not work - slurred SQLite crashes began during intensive work with the program. It was not possible to quickly fix them, I had to rewrite Model to QSql, and since all the same one fails at all.
Release date
After the new year, the release finally turned out, and by March they even hit Habr:
It became even more obvious that you need a mobile application, on the other hand, by this time Digia released Qt 5.2, which announced support for mobile platforms. Already guess what the rake was next?) That's right, we thought that since this is the case, we need to write a cross-platform application on Qt Quick, focusing on starting with Android. Alas, the technology turned out to be very, very crude, especially for writing traditional applications. Very many widgets have to be implemented independently, and those that are or look not native or do not work as they should. Yes, and I had to tinker with scaling for different screens. As a result, we got the maximum viewer of the schedule and tasks, with which we can’t see high places in the store as our ears. For the prospect of at least some kind of monetization, we need not only full-fledged and native mobile applications, but also access to Western students. Well and also such trifles as motivation and free time.
The king is dead. Long live the king!
It will be a pity if, in general, a good and probably necessary product will wither, we decided to open the sources of all the developments, both desktop and mobile. You only need to comb the code a bit, in general coming soon. To ensure synchronization through DropBox, which is implemented, but is not yet available in the current version, the official assembly will be published, because it will not work to lay out the secret key in open access.
Of course, we got an armful of invaluable experience in the full cycle of application development, defended a diploma on this topic, and just went for just fun, so we have no regrets. The experience of creating a desktop application is described in thisa series of articles, and I plan to start sharing my QML development experiences in the coming weeks, including in articles on writing custom components. I am glad that the guys from Digia are not sitting idly by, actively fixing bugs, and this fall there will be a version with support for the native theme of Android design, so the direction is very interesting.
You can download the organizer here , vk group here .
PS Many thanks to TM for their startup support program - this is really great.