Commit | Line | Data |
---|---|---|
55d2b2c2 KL |
1 | Jexer Work Log |
2 | ============== | |
3 | ||
d36057df KL |
4 | October 17, 2017 |
5 | ||
6 | I finally gave up the ghost on using gcj as the default compiler due | |
7 | to its awesome unused imports messages, and learned how to get PMD to | |
8 | do that job. Which promptly created 1000+ warning messages related to | |
9 | class item order (variables, constructors, methods), nested ifs, | |
10 | useless checks, and so on. So now we go on a code sweep to fix those, | |
11 | and along the way set a new class template. Since this is so large | |
12 | and invasive, I will bite the bullet now and get it done before the | |
13 | next release which will get it out on Maven finally. | |
14 | ||
be72cb5c KL |
15 | August 16, 2017 |
16 | ||
17 | Holy balls this has gotten so much faster! It is FINALLY visibly | |
18 | identical in speed to the original d-tui: on xterm it is glass | |
19 | smooth. CPU load is about +/- 10%, idling around 5%. | |
20 | ||
21 | I had to dramatically rework the event processing order, but now it | |
22 | makes much more sense. TApplication.run()'s sole job is to listen for | |
23 | backend I/O, push it into drainEventQueue, and wake up the consumer | |
24 | thread. The consumer thread's run() has the job of dealing with the | |
25 | event, AND THEN calling doIdles and updating the screen. That was the | |
26 | big breakthrough: why bother having main thread do screen updates? It | |
27 | just leads to contention everywhere as it tries to tell the consumer | |
28 | thread to lay off its data structures, when in reality the consumer | |
29 | thread should have been the real owner of those structures in the | |
30 | first place! This was mainly an artifact of the d-tui fiber threading | |
31 | design. | |
32 | ||
33 | So now we have nice flow of events: | |
34 | ||
35 | * I/O enters the backend, backend wakes up main thread. | |
36 | ||
37 | * Main thread grabs events, wakes up consumer thread. | |
38 | ||
39 | * Consumer thread does work, updates screen. | |
40 | ||
41 | * Anyone can call doRepaint() to get a screen update shortly | |
42 | thereafter. | |
43 | ||
44 | * Same flow for TTerminalWindow: ECMA48 gets remote I/O, calls back | |
45 | into TTerminalWindow, which then calls doRepaint(). So in this case | |
46 | we have a completely external thread asking for a screen update, and | |
47 | it is working. | |
48 | ||
49 | Along the way I also eliminated the Screen.dirty flag and cut out | |
50 | calls to CellAttribute checks. Overall we now have about 80% less CPU | |
51 | being burned and way less latency. Both HPROF samples and times puts | |
52 | my code at roughly 5% of the total, all the rest is the | |
53 | sleeping/locking infrastructure. | |
54 | ||
f6d90207 KL |
55 | August 15, 2017 |
56 | ||
57 | I cut 0.0.5 just now, and also applied for a Sonatype repository. | |
58 | It was a reasonable spot: TEditor was working albeit buggy, and a bug | |
59 | had just come in on the main TApplication run loop. So we are about | |
60 | to embark upon some performance work again, it's been probably version | |
61 | 0.0.2 or so since the last cycle. | |
62 | ||
63 | Code size: 40446 lines. | |
64 | ||
65 | Now switching head to 0.0.6 and taking a small break. | |
66 | ||
df602ccf KL |
67 | August 14, 2017 |
68 | ||
69 | TEditor is basically done. Mouse movement, keyboard movement, | |
70 | backspace / delete / enter / etc. are all in. Things are starting to | |
71 | look pretty good. | |
72 | ||
73 | I'm going to prep for a final cut and release tag tomorrow or the next | |
74 | evening. I need to take a break and get some meatspace life dealt | |
75 | with. | |
76 | ||
e8a11f98 KL |
77 | August 12, 2017 |
78 | ||
79 | TEditor is stubbed in about 50% complete now. I have a Highlighter | |
80 | class that provides different colors based on Word text values, but it | |
81 | is a lot too simple to do true syntax highlighting. I am noodling on | |
82 | the right design that would let TEditor be both a programmer's editor | |
83 | (so Highlighter needs to have state and do a lexical scan) and a word | |
84 | processor (where Word needs to tokenize on whitespace). I estimate | |
85 | probably a good 2-4 weeks left to get the editor behavior where I want | |
86 | it, and then after that will be the 0.0.5 release. | |
87 | ||
88 | Finding more minor paper cuts and fixing them: the mouse cursor being | |
89 | ahead of a window drag event, SwingTerminal resetting blink on new | |
90 | input, prevent TWindow from resizing down into the status bar. | |
91 | ||
3e074355 KL |
92 | August 8, 2017 |
93 | ||
94 | Multiscreen is looking really cool! Demo6 now brings up three | |
95 | screens, including one that is inside a TWindow of a different | |
96 | application. | |
97 | ||
42873e30 KL |
98 | August 7, 2017 |
99 | ||
100 | Had trouble sleeping, what with a bunch of imaginative thoughts for | |
101 | this release. jexer.backend will be the ultimate destination for | |
102 | jexer.session and most of jexer.io. TerminalReader will be the | |
103 | interface for keyboard and mouse events. cmScreenConnected and | |
104 | cmScreenDisconnected will be new events to represent a screen | |
105 | appearing/disappearing, and MultiBackend will be a new backend | |
106 | multiplexer that goes full XRandR. Several new demos demonstrating | |
107 | multi-screen support will be coming along. | |
108 | ||
109 | August 6, 2017 | |
110 | ||
111 | Time to clean up more API, particularly between Backend and Screen. | |
112 | Both of these will be interfaces soon, so that one could easily | |
113 | subclass JComponent and implement both Screen and Backend. The | |
114 | original code evolved out of Qodem, where screen.c and input.c were | |
115 | two different things leading to ECMA48Screen and ECMA48Terminal, but | |
116 | now there is really no need to keep them separate. It also | |
117 | complicates the constructors, as these are basically friend classes | |
118 | that have used package private access to get around their artificial | |
119 | separation. | |
120 | ||
121 | When I get this done it should be a lot easier to do any of: | |
122 | ||
123 | * Pass a JFrame or JComponent to SwingBackend and have it add itself, | |
124 | like any other Swing widget. | |
125 | ||
126 | * Construct a SwingBackend and add it to any regular JComponent. | |
127 | ||
128 | * Have multiple TApplications running inside the same Swing | |
129 | application, including having actions affect each other. (Will also | |
130 | need to ensure that TWidgets/TWindows are not in different | |
131 | TApplication collections.) | |
132 | ||
133 | * Build a Backend/Screen multiplexer, so that one could have a ECMA48 | |
134 | TApplication listening on a port and a local Swing monitor for it. | |
135 | ||
136 | * Build a Backend/Screen manager, so that one could have multiple | |
137 | ECMA48 screens acting as a single large screen (e.g. XRandR). | |
138 | ||
139 | Now I need to decide which package will collect Backend, SessionInfo, | |
140 | and Screen. jexer.io has some java.io stuff, so it stays anyway. | |
141 | ||
2fef9c6e KL |
142 | July 28, 2017 |
143 | ||
144 | Got very busy with my meatspace life, now getting a chance to come | |
145 | back around. | |
146 | ||
147 | I gave up on TEditor knowing about graphemes, instead pulling back to | |
148 | simple Cells. This will be better anyway in the long run, as getting | |
149 | grapheme support in Screen someday will also get it for me in TEditor | |
150 | for free. But it does mean that TEditor will chew through much more | |
151 | RAM than it needs to for a text file. Performance optimization will | |
152 | come someday. But this means I can also go back to gcj, because I | |
153 | really like its warnings about unused imports. | |
154 | ||
155 | I've got a POM stubbed in, and created an account over at sonatype. | |
156 | If it isn't too hard, I will try to get 0.0.5 released into the maven | |
157 | universe. But that is still a bit away, I need TEditor running with | |
158 | syntax highlighting first. | |
159 | ||
160 | July 17, 2017 | |
161 | ||
162 | Focus-follows-mouse is in, as is NOCLOSEBOX. | |
163 | ||
8c236a98 KL |
164 | July 15, 2017 |
165 | ||
166 | I think I have cleaned up most of the window show/hide/activate mess | |
167 | in TApplication. Demo4 has some cool interactions between a | |
168 | background TDesktop and several foreground TWindows, which helped | |
169 | expose bugs. | |
170 | ||
171 | July 9, 2017 | |
172 | ||
173 | While working on TWindow.hide/show I decided that I am sick of | |
174 | TApplication's active window handling. TApplication makes lots of | |
175 | assumptions, things are too fragile between modal and not, and one | |
176 | cannot easily say window.activate(). So I will also be changing that | |
177 | too. ... Code is still a bit of a mess, but hooks are in place at | |
178 | least for show/hide/activate. | |
179 | ||
180 | July 8, 2017 | |
181 | ||
182 | Qodem 1.0.0 released last month, I had a vacation, and a Jexer user | |
183 | (nikiroo) started opening up pull requests. :-) So back unto the | |
184 | breach we go! | |
185 | ||
186 | TButton is now animated so that there is some feedback when selected | |
187 | via keyboard. StringJustifier was written which permits TText's to | |
188 | have left/centered/right and full justification. TDesktop is now in | |
189 | too which can act as a permanent max-sized window without borders. | |
190 | ||
191 | Next up is Viewport, an interface to collect scrollbar API, and then a | |
192 | cleaner API for scrollable widgets and windows. After that is more | |
193 | window API: hide/show/maximize/restore, and unclosable windows. I am | |
194 | cherry-picking bits from @nikiroo's PRs, which will likely break them | |
195 | before it fixes things, but I will find some way to get Niki credited | |
196 | with those pieces. | |
197 | ||
e685a47d KL |
198 | March 21, 2017 |
199 | ||
200 | I am starting to gear up for making Jexer a serious project now. I've | |
201 | created its SourceForge project, linked it back to GitHub, have most | |
202 | of its web page set up (looks like Qodem's), and released 0.0.4. And | |
203 | then this morning saw an out-of-bounds exception if you kill the main | |
204 | demo window. Glad I marked it Alpha on SourceForge... | |
205 | ||
206 | Yesterday I was digging around the other Turbo Vision derived projects | |
207 | while populating the about page, and made a sad/happy-ish realization: | |
208 | Embarcadero could probably get all of them shut down if it really | |
209 | wanted to, including Free Vision. I uncovered some hidden history in | |
210 | Free Vision, such that it appears that Graphics Vision had some | |
211 | licensed Borland code in it, so there might be enough mud in the air | |
212 | that Free Vision could be shut down the same way RHTVision was. But | |
213 | even worse is the SCOTUS ruling on Oracle vs Google: if APIs are | |
214 | copyrighted (regardless of their thoughts on fair use), then any | |
215 | software that matches the API of a proprietary project might find | |
216 | itself subject to an infringement case. So that too could shut down | |
217 | the other API-compatible TV clones. | |
218 | ||
219 | Fortunately, Jexer (and D-TUI) is completely new, and has no API | |
220 | compatibility with Turbo Vision. Jexer could be a new root to a whole | |
221 | generation of TUI applications. | |
222 | ||
a7986f7b KL |
223 | March 18, 2017 |
224 | ||
225 | TStatusBar is working, as is "smart" window placement. Overall this | |
226 | is looking quite nice. Found a lot of other small paper cut items and | |
227 | fixed them. It looks absolutely gorgeous on Mac now. | |
228 | ||
229 | Tomorrow I will get to the public wifi and get this uploaded. | |
230 | ||
231 | Time to call this 0.0.4 now though. We are up to 32,123 lines of | |
232 | code. | |
233 | ||
55d2b2c2 KL |
234 | March 17, 2017 |
235 | ||
236 | Jexer is coming back to active development status. I had a lot of | |
237 | other projects ahead of it in the queue, mostly Qodem but also Jermit | |
238 | and of course lots of actual day job work keeping me too tired for | |
239 | afterhours stuff. But here we are now, and I want to get Jexer to its | |
240 | 1.0.0 release before the end of 2018. After that it will be a | |
241 | critical bit of function for IWP and NIB, if I ever get those going. | |
242 | I need to re-organize the demo app a bit so that it fits within 80x25, | |
243 | and then get to TStatusBar. | |
244 | ||
245 | A status bar will be an optional part of TWindow. If it exists, then | |
246 | it will be drawn last by TApplication and get events routed to it from | |
247 | TWindow's event handlers. This will have the nice effect that the | |
248 | status bar can change depending on which window is active, without any | |
249 | real extra work on TApplication's part. | |
250 | ||
251 | Putting together a proper TODO now, with release and regression | |
252 | checklists. I think I will see if jexer is available at SourceForge, | |
253 | and if so grab it. Perhaps I can put together some good Turbo Vision | |
254 | resources too. At the very least direct people to the Borland-derived | |
255 | C++ releases and Free Vision. |